Download: T he Art of Digital Audio Recording

T he Art of Digital Audio Recording The Art of Digital Audio Recording A Practical Guide for Home and Studio Steve Savage With photos by Robert Johnson and diagrams by Iain Fergusson Oxford University Press, Inc., publishes works that further O xford University’s objective of excellence in research, scholarship, and education. Oxford New York Auckland Cape Town Dar es Salaam Hong Kong Karachi K uala Lumpur Madrid Melbourne Mexico City Nairobi New Delhi Shanghai Taipei Toronto With offi ces in A rgentina Austria Brazil Chile Czech Republic France Greece G uatemala Hungary Italy Japan Poland Por...
Author: Thistarry Shared: 7/30/19
Downloads: 371 Views: 3353


T he Art of Digital Audio Recording,

The Art of Digital Audio Recording

A Practical Guide for Home and Studio Steve Savage With photos by Robert Johnson and diagrams by Iain Fergusson, Oxford University Press, Inc., publishes works that further O xford University’s objective of excellence in research, scholarship, and education. Oxford New York Auckland Cape Town Dar es Salaam Hong Kong Karachi K uala Lumpur Madrid Melbourne Mexico City Nairobi New Delhi Shanghai Taipei Toronto With offi ces in A rgentina Austria Brazil Chile Czech Republic France Greece G uatemala Hungary Italy Japan Poland Portugal Singapore South Korea Switzerland Th ailand Turkey Ukraine Vietnam

Copyright © 2011 by Steve Savage

Published by Oxford University Press, Inc. 198 Madison Avenue, New York, New York 10016 w Oxford is a registered trademark of Oxford University Press. All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, e lectronic, mechanical, photocopying, recording, or otherwise, w ithout the prior permission of Oxford University Press.

Library of Congress Cataloging-in-Publication Data

Savage, Steve. Th e art of digital audio recording: a practical guide for home and studio / Steve Savage; with photos by Robert Johnson and diagrams by Iain Fergusson. p. cm. I ncludes bibliographical references and index. ISBN 978-0-19-539409-2; 978-0-19-539410-8 (pbk.) 1. Sound studios. 2. Sound—Recording and reproducing—Digital techniques. I. Title. T K7881.4.S38 2010 621.389'3—dc22 2010032535987654321Printed in the United States of America on acid-free paper, For my daughters Sophia and Thalia. Thanks for all the hours of training!, This page intentionally left blank,


vii Th is book was written because Norm Hirschy at Oxford University Press read something else that I had written. He asked me if I was interested in writing a practical guide to recording, and I was very interested. I had been a professional recording engineer for twenty years, and I had been teaching recording for ten years, so I felt ready to tackle a book of this nature. Norm initiated the project and has nurtured it through each stage—thank you! My fi rst mentor was Brian Risner, who mixed a record I had produced with the artist Bonnie Hayes. Brian had worked extensively with the band Weather Report, and over the course of several projects with him I began to learn how creative the art of recording could be. Brian’s ability to create a very productive and positive environment in the studio—while eff ortlessly handling all the technical requirements—has been a model for all of my work. My good fortune to be teaching in the out- standing Recording Arts program at Los Medanos College has provided the proving ground for much of what is contained here, and it was the site used for much of the photography. I was delighted to discover Iain Fergusson’s diagrams on Wikipedia, and I was able to track him down in New Zealand and engage him to do the diagrams for this book. His work exceeded my expectations and is a model of clarity. Th e diagrams add enormously to the sometimes laborious descriptions of many recording functions. My only regret is that we have yet to meet in person (the joys and vagaries of the Internet)! Robert Johnson is one of the most outstand- ing students to have come through my recording classes, and just happened to be a very accomplished photographer as well. His photographs capture details of the recording process that can only be suggested in words. I was fortunate to have a long-term working relationship with Fantasy Studios in Berkeley, Cali- fornia, and was able to access their spectacular studios and mic closet for ad- ditional photos included here. I was aided in creating many of the screenshots by long-time musical collaborators Curtis Ohlson and Paul Robinson. Curtis runs Digital Performer in his home studio, as well as being a gift ed bass player and producer. Paul Robinson is a Logic user, as well as a wonderfully versatile and talented guitar player. I am indebted to a long list of artists and producers whom I have worked with over the years for all of the wonderful hours we have spent together in the studio. I have attempted to condense something of the breadth of those experi- ences and the joy of making records into these pages., This page intentionally left blank,


ix Introduction • xi 1 Th e Starting Point: Sound Meets the Computer • 3 1.1 Why Computers • 3 1.2 What Does It Sound Like? • 6 1.3 Signal Path • 7 2 Th e Essentials: Where and How Recordings Are Made • 10 2 .1 Recording Rooms and Control Rooms • 10 2.2 Studio Monitors • 14 2.3 Microphones and Mic Placement • 18 2.4 Mixing Boards and Control Surfaces • 29 2.5 EQ: General Information • 44 2 .6 Dynamics (Compressors and Noise Gates) • 55 2.7 FX: Delays • 67 2.8 FX: Reverb • 69 2.9 Beyond Traditional DSP • 71 3 Recording Sessions: A Practical Guide • 76 3.1 Setup • 76 3.2 Headphone Mixes • 89 3.3 Survey of Recording Techniques for Instruments and Voice • 93 3.4 Beyond • 117 4 Editing: Th e New Frontier • 119 4.1 Editing Features • 119 4.2 Screen “Real Estate” • 130 4.3 Virtual Tracks (Playlists) • 132 4.4 Advanced Editing • 138 5 Th ree Fundamentals: Techniques Every Recordist Needs to Know • 154 5.1 Inserts/Plug-ins • 154 5.2 Send and Return • 159 5.3 Auto-switching (Auto-input) • 165 6 Mixing: Th e Most Creative and the Most Challenging Stage • 170 6.1 Mixing Requirements • 171 6 .2 Building a Mix • 176, THE ART OF DIGITAL AUDIO RECORDING 6 .3 Automation and Recall • 199 6 .4 Mix Collaboration, Communication, and Delivery • 205 7 Mastering: One Last Session • 210 7.1 What, Why, How, and Where • 210x8Th ree Best Practices: Easy Ways to Raise the Level of Your Sessions • 222 8.1 Session Flow • 222 8.2 Talkback • 228 8 .3 Playback Volume • 234 Addendum 1 How to Walk into a Commercial Studio and Be the Engineer • 239 Addendum 2 Researching and Buying Gear: Internet vs. Brick and Mortar • 242 A ppendix Digital Audio Formats, Delivery, and Storage • 246 Online Glossary Link • 254 Index • 255,



About Th is Book

Making great recordings requires striking the right balance between techni- cal know-how and a practical understanding of recording sessions. Even in the digital age, some of the most important aspects of creating and recording music are completely nontechnical and, as a result, are oft en ignored by traditional recording manuals. Getting the best audio recording results oft en requires as much common sense and attention to the recording environment as it does a deep understanding of the technical elements involved. Too many books about recording provide technical information but don’t supply the practical context for how and when to apply the tools and techniques described. Th is can leave the reader without a sense of priority, trying to fi gure out what is actually important to the recording process in specifi c situations. Th e Art of Digital Audio Record- ing c an teach readers what they really need to know to make great-sounding recordings with their computers—the essential practical, as well as technical, information, including: • What to look and listen for in your recording environment • Straightforward advice on recording almost any instrument • Th e essentials of digital audio workstations (DAWs) • Th e essentials regarding recording gear: microphones, mixers, and speakers • Th e fundamentals of understanding and applying EQ, compres- sion, delay, and reverb • Th e secrets to running creative recording sessions • Th e practical application of digital editing, mixing, and mastering • A special section that identifi es the most common challenges of the recording studio. • Addendum: • How to walk into a commercial studio and be the engineer • Researching and buying gear: Internet vs. brick and mortar. • Appendix • Digital formats, delivery, and storage Th e Art of Digital Audio Recording is a reference manual for the home recordist, a textbook for any basic to intermediate DAW training class, and a primer for the musician who is either doing his or her own recordings or simply wishes to be better informed when working in the studio., THE ART OF DIGITAL AUDIO RECORDING

About the Author

My personal path into recording and audio production, and from there to this book, began with a career as a drummer. I played in numerous unsuccessful xii rock bands, learned some jazz without ever coming close to mastering it, stud- ied and performed African music with a master drummer from Ghana, and spent a couple of years actually making a living as a musician, playing in a dance band. Aft er a short but glorious stint in a punk band, my career transitioned into recording and production. I discovered that the other side of the glass—the control room rather than the recording room—fi t me better, and my career slowly built up around re- cording. I had a 12-track studio in my garage (equipped with the short-lived Akai recording format) and recorded demos for rock bands for dirt-cheap. One of those bands put its resources together to go into a professional studio to record a single and asked me to be the engineer/producer. Th ere, I got my fi rst taste of making commercial recordings and I was hooked. I recorded a variety of fl edgling “new wave” artists’ singles and albums in the heady early 1980s, and I cut my teeth on 24-track analog recording. Aft er a stint as house producer for a small indie label—where I built and learned to operate a lovely little state-of- the-art SSL studio (Solid State Logic makes some of the best and most expen- sive consoles and control surfaces)—I became a full-time independent record producer and engineer. One tends to get work in areas where one has some successes, so it was through my work with the very talented songwriter Bonnie Hayes that I have ended up working on many singer/songwriter music projects, and aft er three Grammy-nominated CDs with the master blues artist Robert Cray, I have had the pleasure of working on many blues records. I have also recorded jazz, R&B, rap, hip-hop, country, opera, music for musicals, and children’s records. I have been the engineer and/or producer on over 100 commercial releases and have served as the primary recording engineer and mixer on seven Grammy- nominated CDs, including records for Robert Cray, John Hammond Jr., Elvin Bishop, and Th e Gospel Hummingbirds. I have also taught recording in the Re- cording Arts Department at Los Medanos College in Pittsburg, California, one night a week for the past ten years. Th is book is a result of those experiences, both in the studio and in the classroom, along with the countless hours read- ing various books, trade magazines, and (increasingly) Web sites that provide an endless supply of information and opinion about the world of recording. Th rough it all, it is my love of music that makes me love my work. I am deeply grateful for the opportunity to have participated in the making of recordings with so many talented artists., T he Art of Digital Audio Recording, This page intentionally left blank,

C hapter 1 The Starting Point Sound Meets the Computer

1 .1 Why Computers Th e title of this book is Th e Art of Digital Audio Recording, but it will be appar- ent to even the most casual reader that the book covers a wide variety of topics that extend beyond the specifi cs of computer-based, digital recording. None- theless, the title indicates this book’s orientation and that all of the informa- tion here is presented primarily in the context of the digital audio workstation (DAW). Even the most basic recording practices have been infl uenced by the migration from analog to digital recording, and this book maintains its focus on computer-based audio production throughout. W hile I don’t think I need to convince you that audio production is domi- nated by computer-based systems, analog gear remains an important part of the recording process. Aft er all, sound itself is an analog phenomenon—created by disturbances in the air—and certain elements such as microphones and speak- ers remain essentially analog. With other primary recording technologies, such as EQ, the debate regarding preferences for analog versus digital gear is not over (and probably never will be), despite the fact that digital dominates almost every recording environment today. But wherever you stand on the aesthet- ics of analog versus digital, it is valuable to examine why DAWs represent the standard in contemporary audio production. By detailing the primary advan- tages of DAW recording over its analog predecessors, I set the context for the remainder of this book. A brief survey of the primary audio practices includes recording, edit- ing, signal processing, mixing, and mastering. In each of these areas, the DAW, THE ART OF DIGITAL AUDIO RECORDING has introduced revolutionary capabilities. Th e most fundamental change from analog production has come in the nondestructive capabilities of DAW record- ing and editing, but signal processing, mixing, and mastering have also seen dramatic changes in the digital world.


D AWs generally record to hard drives, which allow data to be stored in any available area of the medium. Th ere is no “erase” head on a DAW recorder— which is to say that it is no longer necessary to erase (or destroy) previous re- cordings when making new recordings. As long as there is drive space available, further recordings can be made. With the enormous capacity and relative low cost of current hard drives, this eff ectively means that no recordings need ever be eliminated. A long with doing away with the need to ever erase anything, nondestruc- tive recording has transformed the recording process by allowing for many more recorded elements to be available in any given project. As you will see in more detail in chapter 4, when I explore virtual tracks, nondestructive record- ing changes the way people work with audio in more ways than just eliminating the problem of running out of analog-tape tracks. Whole new working proce- dures have evolved within the nondestructive environment of the DAW. One such example is the way that nondestructive audio has transformed one of the most basic production practices: punching-in. Punching-in typically involves the rerecording of parts of previously recorded elements. A common example is replacing a line from an already recorded vocal performance. On an analog tape recorder, punching-in required erasing what was previously re- corded. Th is sometimes led to diffi cult decisions about whether it was worth losing the previous performance in the hope of getting something better. Ana- log punching-in also involved the potential risk of accidentally losing parts of the recording, because the beginning or ending of material around the part to be replaced might get clipped off if the punch-in was not done accurately enough. With nondestructive recording, these problems have been eliminated. Parts of recordings may be replaced without losing (erasing) the part that has been replaced; you never actually have to “record over” any element, as each element remains stored and accessible from the hard drive. Also, accidental “punches” (recordings) don’t eliminate previously recorded material for the same reason—the process is nondestructive so nothing is actually lost. Nonde- structive recording has eliminated many of the most basic limitations of the analog recording process .


In regard to editing, new capabilities in the DAW are even more signifi cant than the changes DAW brought to recording. Th e nondestructive quality of DAW-, Th e Starting Point based editing provides vast new opportunities for audio manipulation. With nondestructive DAW editing, you simply create alternative instructions as to how to play back the audio that has been recorded. Because the manipulation of audio in a DAW is separate from the storage of that audio on the hard drive, you can edit without altering the original recording. Th is is a major improve- 5 ment over tape-based editing, which required the physical cutting and splicing of tape. Not only do you no longer endanger the storage medium by cutting tape, you are able to edit much faster and in many more fl exible ways than ever possible with tape splicing. Whole new recording and working procedures are now built around these editing capabilities. I explore this new world of editing capabilities in much greater detail in chapter 4.

Signal processing

S ignal processing has also been transformed by the DAW, though that has been a slower process of change than with recording or editing. Digital EQ, dynamics processing (compression, etc.), and ambient eff ects (reverbs, delays, etc.) oper- ate in much the same way as they did in the analog world. While it has taken a considerable amount of time and development to produce digital equivalents of these signal processors that compare in quality to their analog relatives, they have fi nally arrived, though whether they are truly a match for the best of the analog versions is a still very much debated. Th ese processors were already used nondestructively in analog production—applied to already recorded signals and easily altered or removed at any time. Th e big changes in signal processing have come with wholly new capabilities that were not at all available in analog. Th ese include the ability to speed up or slow down audio without changing pitch and the ability to analyze and alter the subtleties of pitch with tools such as Auto-Tune. Th ere are also an increasing number of processing tools that oper- ate based on a detailed analysis of audio content that is available only through computerized technology. I look more thoroughly at some of these develop- ments at the end of chapter 2, when the discussion goes “beyond” the familiar kinds of signal processing.

M ixing

Th e DAW has advanced the kinds of control over the mixing stage—controls that were begun when automation and recall began to be implemented in ana- log consoles. Automation allows for the “automatic” replaying of changes in volume and other typical mixing moves, while recall enables the recordist to regain all of the mix settings at a later time—in order to revise mixes. Suffi ce it to say that even the early implementation of automation and recall in the analog realm required the interfacing of a computer to control these functions. Now that the entire mixing process may be computer based, the implementation of automation and recall have become much more elaborate and also more reliable., THE ART OF DIGITAL AUDIO RECORDING Th e DAW has also vastly improved the ability to automate mixing moves off - line, using a graphic interface that provides extremely fi ne control over desired changes. Th ese features and the evolution of mixing in the DAW are covered thoroughly in chapter 6.


Th e fi nal stage of production—mastering—prepares the fi nal mixes for manu- facturing. Th e combination of digital delivery (from CDs to mp3s and beyond) and DAW production has meant that just about anyone can create a master that is usable for CD manufacturing or online delivery. Th e large lathes required to create vinyl LP masters are still used for that format, but that has become a very small part of the audio marketplace. New tools for mastering to digital formats such as CDs have resulted in what many believe to be both a blessing and a curse—a blessing for the technologies that allow CDs to sound better than ever, and a curse for the ability to overuse some of these technologies at some signifi cant cost to the original musical dynamics. All of these techniques and controversies are covered in chapter 7. It is noteworthy that books such as this one now cover the practical application of mastering techniques for a broad audience, as these technologies have only recently become available outside of what was once a very specialized (and expensive) mastering facility.

Digital versus analog

Th e overwhelming advantages of DAW production have resulted in the pre- dominance of computer-based audio production in both amateur and profes- sional music recording. Still, this leaves the question: Does digital sound better or worse than analog? Th e wide range of opinions you fi nd in a typical audio discussion group suggests that there is no one answer to this question, though I would maintain the following: (1) Th ere are so many factors in creating good- sounding audio (and even in defi ning what is meant by “good-sounding”) that the analog/digital divide is a relatively small element in the overall mix of fac- tors pertaining to quality; and (2) like it or not, we live in a digital audio world and most of us will spend most of our time recording, editing, processing, mix- ing, and mastering audio in a DAW! 1 .2 What Does It Sound Like? While many things in the digital domain are held over from the analog era, at the same time much has been changed by the DAW environment. For all the changes, one thing—the most important thing—remains the same. Th is is the guiding principle in audio production: W hat does it sound like ? Th ese are the words spoken by Ray Charles in the extraordinary documentary Tom Dowd & the Language of Music, which traces Dowd’s remarkable career in audio produc-, Th e Starting Point tion. Ray is summarizing his point of view about recording and expressing his aff ection for Tom Dowd, who shared his passion for sound. Ray reminds us to keep the focus where it belongs, on the sound, instead of on preconceived or technically drilled notions of what “proper” technique is. Aft er all, it is only the sound of the recording that the listener hears. 7Sothroughout this book, while the bulk of the time is spent on the tech- nicalities of recording, I have tried not to lose sight of this much more subjec- tive and much more important element in audio production: creative listening. Th ere’s a saying in jazz that in order to play “outside,” you must fi rst learn to play “inside.” Th is means that the important business of pressing the boundar- ies and breaking the rules works best when the boundaries and rules are well understood. As with playing music, the art of recording music requires that rules be broken, as well as followed; and as with music, the better the rules are understood, the more eff ective will be the bending and breaking of those rules. So dive into the technique and the theory, but don’t forget to come up for some creative breaths of fresh air! 1.3 Signal Path T echnically speaking, the entire job of a recording engineer is summed up in these two words: signal path . Th e engineers are responsible for what is happen- ing to audio from the beginning to the end—from the creation of the sound waves by the musician playing his or her instrument to the recreation of the sound waves by the speakers in the listener’s living room. You might pick up and/or leave the audio chain at intermediate points—perhaps starting as sam- ples used in drum loops and ending when you turn the project over to a mixing or mastering engineer—but in any event, when you work on sound you work within the context of a signal path. One of the fi rst challenges of signal path is simply getting the sound from one place to the next. Getting the sound from the microphone to the recorder and from the recorder to the playback system can be a challenge in itself. Add a lot of processing gear, such as compressors and EQs, and monitoring demands, such as headphone mixes for musicians, and setting up the correct signal path can be complicated. I can’t cover all the contingencies here, but there is much more said about signal path in almost every section of this book. Here, at the beginning, I lay out some basics.

Input and output (I/O)

T o start with, signal path is controlled by the most essential technical element in audio production: i nput and output (oft en shortened to I/O). Following the audio’s signal path (also referred to as s ignal fl ow ) is the same as following a se- ries of inputs and outputs, and it is oft en referred to with another essential audio term, r outing . I/O routing can be pretty straightforward in some cases. For ex-, THE ART OF DIGITAL AUDIO RECORDING DIAGRAM 1.1 A simple signal path: DAW mic input to speakers ample, in a system where the DAW interface has a microphone preamp built in, the signal path may be as simple as: sound source inputs to microphone, microphone outputs to the mic input of the DAW audio interface, the interface outputs to the computer soft ware that then handles the signal path until it is output back to the interface, and from there output to the playback system. In this example, assuming the DAW interface is already set up, the only external connection the engineer might have to make is connecting the mic to the mic cable and the other end of the mic cable to the audio interface. On the other hand, the signal path’s I/O routing may be very complicated, involving multiple inserts, patch bays, talkback systems, cue systems, and so on; and each of these may be either hardware of soft ware based (or both)! All of these topics are considered later in this book, with the focus on the soft - ware/DAW side, but it is not possible for any book to cover all possible routing schemes. What’s more, the internal routing systems within each brand of DAW may diff er in both terminology and implementation. You will have to learn the I/O intricacies for your own setup, but it is most helpful to begin with this basic understanding: every thing you do starts with signal path, and signal path is de- fi ned by the input and output routing series . Th e I/O model of signal path is also in operation on a micro scale within each dedicated audio element, from stomp box to DAW. You may have seen schematics for individual pieces of gear or computers; they are complex grids of inputs and outputs. Audio engineers do not necessarily need to be familiar with the internal workings of audio or computer hardware, though sometimes that knowledge can be helpful. In any event, a strong understanding of signal fl ow between gear and within soft ware is essential for making good recordings.

T roubleshooting

Troubleshooting—an unfortunate but inevitable part of every engineer’s job—also starts with signal path. Th e best way to troubleshoot most technical problems is to investigate each step of the signal path, starting with the sound, Th e Starting Point source, in order to determine where the problem lies. Whether it’s poor-sound- ing audio, noisy audio, or simply no audio at all, the problem lies somewhere along the signal path. A systematic approach that examines the I/Os from the beginning of the chain is the best and most effi cient approach to solving almost all technical problems. 9

Combining the technical and the aesthetic

R ecording always entails fi nding the proper balance between creative and tech- nical demands. Considering the question “What does it sound like?” takes you to the essence of the creative process—ultimately, that is all that matters. Un- derstanding the basis of signal path takes you to the essence of the technical process; these are the nuts and bolts that must serve the aesthetic. With this grounding in both the aesthetic and the technical, you are ready to tackle some much more specifi c elements in audio production, beginning with the essen- tials of where and how recordings are made.,

Chapter 2 The Essentials Where and How Recordings Are Made

2.1 Recording Rooms and Control Rooms Th is opening section is going to be relatively brief—there are many other re- sources for delving more deeply into the technicalities of acoustics. For most of us, the idea of constructing a space for recording is not part of our work. We recordists are either stuck with certain spaces because we need to work there or perhaps we live there, or we choose to work at studio spaces based on expe- rience or reputation. Nonetheless, there are some fundamentals about sound and space that every recordist should be familiar with, and some helpful ways of dealing with basic problems. I summarize the issues concerning the physi- cal space that we work in, dividing them into three basic topics: isolation, fre- quency response, and ambient characteristics. IsolationInregard to isolation, there are two main considerations and one basic rule. Th e things to consider are isolation from outside noise leaking in, and isolation of inside noise leaking out. Either or both may be problematic, but the solution for both—the one basic rule—is the same. Th at rule is that isolation is created by a combination of mass and density. Th at is to say, the way sound leakage (in either direction) is prevented is with suffi cient mass that is suffi ciently dense. What this means in practical terms is that a 12-inch-thick wall of dense con- crete will isolate sound much better than a typical wall with two sides of sheet- rock and an air cavity in between. Studios in highly problematic environments have been known to resort to sheets of lead as part of the wall structure. Th is, Th e Essentials can work well, but can also be very expensive. If you are fortunate to work in an environment with little external noise and without sensitive neighbors, you may have far fewer concerns about isolation. If not, density and mass are your primary allies. Th ere is sometimes the notion that more absorption inside a room (from acoustic panels, to foam, to rugs, to egg cartons) will help solve leakage prob- lems. Unfortunately, this not the case because materials that absorb sound do 11 so primarily in the higher frequencies and leakage decreases as the frequencies rise. Th at is why, if you are standing outside a rehearsal studio with a rock band playing inside, what you hear is primarily the bass guitar and the kick drum. It is the low frequencies that permeate walls, mess with recordings, and anger neighbors, no matter how much dampening material you have inside the room. Only mass plus density will do an eff ective job of decreasing low-frequency transmission. Isolation does have an eff ect on the sound in the room, as well. Th e more low frequencies are prevented from escaping because they are refl ected with suf- fi cient density and mass (such as a concrete wall), the more problems with bass buildup within the room itself. Solving transmission problems to and from the outside also engages you in absorption and refl ection issues within the room. Th ere are many other technical elements that will aff ect transmission, re- fl ection, and absorption; and there are a variety of books that describe com- mon approaches to designing and constructing walls, fl oors, ceilings, doors, windows, and HVAC (heating/venting/air-conditioning) systems for recording studios. Th ese topics are beyond the scope of this book, but very much worth exploring if you are building or remodeling a space to be used for recording.

Frequency response of a room

Th e frequency response of a room refers to the way diff erent frequencies, from low to high, respond to the absorptive and refl ective qualities of room surfaces. Every room has diff erent frequency responses—the room’s physical character- istics cause boosts or dips at certain frequencies—and these are variable to a certain degree, depending on where you are in the room. Generally, a room with relatively even frequency response across the spectrum is desirable, and this can be achieved by controlling the absorption and refl ection of sound in the room. Th ere are some basic principles in this regard, though the details of designing and controlling room acoustics can get very complex and the results are never thoroughly predictable. Th ere are two main enemies of a smooth and even frequency response. Th ese are right-angle corners and parallel surfaces. Right angles, such as at most wall-to-wall, fl oor-to-wall and ceiling-to-wall intersections, will refl ect sound back in the same direction as it has come from and will cause the most promi- nent frequencies of the original sound to build up, disrupting an even frequency, THE ART OF DIGITAL AUDIO RECORDING response. Opposing parallel walls (or fl oor and ceiling) create s tanding waves by refl ecting the sound back into its own original path. Standing waves also amplify certain frequen- cies and disrupt an even frequency response. Unfortunately, most typi- cal room construction uses a lot of right angles and parallel surfaces. Bass frequency buildup and other unwanted room resonances are an especially common problem that may be made worse by right an- gles and parallel surfaces, but they are not necessarily eliminated by a PHOTO 2.1 and 2.2 room with neither of those design characteristics. A whole world of Various wall treatments “bass trap” solutions has evolved, and there is some debate as to how eff ective any or all of these solutions may be. Th ere are companies that spe- cialize in products to aid in improving room acoustics without your having to tear down walls and rebuild. Th ese are defi nitely worth exploring unless you are working in an already well-designed acoustic environment. Most home and project studios need some acoustic treatment. B esides creating problems, room refl ections can be used to help solve problems. While many room frequency imbalances caused by refl ections may be solved using absorptive material, too much absorption can make a room, Th e Essentials P HOTO 2.3 Diffuser sound “dead,” and that may not be desirable either. For many recording appli- cations, the recording environment works best when it is enhancing the natu- ral acoustics of the musical instruments. Th ere has been a trend toward using diff users to balance the frequencies of room refl ections. Diff users are specially built wall treatments that break up frequencies and scatter them to reduce un- wanted frequency buildup. Th e physical dimensions of the wells of the diff user (width and depth) determine the frequencies that are aff ected. Diff users have the advantage over absorption materials in that they don’t make rooms exces- sively dead sounding, but absorption materials can eliminate some problems too severe for diff users to manage. Th e best solution for treatment of critical audio spaces usually involves a combination of absorption, bass trapping, and diff usion.

Room ambience (reverberation)

Th e ambient characteristics of a room refer to the quality and length of the delays created when sound is produced in the room. R everberation is the audio term used to describe these characteristics. It is the refl ections of sound off all of the various surfaces in a space, returning with varying degrees of intensity and delay to the listener, which create reverberation. Th e ambience created by room acoustics is the “natural” reverb, whereas the addition of “artifi cial,” or simu- lated reverb, will be covered later in this chapter (section 2.7). As noted, room acoustics may create problems for recordings (standing waves, bass buildup, etc.) or may enhance recordings by the addition of a pleasing spatial quality. Using microphones to capture the ambient characteristics of a room is covered later (section 2.3)., THE ART OF DIGITAL AUDIO RECORDING Th e reliance of recordings on room acoustics for ambience var- ies enormously. Vocals are oft en recorded in small booths with lots of absorptive material on the fl oor, walls, and even ceiling. Th e 14 microphone is close to the singer’s mouth, so the minimal room re- fl ections are virtually nonexistent relative to the direct sound of the voice. In contrast, many orches- DIAGRAM 2.1 tral recordings are made primarily Refl ecting sound using microphones at some dis- tance from the orchestra, and the room ambience is a major portion of the sound that is captured along with the direct sound from the instruments. Along this continuum lies the world of aesthetic decisions about how to place the musicians and microphones and how to capture or minimize the eff ect of room acoustics on recordings. Such decisions begin with your feelings about the particular acoustics of the room you are recording in. In most instances, it is impossible to completely separate decisions about how to record from con- siderations regarding room acoustics, so the aesthetics of recording are always intertwined with the sound of the recording room.

Control room acoustics

For many home recording environments, there is no diff erence between the re- cording room and the control room—that is, they are the same room. Th is can be a workable recording situation, but it does challenge the acoustic priorities of the two functions—recording and listening. In general terms, it is desirable to minimize the eff ects of room acoustics in the listening environment (control room), whereas room acoustics are oft en used to enhance the recording envi- ronment (studio room). Th ose using one-room studios inevitably have to seek some compromise between these two priorities. Certain trends have encour- aged a relatively easy mix: using more diff usion in control rooms has made them more “live” sounding without too many frequency irregularities. Th is in- creases the aesthetics of listening compared to overly dead rooms and makes the room more suitable for recording, as well. 2.2 Studio Monitors Studio monitor speaker selection and placement are critical to your work envi- ronment. Your primary studio monitors—usually the near-fi eld speakers—are your most consistent and important reference point for what your recordings, Th e Essentials sound like. Th ere are a variety of factors to consider in achieving a monitoring environment that you can trust as reasonably accurate.

Near-fi eld monitors

Near-fi eld monitors have long been the principal means of limiting the eff ects of room acoustics on listening. Th ey also provide a better reference to the “real” 15 world of consumer speakers, which will be what is used by most of those who listen to your recordings. Th at said, it should be remembered that no speakers eliminate the eff ects of room acoustics, no matter how near to your ears they are, and no speakers can give you a complete picture of what your recordings are going to sound like out in the real world, because of the wide variety of play- back systems (and problems). Studio monitors diff er from most consumer speakers in their basic phi- losophy. Studio monitors seek a balanced sound, whereas consumer speakers oft en enhance frequency ranges, eff ectively “hyping” the sound for the lis- tener (most oft en with high- and low-frequency boosts). Despite the inten- tions for studio monitors to be “fl at” across the frequency range, this ideal is impossible to achieve. Inevitably, speakers have some variation in frequency response across the spectrum, and crossover points (between the woofer and tweeter or other speaker combinations) provide the greatest challenges in speaker design; they are always compromised in some ways. Th is is why so many studio monitors are two-way speakers—the more crossovers, the more potential problems. Th e overall sound of the speaker comprises its timbre characteristics. Th ese can be described in various ways, but typically you might judge speakers on a scale from smooth to harsh. You might think, “Th e smoother, the better,” but not all recordists would agree. Some fi nd that speakers that have very smooth timbre characteristics don’t necessarily translate that well to a wide range of other playback systems. Smooth timbre is good for long listening sessions, but a slightly harsher timbre characteristic might be more “real world”—have more in common with the majority of lower cost consumer playback systems—and therefore translate better in more circumstances outside the studio. I fi nd that some of the fi nest speakers have a tendency to lull me into a false sense of secu- rity—everything sounds good!—so I prefer studio monitors that have a bit of a bite to them, though not too much bite so that they can be listened to for long periods of time with minimum ear fatigue. When making live recordings in a one-room studio, it is usually necessary for everyone to use headphones (no speakers), so as to limit bleed from the speakers back into the recordings and to prevent feedback. Th is requires quite a bit of switching back and forth between headphone listening (to record) and speaker listening (to get a better sense of the sound of the recording), but it can be a workable situation. It is important to reference your recordings on your, THE ART OF DIGITAL AUDIO RECORDING speakers, and not to make sonic judgments based solely on monitoring with headphones.

P owered studio monitors

Th ere has been a growing tendency for studio monitors to come in powered 16 versions—that is, the power amp is built right into the speakers (some manu- facturers only make powered speakers). Th e motivation for this is simple: pow- ered speakers ensure that the amplifi cation for the speaker is properly matched to the speaker design and capabilities. In general, this is a very good develop- ment; the only real drawbacks are that it makes the speakers more expensive (though they do have to be powered one way or another, anyway), and it makes them heavier (which can be a bit unfortunate if you are traveling between stu- dios and like bringing your speakers with you). I recommend getting powered studio monitors, if possible.

Near-fi eld monitor setup

Positioning of near-fi eld monitors is an important part of getting an accurate representation of the recorded sound. Th e basic rule is that the speakers should be the same distance from you as they are from each other, creating an equi- lateral triangle. Th is arrangement provides the optimal stereo imaging. If the speakers are too close to each other, the stereo fi eld will sound collapsed; if they’re too far apart, it will sound unnaturally spread out. Th e speakers should be angled toward you (though some recordists like them to point slightly be- hind their head to lessen fatigue). Proper aiming of the speakers aff ects the perception of the stereo image and reduces frequency smearing. Th e speakers should be isolated (decoupled) from whatever they are sitting on. Th e best way to do this is with speaker pads such as those sold by Auralex. If the speakers are not isolated, the sound will be transmitted through whatever they are sitting on and it will arrive at your ears prior to the direct sound from the speaker (sound travels faster through solid material). Because the sound is arriving at a diff erent time, there will be phase problems. It is generally recommended that you set up your playback system along the longer wall of your room so as to minimize refl ections off the side walls, but if your room is very narrow, the refl ections off the back wall might be a bigger problem and you would be bet- DIAGRAM 2.2 ter off setting up facing the narrow side. Refl ections off of your console, desk, or Near-fi eld monitor setup tabletop might also create phase prob-, Th e Essentials lems. Th is can be minimized by angling the speakers up slightly with the tweet- ers pointing at or just behind your ears. You can also experiment with using extenders to move the speakers closer to you or use stands to move them back if you feel as if you’re getting to much refl ection from the work surface. Similarly, refl ections off the wall behind the speakers or from corners will create phase problems, so it’s best to keep the monitors somewhat out in the room and away from walls. 17 WHAT NOT TO DO Do not ignore the basics of speaker placement. Do not place near- fi eld speakers up against a wall or in a corner. Be sure that your speakers are isolated from their mounting surface. Take care to have your speakers placed at an equal distance from the listening position.

C hoosing near-fi eld monitors

Probably the most infl uential element in the eff ectiveness of near-fi eld monitor- ing is the familiarity of the recordist with the speakers. Consider the informa- tion above and then fi nd speakers you like and stick with them. It’s best if you can go to a studio-equipment dealer and audition a bunch at once. Over time you will be able to really trust what you hear from the speakers because you are familiar with them. Eventually you will have heard a lot of diff erent instruments and music through your speakers, and also have had the chance to hear your mixes on a variety of systems. It’s important that the speakers and the room have a reasonably fl at response, and that they be positioned properly, but be- yond that, it is familiarity that will serve you best.

Large monitors

A lmost all critical listening is done on the near-fi eld monitors. Large (wall- mounted or soffi t-mounted) speakers are nonetheless useful for a variety of other purposes. Large monitors may be used for referencing low frequencies that may not be suffi ciently reproduced in the near-fi eld monitor, though sub- woofers have become a common alternative for doing this. I generally use large monitors for playback when musicians are recording live in the control room, if they are used to hearing their instruments rather loud, such as with electric guitar players in many rock bands. When there are no problems with leakage or feedback, such as when a guitar player is in the control room but his or her amp is isolated in another room, it can be very convenient to have the musician playing in the control room. Th is bypasses the use of a talkback system, making, THE ART OF DIGITAL AUDIO RECORDING communication between you and the musician easier. (I explore this practice more thoroughly in chapter 8, on best practices.) Large monitors are also useful for impressing clients; there’s nothing quite like loud playback over big, high- quality speakers, done preferably at the end of the session so as to avoid too much ear fatigue. Big monitors can be useful for more general listening evaluations if they’re 18 accurate across the frequency range, but this is not easy to accomplish. Large monitors are typically farther away from the listening position, so they interact much more with room acoustics than near-fi eld monitors, and this oft en causes complications in achieving a well-balanced frequency response. Large monitors also usually need to be wall or soffi t mounted, and this also can cause problems as the sound interacts with the walls. As a result, it is almost always necessary to EQ the large speakers to fi x unbalanced frequency response. To do so properly requires “shooting the room.” Th is is done by broadcasting and measuring vari- ous kinds of noise (white noise, pink noise, etc.) through the speakers and cap- turing it with a well-balanced microphone, reading the results via a spectrum analyzer, and adjusting the frequency balance accordingly, using EQ. It sounds scientifi c, and it is up to a point, but the variables are enormous: small varia- tions in mic placement can cause diff erent readings, and so on. Shooting a room has become a highly developed craft , with a variety of tools available to aid in the process and with certain practitioners gaining reputations for producing particularly pleasing results. Th e same set of speakers in the same room can end up with pretty diff erent EQ curve corrections, depending on who “shoots the room.” 2.3 Microphones and Mic Placement Microphones are oft en at the beginning of the recording chain, and there are an enormous number of microphone brands and types to choose from. Th ere may be no more important element in many recording situations than the selection and placement of microphones. Th ere are complete books about microphones, but here I focus on the practical side of the most common kinds of studio mi- crophones and their uses.

Microphone types

T here are two types of microphones used the majority of the time for record- ing: condenser mics and dynamic mics. Condenser mics use a diaphragm that vibrates next to a solid backplate and the mic measures the electrical charge of the movement of the diaphragm relative to the backplate, chang- ing these measurements into an electrical representation of the sound. Condenser mics require external power, called phantom power, which is supplied as an option by most mic preamps. Dynamic mics, which are also referred to as moving coil mics, capture sound by using a coil attached to, Th e Essentials the diaphragm that is vibrated in a magnetic field by the movement of the diaphragm. The moving coil creates an electrical current that is a represen- tation of the sound. Here is a list of the primary differences between con- denser and dynamic mics: C ondenser Mics: • R equire external (phantom) power 19 • P rovide the greatest detail of frequency response • Respond quickly to capture leading-edge transients • May be sensitive to loud sounds • Are somewhat fragile Dynamic Mics: • Do not require external power • Provide less detail than condenser mics • Do not respond as quickly to transients • Are able to withstand loud sounds • Are quite rugged Th ere are two primary types of condenser microphones: large-diaphragm con- densers and small-diaphragm (pencil) condensers. Th e primary diff erences be- tween the two are: L arge-diaphragm Condensers • Have less self noise and high output • Have slightly diminished high-frequency response • May have poor frequency response for off -axis sounds • May have multipattern switching capabilities Small-diaphragm Condensers: • Have slightly more self noise and lower output • Have a slightly extended high-frequency response • Tend to have pleasing off -axis capture capabilities • Most versions require changing capsules to achieve diff erent patternsOnthe basis of this information, you can understand why condenser mi- crophones are used most of the time in the studio. Th e exceptions come pri- marily when the sound to be recorded is too loud for the sensitive condenser capsule. Th e most common application for dynamic microphones in the studio is for drums and for miking electric guitar amp speakers. However, this is de- ceptive, as there are now many new designs of condenser microphones that can withstand high volumes, yet dynamic mics are still most oft en used for drums and guitar amps. And dynamic mics are sometimes used for almost every other kind of studio recording, including vocals. Th is is because fi delity—breadth and detail in frequency and transient response—is not the only consideration in choosing microphones. Th ink back to the “What does it sound like?” criterion,


S ome common large- diaphragm condensers, left to right: Telefunken U-47, AKG 414, Neumann TLM- 103, Neumann U-87


Some common small- diaphragm condensers, left to right: Sony ECM-22P (electret), Neumann KM- 84, AKG C452, Bruel & Kjaer (B&K) 4011

P HOTO 2.6

Some common dynamic mics, clockwise: Electrovoice RE-20, Shure SM-57, Shure SM-58, Shure Beta-58, Shure SM-7, AKG D112, Sennheiser MD-421, Th e Essentials from the last chapter; there’s a preference for the sound of a less detailed, lower fi delity microphone in certain (sometimes many) studio applications. Th ere are microphones with technologies other than those used by tradi- tional condenser or dynamic mics, such as ribbon mics, PZM mics (pressure- zone microphones), specialized technologies for miniaturized mics, shotgun mics, and so on. Ribbon microphones, which are a variation on a dynamic mic, have been gaining in popularity and there have been advances made in their 21 ability to withstand higher volume levels and to be more rugged. Th ey have become fairly widely used—especially on guitar amps, as well as for reed and brass instruments—as a result of their balancing the warmth of a dynamic mic and the detail of a condenser mic.

M icrophone patterns

Th ere are two primary mic patterns: cardioid and omni-directional. C ardioid mics have a directional pickup pattern, meaning they are optimized to pick up sound coming from within the bounds of a directional pattern. Th ese provide excellent fi delity from sounds oriented within the pickup pattern and consider- ably lesser fi delity for sounds that might be coming off -axis (response to sounds coming from a direction outside the optimal pickup pattern of a directional microphone). Omni-directional mics pick up sounds relatively evenly from any direction. Some large-diaphragm condenser mics have variable pattern selec- tion, and some pencil condensers have swappable capsules that provide either cardioid or omni performance. W hile microphones operating in omni mode have slightly better frequency response and smoother overall characteristics, they have the disadvantage of pick- ing up a lot of room ambience and limited control over the volume of sounds coming to the mic from diff erent positions. When neither of these things are a problem—such as with orchestral recording, where the idea is to capture the sound of the ensemble and the room acoustics are considered an integral part of the sound—selecting an omni pattern may be a good choice. Orienting omni mics closer or farther from the sound source can also give the recordist a fair amount of control over room acoustics. In most recording instances, however, cardioid (directional) mics are preferred for their ability to capture the maximum direct sound and to minimize room sound and leakage of unwanted, off -axis sounds. Many microphones off er variations on the standard cardioid pattern, providing even tighter directionality, such as with hypercardioid or supercardioid patterns. Th ere are other mic patterns, such as the fi gure-8 or bi-direc- tional pattern, which provide two opposing pickup D IAGRAM 2.3 patterns, but cardioid and omni-directional pat- C ardioid and omni- directional pickup patterns terns are by far the most frequently utilized., THE ART OF DIGITAL AUDIO RECORDING

M icrophone selection

Remember: Th ere is no “right” mic for the job, as microphone selection is highly subjective. Generally speaking, for the greatest detail and fi delity, you would use a condenser. Typically, large-diaphragm condensers are used for vocals, but the sound of the voice, the desired sound, and the available microphones might 22 dictate the use of any of the other types of mics for recording vocals. Where o ff -axis (at an angle to the plane of the element being recorded; o n-axis means the plane of the microphone diaphragm is parallel to the recorded element) response is a problem, such as with multi-mic setups for ensembles, then pencil condensers might be the best choice. Dynamic mics are oft en a good choice for loud sounds with a lot of transients, such as drums and guitar amps, and sometimes for horns. A s with speaker selection, familiarity becomes the recordist’s greatest asset in choosing and using microphones. Not just familiarity with individual micro- phones but also developing a familiarity with the quality of sound that diff erent microphone types capture contribute to the recordist’s ability to make aesthetic decisions about mics and their eff ects. (See section 3.3 for more specifi c infor- mation on choosing microphones for individual instruments.)

Microphone placement

Aft er choosing the microphone you are going to use, you have to decide where to place it. Th e most basic part of that decision is how close to the sound source to place the mic. Th e proximity to the sound source aff ects both the detail that the mic is able to capture and the amount of room ambience relative to the direct sound. Studio practices have gravitated toward closer and closer miking techniques in order to capture the most detail from an instrument and to mini- mize the eff ects of room ambience—especially now that there are so many al- ternatives for adding ambience eff ects later via reverb and delay plug-ins. While close miking is the norm for individual instruments and voices, and it provides excellent results in most cases, it is certainly not the only approach. Maximum detail is not always desirable. Th e classic example is in record- ing stringed instruments. In most cases, you don’t want too much detail com- ing from a violin, where close miking may emphasize the scraping bow on the strings. (Th is is explored more thoroughly in section 3.3.) Similarly, minimizing room ambience is not always desirable. While it gives you the most options for controlling ambience later, sometimes room ambience plays an integral role in the sound and is best captured in the initial recording. Because it is impossible to truly eliminate all room ambience, some decision about balancing direct sound and room ambience is inherent in the microphone placement. When a mic is placed close to the sound source, a diff erence of 1 inch can have an audible eff ect on the sound captured. Experience and sensitive listening follow, Th e Essentials attention to microphone placement in order to capture the desired results. (Sec- tion 3.3 has more specifi c information on microphone placement for individual instruments, as well as diagrams and photographs.) A s will be emphasized in the discussion on session fl ow in section 8.1, it is important to keep your priorities straight when it comes to mic placement. Yes, small movements in microphone location will aff ect the sound captured, but optimal session fl ow oft en dictates against taking the time to do a lot of tweak- 23 ing of mic placement. A musician’s state of mind is more critical than small improvements in sound quality. Th is is why experience is so valuable—it allows you to make good choices quickly, thereby maintaining the creative fl ow of the session. Sometimes musicians thrive on taking the time for a lot of experimen- tation with mic placement (and sometimes the budget allows it, as well), but it is up to the recordist to help determine the proper balance between tweaking and keeping the session moving.

Phase and polarity

Phase and polarity are two key elements of concern whenever there are two sources for the same sound. Th ese are central considerations in the stereo mik- ing techniques covered in the sections immediately following this one. Phase issues are also key in the next chapter, which discusses strategies for various in- strument recordings, many of which use more than one source and thereby cre- ate issues concerning phase relationships. Before I cover stereo mic techniques, though, you need to be clear on how phase and polarity work. A phase relationship in recordings generally refers to the potential time diff erence between when a single sound source is received by two diff erent microphones (or other signal path). Variations in mic placement or other fac- tors may introduce diff ering amounts of delay before the signals are recorded. If the peaks and troughs of the waveforms are received at the same time, they are said to be “in phase” and the sound is reinforced by the two sources. If the sound is received at two diff erent times, depending on the relationship of the waves’ peaks and troughs, the result may produce phase problems (phase cancellation ). If the waveforms are somewhat off set, then certain fre- quencies will be canceled and others reinforced. If the waveforms are off - set completely, then there is the pos- sibility of complete cancellation. Th e reality is that rarely are two D IAGRAM 2.4 sound sources perfectly in or out of P hase phase, so the degree of phase coher-, THE ART OF DIGITAL AUDIO RECORDING ency is the primary concern. In fact, it is the slightly out-of-phase quality that gives stereo recordings their character. If the two signals are perfectly in phase, they would be identical and therefore would 24 be a mono signal. Sometimes phase problems can be detected by careful listening, but there is also a simple test to see if the two signals are generally more or less DIAGRAM 2.5 in phase. You pan the two signals Polarity hard left and right, and then switch your monitoring to mono. While monitoring in mono, you reverse the phase or polarity on one of the channels. Whichever setting is louder—the combined signal with one channel’s polarity switched or unswitched—is the one in which the signals are more in phase. If more frequencies are reinforcing each other, the sound will be louder. P olarity is not the same as phase, though the eff ect is related. Phase is the complex relationships of time between identical sources at their destination; polarity refers to the simple positive and negative voltage values of a signal. Phase diff erences will vary at diff erence frequencies when the time diff erence is constant—smaller amounts of phase for low frequencies and larger amounts of phase for high frequencies. Two signals with reversed polarity—caused when the positive and negative voltages are reversed—exhibit the same kind of can- cellation eff ect of signals completely (180 degrees) out of phase. Switching the polarity is the same as reversing the phase .

S tereo miking techniques

S tereo miking refers to the practice of using two microphones to create a stereo image. To get the maximum stereo eff ect, the two tracks that are recorded are panned hard left and hard right (all the way to the left and all the way to the right), but other approaches to panning stereo tracks may also be used. (See the mid/side stereo technique below for an exception to the hard left /hard right rule; and section 6.1 on mixing for more information about panning strategies.) Stereo miking can be used to capture ensembles when the sound is coming from a variety of sources, or it can be used to record a single sound source. With single sound sources, the stereo spread is created by variations in room ambience based on the orientation of the mic to the sound source. Variations in stereo miking techniques generally seek to address two primary concerns: fi rst, the breadth or width of the stereo image versus the desire for a stable, coherent center image; and second, the problems created by out-of-phase information, Th e Essentials caused when two microphones pick up the same sounds at diff erent locations. Th ere are four common stereo miking techniques covered below, with informa- tion about how they deal with these and other concerns. Th e coincident pair or X/Y confi guration Th e X/Y, or c oincident pair, technique is one of the most common and most reliable stereo miking techniques. It does a very good job of controlling prob- 25 lems in maintaining a coherent center image and with phase cancellation. Two cardioid microphones are set up with their diaphragms at a 90 degree angle and as close together as possible. Other angles may be used, broadening or narrow- ing the stereo fi eld, but common practice maintains the 90 degree model. Pencil condensers are frequently used for stereo recordings using the X/Y confi gura- tion because of their superior off -axis fi delity. Matched pairs of the same make and model of microphone are favored, but any pair of mics can be used. Because the two microphone cap- sules are place so close together, they receive the sound at almost identical times, thus limiting out- of-phase information. Because of their close prox- D IAGRAM 2.6 imity, they are also receiving enough of the same Coincident pair or X/Y information to provide for a coherent center confi guration image. For the same reason—their proximity— there is a limited degree of stereo image between the two channels, but because the mics are aimed at diff erent parts of the room, there is enough variation in what they pick up to make for a pleasing stereo spread. A broader stereo image will be captured as the coincident pair is moved closer to the sound source. As the mics move farther from the sound source, the diff erences in sound from one to the other will diminish. For a dramatic stereo eff ect, with a broad sense of the stereo fi eld, other stereo miking techniques yield superior results (and pose more serious potential problems, as well). Th ere are also single microphones with stereo microphone capabilities. Th ese mics have two diaphragms and two outputs—they are essentially two mics built into one body and are set to an X/Y confi guration (typically at a 90 degree angle, but not always). Some of these mics have the ability to rotate one of the diaphragms from a 90 degree angle into other variations in angle. Stereo mics are convenient, and the two diaphragms are always well matched, but they have the disadvantage of being limited in their approach to stereo mic confi guration. O RTF stereo confi guration Th e ORTF stereo confi guration represents a variation on the coincident pair and is sometimes called n ear-coincident pair. It was developed by the French national public radio and television broadcaster offi ce (acronym ORTF). Th is technique calls for two cardioid mics placed 17 centimeters apart (about 6.5 inches) and, THE ART OF DIGITAL AUDIO RECORDING D IAGRAM 2.7 26 O RTF stereo confi guration at 110 degree angle. Th e mics should be as similar as possible, preferably the same make and model. Some manufacturers sell frequency-matched pairs that are particularly nice for stereo miking. Th e ORTF confi guration reminds us that the distance and angle between two mics used in a coincident-pair confi gura- tion can be adjusted for variation in results. C onsiderable research and testing went into the ORTF standard, and it yields reliably good results, but other variations can be used with great success and there are other standards, as well. Th e advantage of the ORTF technique over the traditional X/Y confi guration is that it has a broader stereo fi eld while maintaining good mono compatibility (minimal phase problems) and a rela- tively stable center image. While strict phase compatibility and center image stability are better with the traditional X/Y, I fi nd that in many cases the more pronounced stereo image is worth the small compromises, and I tend to use the ORTF technique frequently. I’ve also found that the distance between my thumb and my little fi nger, with my hands spread wide, is just about the right distance for an ORTF setup. If you can fi nd some easy way such as this to refer- ence this distance, it will speed your setup. Special mic clips that will hold two pencil condensers in either the X/Y or ORTF confi guration (as well as other variations) are available and are very handy for this application. (More specifi c applications of the ORTF confi guration can be found in section 3.3. ) S paced pair (omni-directional or cardioid) Th e spaced-pair mic placement is especially good for recording ensembles, from bands to orchestras, because the two mics pick up sound more evenly over a larger area than the coincident pairs. Two matched microphones are generally placed between 2 and 12 feet apart, depending on the size of the ensemble. Th e mic pickup patterns may be either cardioid or omni, with omni being the preferred pattern (better frequency response) as long as the additional room ambience picked up in the omni position is not a problem. Many engineers employ the 3-to-1 rule, which holds that if the microphones are three times as far from each other as they are from the sound source, there will be minimal phase problems. In practice, this isn’t always true, as room acoustics and the nature of the sound source also aff ect the phase relationship. Trial and error, by moving microphones and listening, is the best way to fi nd the optimal placement for a spaced pair., Th e Essentials S paced pairs are technically the most problematic of all the commonly used stereo techniques because of the potential phase problems and the possi- bility of an unstable or “blurred” center image, caused by microphones that are far apart from each other (sometimes referred to as a “hole” in the center of the stereo image). Th is is why one of the variations on coincident pairs, such as the Decca Tree (see below), may be preferred. However, when the right positioning is found via trial and error, spaced pairs can produce very good and dramatic 27 results. Checking the summed (mono) response of the two mics in a spaced pair is one good way to determine how much of a problem the phase relation- ship may be. Th e more the sound is diminished in mono, the greater the phase problems. Decca Tree Th e Decca Tree is a variation on the spaced-pair confi guration. Th e recording engineers at English Decca Records developed it in the 1950s, primarily for or- chestral recording. Th e Decca Tree adds a third mic to the spaced pair in order to provide greater center-image stability. In its basic confi guration, the Decca Tree utilizes three omni, large-diaphragm condenser mics with the left and right mic approximately 2 meters (6 feet) apart and the third mic centered about 1.5 meters (4.5 feet) in front of the other two. In practice, many diff erent microphones, in- cluding pencil condensers, and either cardioid, supercardioid, or omni patterns maybe selected. Also, the distance between the mics may be adjusted depend- ing on the size of the ensemble, the room acoustics, and the desired eff ect. Even the standard panning of hard left , hard right, and center may be adjusted. Th e mics are generally aimed in toward the center; even omni mics exhibit a certain amount of directional bias, especially in the higher frequencies. O ther variations on the Decca Tree include the addition of two more mics, usually farther back from the ensemble and spread more widely, to gain greater stereo width and room ambience. Th e center mic may be replaced by a pair of mics in the X/Y confi guration or other variations on a coincident pair. In whatever confi guration that is used, it is the balance between the center and the fl anking microphones that will be adjusted to create more or less stereo spread—more fl anking mics in the balance for greater stereo spread, more cen- ter mic for greater center stability. Again, monophonic summing (listening to all the mics in mono) will reveal problems in phase coherence and may cause you to increase or decrease the relative level between the mics. Orchestral re- cordings for use in fi lm soundtracks oft en employ the Decca Tree because it can produce a stable stereo image that holds up well when processed for surround- sound applications. Mid/Side (M/S) Th e mid/side technique uses two mics with two diff erent microphone patterns, one cardioid and one fi gure-8 (sometimes called bipolar or bi-directional). Th e, THE ART OF DIGITAL AUDIO RECORDING cardioid mic is for center (or mono) information and is generally aimed at the sound source. Th e fi gure-8 mic is placed in close proximity (usually above or below) the cardioid mic and aimed at a 90 degree angle to the cardioid so that the two areas that it picks up are each off set 90 degrees from the center mic. Th e fi gure-8 microphone encodes stereo information by picking up from the two opposing sides of the microphone’s capsule. Th e single channel that is recorded 28 by the fi gure-8 mic (side channel) is decoded by duplicating that channel and reversing the polarity (also called “inverting the phase”) on the duplicated chan- nel and then panning the original and polarity-reversed channel hard left and hard right, respectively. Some DAWs, such as Nuendo, have a Stereo Tools VST plug-in that will automatically confi gure the side channel (fi gure-8 recording) as described here. If you group the left and right (side) channels together, you can raise or lower their volume relative to the mid channel (panned center), and in doing so you will increase or decrease the sense of stereo spread. Th e biggest advantages to the M/S technique are in mono compatibility and in the way you can control the stereo versus mono relationship. Because all of the stereo information is provided by two identical but reversed-polarity tracks, they completely cancel each other out when played back in mono (such as playback on old mono AM radio receivers or television sets). Th is leaves only the original mid or mono channel, without any of the phase anomalies of other two-channel stereo miking techniques. It also eliminates any room ambience that has been added by the side channels, which may or may not be desirable in the mono playback setting. Because all of the stereo information comes from one microphone, and all of the mono information comes from another, you can balance the two, keeping a clear diff erentiation between stable mono and highly phased stereo. Mics and DAWs M ic selection and positioning are critical elements that aff ect the quality of your recording. Quality, in this case, means both the fi delity and the aesthetics, or “sound,” of what has been captured by each microphone. In many instances, the signal path from the mic into the computer is the only time your audio will be processed in the analog domain. You may wish to access analog gear WHAT NOT TO DO Do not get hung up on having to use your stereo recording in maximum stereo confi guration—with the two channels panned hard left and hard right. “Collapsing” the stereo image by bringing the panning of either or both channels in from hard left or hard right is often desirable in mix situations., Th e Essentials such as compressors or EQs as part of this chain to avoid sending your signal back from digital to analog for this kind of processing. Th ere are other, less frequently used kinds of mics (Soundfi eld mic, binaural mics, etc.) and miking confi gurations (Blumlein pair, baffl ed stereo confi gurations, etc.) to explore, but they fall beyond the scope of this book. In any event, you will want to pay close attention to the role that the microphones are playing in your overall recording strategy. 29 2.4 Mixing Boards and Control Surfaces Traditional routing for analog recordings goes from the microphone to a mix- ing board (mixer) to a tape recorder. Microphones may be connected to a DAW in a wide variety of ways, and the mixer/recorder paradigm from the analog world has been expanded. DAWs include a mixer-style interface as part of the soft ware; some DAW interfaces include hardware mixers and a whole new world of control surfaces that may replace a traditional mixer in a DAW setup.

Mixing boards and control surfaces: What are they?

To begin, it is necessary to defi ne what is meant by mixing boards (usually re- ferred to as mixers, but also called consoles, desks, etc.) and control surfaces. A traditional mixer includes all of the elements necessary for routing audio to and from the tape recorder and the speaker/amplifi cation system, as well as the capability of controlling the audio for most other routing or processing that may be desired. Th is means that most mixers have microphone preamps, some amount of signal-processing capabilities (generally at least some EQ), and rout- ing capabilities for incorporating all varieties of external gear, such as other signal processors, cue/headphone systems, and other recorder and/or playback devices. A c ontrol surface is a subset of a mixer that generally provides only for the control and routing of the audio, without the mic preamps or signal- processing capabilities. Th ere are numerous hybrid products that incorporate some, but not all, of the capabilities of a traditional mixer. Soft ware mixers (such as the “mixer” page in your DAW) are really just virtual control surfaces, although they become more mixerlike by using plug-ins to give the user signal- processing capabilities. Mic preamps are hardware by nature—the mic must be able to physically plug into them. Here, I am concerned primarily with soft ware mixers (the mixer in your DAW), which might be better described as a virtual control surface. However, your DAW’s mixer is modeled aft er its hardware predecessors, so much of what I cover here translates to the hardware world as well, and I include a discussion of mic preamps also. I follow a typical order of controls from top to bottom on a typical channel strip, but this order will vary with diff erent soft ware. Th e general function of each of these controls is found in almost every mixer. At, THE ART OF DIGITAL AUDIO RECORDING PHOTO 2.7 A Solid State Logic (SSL) G+ series analog mixer PHOTO 2.8 A Digidesign C24 digital control surface the end of this chapter, I have a more thorough examination of soft ware versus hardware mixers.

Th e mixer channel strip

E ach channel strip duplicates a set of controls for the individual channels on a mixer. Th e number of channel strips defi nes the capacity of a hardware mixer (e.g., a 16-channel mixer or a 24-channel mixer), but the soft ware world has pretty much ended that distinction. With most DAW programs, channels can be added as needed, oft en up to a very large capacity. Even some systems that, Th e Essentials restrict the number of audio channels still provide a large number of auxiliary channels, as well as virtual tracks (covered in section 4.2) that multi- ply the mixer’s capacity enormously. By examin- ing each of the principal functions of the channel strip, I survey all of the primary operations that a mixer is used for. 31 Shown on this page are a couple of screen- shots of soft ware channel strips from two dif- ferent DAWs. Note the labels for the functions, including Inserts, Sends, I/O for inputs and out- puts, Panning, Solo, and Mute, plus the main fader including volume readout, the scribble strip for labeling the channel, and other functions de- pending on the DAW. Note that many DAWs do not have labels for every function on the channel strip, requiring you to learn your way around the DAW, using the manual and/or trial and error. Types of mixer channels B ecause of the increasing number of capabilities within a DAW, there has been an increase in the number of channel types. It is not possible to thor- oughly discuss them all in this context, but you should be aware that great fl exibility is derived S CREENSHOT 2.1 from using the proper channel for the proper A channel strip: Pro Tools function. Here, I cover audio channels in depth; and in the section on sends and returns (section S CREENSHOT 2.2 5.2) and building a mix (section 6.1), I look at uses A channel strip: Digital for auxiliary input channels (aux channels). Mas- Performer ter fader channels are also covered under the topic of building a mix (section 6.1). Besides these channels, your DAW may include the ability to create channels specifi cally for MIDI use and for instrument use (usually “soft synths,” or soft ware-based synthesizer and sampler programs that operate within the DAW environment). See your DAW user guide for more spe- cifi c information on these and other specialized channel strip capabilities.

I /O—input and output

I n the previous section on signal path, I covered some of the general principles of input and output (I/O). Somewhere in your channel strip you must have the option for choosing the primary input and output for that channel (as seen on the previous screenshots, this may be located at diff erent places in the chan-, THE ART OF DIGITAL AUDIO RECORDING nel strip—top or middle—on diff erent DAWs). Th e primary input sets the path that audio takes to get into each channel. Th e signal is typically coming from a microphone, but it could be from a synthesizer, from another already recorded audio track, or from any other audio source. Th e primary output sets the audio destination when it leaves the channel. U sually this would be the stereo buss that feeds the playback (speakers), but it could be going to an outboard processing 32 box, another track, or any other audio destination. Interface or buss routingInthe digital world, there is an important distinction made with regard to in- puts and outputs that did not exist in the world of hardware mixers. Within the soft ware mixer, the choices for I/O routing may be either through inter- faces or through busses. Th is distinguishes audio routing that takes the audio out of the computer (external) from routing that keeps the audio within the computer (internal). E xternal routing— routing out the audio interface through which all audio must travel to get in or out of the computer—is used when the audio needs to access external gear, such as speakers, amplifi ers for headphone mixers, or any analog processing. Internal routing uses busses to move audio around within the computer soft ware—such as to other tracks or to computer- based processing tools (plug-ins). DIAGRAM 2.8 Internal and external routing Mono or stereo A udio channel strips may be confi gured as mono or stereo in both their input and their output status. Mono input and stereo output is the most common confi guration, but stereo inputs are also common when stereo recordings are, Th e Essentials being made or when stereo samples are accessed. Mono outputs are available and are valuable if you are routing audio from the DAW to a mixer or con- trol surface channel where they are then given stereo-output capabilities. Most hardware channel strips are confi gured mono in and stereo out, so the output of the DAW to the input of the hardware channel strip is a mono signal path. (See the section below on panning for more on mono and stereo outputs.) Th ere is increasing need for and use of expanded I/O options to deal with surround 33 sound (5.1, 7.1, etc.), but that is beyond the scope of this book.

Mic preamps

Microphone preamps are necessary to amplify the low-level output from a microphone. Th ey provide a variable level to supply the proper output to be eff ectively recorded. Diff erent microphones have very diff erent levels of out- put—and, of course, sound sources diff er enormously in volume as well—so the ability to control the output from a mic with a preamp is essential to the recording process. With the prevalence of DAW systems, where mic preamps are not necessarily a part of the hardware interface, more attention has been paid to outboard preamps (any mixer-related hardware, such as mic preamps or processing units, that is not built into a mixing board is referred to as o utboard equipment ). Mic preamps (whether onboard or outboard) are also able to sup- ply the special phantom power needed for condenser microphones, and they oft en include a phase-reverse switch, as well. M ic preamps, as with most electronic audio gear, come in two basic de- signs: vacuum tube and solid state. Th ey also come in a staggering array of qual- ity and price ranges. As with microphones, the selection of mic preamps should be based on a combination of access, intended use, and experience. In selecting all audio gear for purchase, it is a good idea to keep in mind that every link in the chain is critical. It prob- ably does not make sense to buy a $1,000 mic pre to am- plify a $100 mic (though it does no harm), and it cer- tainly isn’t advisable to buy a  $25 mic pre for use with your $1,000 mic. Because of the proliferation of inter- faces that provide only line- level input to the DAW, it has become more common for studios to accumulateaPHOTO 2.9 variety of external mic pre’s A Solid State Logic (SSL) mic preamp in order to have a range of, THE ART OF DIGITAL AUDIO RECORDING options for diff erent mics and diff erent situations. Th is proliferation of stand- alone mic preamps also eliminates the need for the typical mixer with built-in mic pre’s, and this is part of the reason for the rise in hardware-control surfaces as alternatives to mixers. In a typical hardware mixer each channel contains an onboard mic pre- amp. It is generally located at or near the top of the channel strip. Th e onboard 34 mic pre ranges from a barebones model that has only a gain control to a more elaborate preamp with individual controls over phantom power, pad, and phase reversal and a separate level control for line-level signals, such as in the photo on the previous page.


Inserts are such an important part of a soft ware mixer’s applications that I devote a separate section to their use (see section 5.1). Here, I simply note that the in- sert portion of the soft ware mixer is the point at which all manner of processing functions, as well as soft ware instruments, are integrated into the mixer envi- ronment. Th is is one of the areas in which soft ware mixers diff er considerably from hardware mixers. Th e use of inserts to dramatically increase the control over and creation of audio has far exceeded the comparably minimal use that inserts found in the hardware world. (Th ere are more details on this at section 5.1.)

Auxiliary sends

Auxiliary sends (or “aux sends,” or most commonly, just “sends”) are another essential part of mixer functionality, and their various uses are outlined in two independent sections in this book (sections 3.2 and 5.2). Here, I cover only the basic controls found on a typical aux send. An aux send functions similarly to the main fader on any mixer channel. Th e primary routing for audio on any given channel is through the primary channel output, and the main channel fader controls the level of that output. Th e channel aux sends provide further routing options for the same audio—the audio on that particular channel. Th is is why they are called aux sends—they are auxiliary (or “in addition”) to the main send, which is output controlled by the channel fader (usually located at the bottom of the channel strip). When an aux send is created on a soft ware mixer, typically a pop-up consisting of a new fader appears along with a variety of other controls. Additional aux send controls include the ability to select the output for the send, panning control, solo and mute capabilities, and the pre-fader or post-fader status for that send. Pre-fader and post-fader aux sends Th e terms p re-fader and post-fader describe a critical element in the routing status of an aux send (and the settings are oft en shortened to simply “pre” or “post” when describing the send’s status in this regard). Because the aux send, Th e Essentials is in addition to the primary channel output, you must set its routing status relative to the primary output. Th e channel’s main fader con- trols the primary output. Any aux sends on that same channel access the audio on that channel, either before (pre-fader) or aft er (post-fader) the audio is routed through the main channel 35 fader. If it is selected to be pre-fader, then its level control of the channel audio is unaff ected by the position or movement of the main fader and the level is controlled only by movement of the sends fader. If it is selected to be post-fader, its send level is aff ected by both the position and the movement in the main fader, as well as by the sends fader. I n practice, the decision to set a send to pre or post depends on the intended use for the audio being sent. Th e two primary uses for aux sends are headphone mixes and access to eff ects via sends and returns; and these are prime examples of the need for the two dif- ferent routing options (pre and post). Because headphone mixes need to be completely inde- pendent of the control-room mix, the sends used will typically be set in the pre-fader posi- tion. Because eff ects added in the sends and returns routing model need to maintain a con- sistent relationship to the level of the primary output, the sends used will typically be set to SCREENSHOT 2.3 post-fader. (Details are covered in the sections S end control on headphone mixes and sends & returns, sec- tions 3.2 and 5.2, respectively.) Aux send outputs S ends have output routing that is separate from the channel’s primary output rout- ing. As with the primary outputs, however, these outputs may be either through the in- terface or via busses (explained in the sec- tion on I/Os, above). When sends are used for things such as headphone mixes, it is DIAGRAM 2.9 necessary to use the interface outputs in Pre-fader and post-fader auxiliary sends order to get to the headphone system. For, THE ART OF DIGITAL AUDIO RECORDING internal processing, such as that done when using the send and return model, busses are used to route the signal. (Again, details are covered in the sections on headphone mixes and sends and returns.)


36 Th e channel pan function controls the placement of the audio in the stereo fi eld. Panning requires a stereo output, allowing you to move the sound from the left speaker through the stereo fi eld to the right speaker. If you create a channel with a mono output, you will notice that the panning function has been eliminated—there can be no panning with a single channel of output. Th is seemingly simple distinction between mono and stereo—and the ability to pan audio in stereo—is oft en misunderstood. A mono sound source (a single sound) can be panned (placed) anywhere in the stereo fi eld as long it has access to a stereo output (typically, outputs 1 and 2—with output 1 feeding the left speaker and output 2 the right speaker). Both elements of a stereo sound source can be panned across the stereo fi eld independently, though placing one hard left and the other hard right is commonplace for stereo audio. (More on this in building a mix, section 6.1.) When something is playing “in mono,” this means that there is no diff erence between what is feeding each of the two speakers. A mono sys- tem has only one output to the speaker(s), whereas a stereo system must have two outputs and two speakers. Here are the input and output options for pan- ning capabilities: Input Output Panning Mono Mono No panning possible Stereo Mono No panning possible Mono Stereo Sound can be panned anywhere in the stereo fi eld S tereo Stereo Sound from each output channel can be panned anywhere in the stereo fi eld W hat, then, is meant when the sound coming from a channel with ste- reo output is playing in mono? This means that the sound is center-panned (panned evenly to both the left and the right channel). This is sometimes where the confusion comes in. When a sound is center-panned, it is ef- fectively “playing in mono”—it is not using the capabilities of panning be- cause the same level of output is feeding each speaker equally. If all channels are center-panned, the entire piece is effectively “playing in mono” (even though channels with stereo outputs have stereo capabilities). As soon as the sound is moved by the use of panning, even slightly, to create an im- balance between the right and left speakers, then the sound is “playing in stereo.”, Th e Essentials

Output fader

Th e main fader control (generally at or close to the bottom of the channel strip) controls the output level. It is important to remember that it is at the end of the channel’s signal path, controlling only the level of the signal as it leaves the channel to its destination, as set by the main output. Th e destination is fre- quently the stereo buss, but it may be any interface or buss output. Th is means 37 that the position of this fader has no eff ect on the input to the channel, and therefore, it has no eff ect on the level of the recording (a common novice mis- take is to try to turn down the recording level by lowering the channel’s main output fader). Th e fader sets the monitor (listening) level and the fi nal output volume when mixing.


I n every DAW, there is the capability for grouping channels together to facili- tate a variety of functions. In the most basic group confi guration, the output faders of each channel are grouped so that moving any one fader moves all fader levels in the group by the same amount. Th is allows you to easily raise or lower the level of many channels used for the same instrument (like multiple tracks oft en used to record a piano or a drum set) or many channels of related elements (such as backing vocal tracks). Other group controls include solo and mute functions, panning, input or output assignments, automation controls, or arming the channels for recording. Whether or not the groups share all of these functions is usually determined by the user and will depend on the nature of the elements in the group. Most of these controls provide added convenience with the exception of panning position, which will oft en be best left individu- ally variable, as diff erent panning positions are usually an important part of the group settings. In many DAWs, the groups made by the user are given a particular des- ignation (a number or a letter) that is indicated on the channel strip. Channels may be color-coded by groups as well. It is necessary to have the ability to tem- porarily suspend the group functions, so that you can adjust individual tracks independently and then return to group status (perhaps the hi-hat is too loud relative to the rest of the drums, or one background singer’s voice is getting lost in the mix and needs to come up in volume). C hannels may be part of more than one group. Larger groups can be helpful in complex projects where groups involving whole sections (such as strings, or percussion) may be used at times, and then suspended while smaller groups (violins, or hand percussion, for example) are kept active for more fi ne-tuning. Th e ability to group all channels in a project allows for ed- iting entire sections of the arrangement, such as when eliminating or rear- ranging whole sections of a piece (as described in section 4.4 under “Global, THE ART OF DIGITAL AUDIO RECORDING Edits”). In almost every recording project, there are instances when using groups makes the workfl ow simpler.

Track name/Track notes

A t the bottom of the fader, there is typically a place that allows you to name the 38 channel. Th is is roughly the equivalent of the track sheet used to keep track of what was recorded on individual tracks of analog tape, or the “scribble strip” at the bottom of a mixing console where tape is generally placed to write track names. Th e track name has the additional function of supplying something other than a default name for the audio fi les as they are recorded. Th is means that if you name your track Gtr (for guitar) or Vox (for voice), then each audio element recorded on that channel will be tagged with the label Gtr or Vox, adding a numbering scheme each time a new recording is made. For example, the fi rst audio recording on the track might be labeled by the DAW as G tr.01, the second as G tr.02, and so on. Th is can be very useful in both fi nding audio fi les at a later time and being able to identify the order in which they were recorded. If the default track name is not changed when creating new tracks, the audio will be labeled with the default name (such as A udio1.01, A udio1.02, etc.). Th is leads to a huge number of audio elements with very similar names and no means of identifying them. Naming tracks before recording is a benefi cial practice in the DAW recording process. SCREENSHOT 2.4 Track naming and scribble strip B elow the track name there is oft en an area for making notes or comments about the track. Again, this is information that was generally kept on the track sheet for recordings made on analog tape recorders. Th e two most common bits of recordkeeping done here are the name of the musician who was recorded and the name of the microphone (if any) that was used. Other information, both technical and creative, can be entered here. You may want more complete input path information, such as the type of mic preamp or compressor that was used, or you may want to make mix notes such as “fi lter the low end rumble.” You may also want to note particularly strong or weak elements (“great solo”). Th e ability to name tracks and make notes and comments becomes even more useful when using virtual tracks (described in section 4.2).

Other kinds of channel strips

Besides the typical audio channel, mixers (both hardware and soft ware) have other kinds of input and output capabilities, such as auxiliary inputs (aux in- puts) or master fader outputs. DAWs now come equipped with a large array of specialty channels. Besides aux inputs and a master fader, a DAW mixer may, Th e Essentials WHAT NOT TO DO Don’t postpone keeping tracks labeled and organized as a session progresses. Even in the heat of a rushed session, it is worth the few seconds it takes to label a new track with the name of what is being recorded, making a quick but essential note in the scribble strip (e.g., “mute this track during the guitar solo”), and creating a group if you’re 39 recording multiple related elements (e.g., three backing vocalists on separate tracks). Labeling saves time in the long run and is always worth the little bit of time it takes. include MIDI channels for handling MIDI data, as well as instrument channels for soft ware instruments/synthesizers (soft synths). Auxiliary input channels Aux inputs provide additional routing capabilities that are used primarily for internal routing and processing duties. An aux input cannot record or playback audio. Instead, the aux track passes audio through a channel and this can be used for processing or monitoring. Whenever you wish to use signal process- ing (EQ, compression, etc.) as part of the recording or on groups of already recorded tracks, an aux channel can provide the appropriate signal path. An aux channel can also be used to monitor a talkback mic that isn’t being recorded. Unlike audio tracks, aux tracks receive audio without having to be in record mode (or record ready), as long as audio is routed to their input path (either interface or buss). Unlike a DAW mixer, every channel of a hardware mixer functions like an aux channel, as opposed to an audio channel. Th is is because the channels of a hardware mixer don’t actually contain audio recordings (the audio is handled by the separate recorder)—they simply pass the audio signal through for pro- cessing and mixing. What are called “aux channels” in a hardware mixer (oft en included in the center section) are really just more input channels with limited routing and processing capabilities. Th e integration of the actual recording, as well as the added fl exibility of soft ware over hardware, gives the DAW chan- nel paths much broader functions than found in any hardware mixing console. (For further information about aux track functions, see section 5.1 on insert/ plug-in uses and section 5.2 on send and return routing for signal processing.) Master fader channel M ost hardware mixers incorporate a stereo master fader that gives you single- fader control over the sum of all the individual channels. DAW mixers do the same thing, though as with all channels in the DAW mixer, it is up to you to create a master fader or to work from a template that already has one created., THE ART OF DIGITAL AUDIO RECORDING Th e master fader is used to control global level movements such as fade-outs at the end of the song. It can also be used to adjust the overall level of a mix before it is rendered or bounced to a fi le for use outside of the DAW (such as burned to a CD, podcast, or e-mailed as an mp3) Th e master fader also allows you to see what the overall (summed) level is so that individual tracks might be adjusted up or down to put level operations into a comfortable range. Th e master fader itself 40 can also be used to adjust overall output if the sum of all the tracks is not at a comfortable operating level. It is a good idea to create a master fader in your DAW at the beginning of your project, to help monitor your overall gain structure. MIDI channels MIDI is an acronym that stands for Musical Instrument Digital Interface. MIDI channels allow for the recording and playback of MIDI data. MIDI is not audio; it is digital data that is used to control synthesizers and other computer-based music gear. MIDI information is stored and controlled diff erently from audio information, so MIDI channels are an essential part of every DAW (many DAWs began as MIDI recorders/sequencers). MIDI production techniques fall outside the scope of this book, which specifi cally addresses audio recording. Th ere are plenty of books about MIDI and I encourage you to study and explore the MIDI capabilities of your particular DAW. Instrument channels O ne of the most explosive areas of development in the world of DAWs has been the integration of soft ware synthesizers (soft -synths). Th ese soft ware in- struments run the gamut from traditional synthesizer-type sound generation to elaborate sample-based instrument playback. Th e instrument programs are oft en capable of running either as stand-alone soft ware or integrated into most DAWs. Many DAWs now include dedicated instrument channels that can be created within the mixer environment to best integrate the functions of the soft -synth program. Implementation will vary depending on the DAW and the particular soft -synth.

H ardware versus software mixers and control surfaces

Th e focus of this book is on the DAW, but every DAW soft ware program re- quires some amount of hardware to get audio in and out of the computer. Th e diff erences between soft ware and hardware control of audio has led to some confusion over the need for a hardware mixer. At the same time, the typical soft ware version of a mixer that you fi nd in a DAW has now been recreated in the hardware world, and to diff erentiate it from the traditional hardware mixer, it has been renamed a “control surface.” So, soft ware “mixers,” or mix pages, or whatever they are labeled in your particular DAW, are really more akin to the new generation of hardware control surfaces than they are to the traditional, Th e Essentials hardware mixer. Th e main diff erences between a control surface and a mixer are in the ability of the mixer to process audio (primarily with EQ, but some mixers have other processing capabilities, as well) and the existence of mic preamps (which by their nature must be hardware). Th e one item on a typical hardware mixer channel strip that cannot be reproduced through soft ware in the DAW is the microphone preamp. Instead, many DAW manufacturers integrate mic preamps into their interface units, and 41 there is a proliferation of stand-alone mic preamps available in all price ranges. Whether through the use of integrated mic preamps or stand-alone units, if you are going to record using a microphone, you will need a hardware mic preamp to amplify the signal before it goes to the DAW. Audio processing (signal process- ing such as EQ or compression) can be handled with hardware or through the use of soft ware plug-ins. Hardware processors can be either digital or analog, but are built into the mixer rather than accessed as plug-ins. Plug-ins can be part of a soft ware mixer (the one built into the DAW) or accessed from a digital mixer or control surface. If this all seems confusing, it’s because it is—there is a lot of cross- over in functions between the hardware and the soft ware world. Here’s a break- down of the main features for the basic types of mixers and control surfaces: • Analog Mixer. Th is is the traditional-style mixing console. Analog mixers generally include mic preamps and some signal process- ing, though most oft en just EQ. • D igital Mixer. Th e digital mixing console includes analog mic preamps so that it can function as a true mixer. Digital mixers sometimes include built-in hardware signal processing, as well as the ability to access and control your DAW’s soft ware processors via plug-ins (some even include plug-in soft ware). • A nalog Control Surface. Th ere is no such thing. If it is a control surface, it is the digital control of a DAW, though it may have some analog elements. See Digital Control Surface, below.. • D igital Control Surface. Th e hardware-based digital control sur- face off ers physical control over a DAW (fader controls, panning knobs, etc.), but typically does not include mic preamps or any processing capabilities—outside of the ability to use and control plug-ins. • S oft ware Mixer or Control Surface. Th e built-in mixer in a DAW is really a control surface (though it is oft en labeled as the mixer or mix page). Th e DAW soft ware cannot include mic preamps and it handles all signal processing via plug-ins. By their nature, soft - ware elements are completely digital. D eciding between using a hardware mixer or control surface (digital or analog) and relying solely on the DAW’s built-in soft ware control surface (the virtual mixing board) has become a major dilemma for many recordists. Below,, THE ART OF DIGITAL AUDIO RECORDING I examine the pros and cons of each and I off er the rationale for my own work- ing methodology. Having a hardware mixer or a control surface does not re- strict you from using the DAW’s control surface, but it does add expense. Using analog mixers Analog mixers off er the most elements not found in a DAW. Besides the es- 42 sential mic preamps, these mixers may provide analog EQ, as well as other pro- cessing gear, such as the compressors and noise gates found in some high-end analog consoles. Th ey also off er the advantage of physical faders that give tactile control and are much easier to operate than trying to move soft ware faders with a mouse. Analog mixers also provide analog summing, which is to say that they combine all the individual track outputs into the stereo buss in the analog realm. Some argue that digital summing is one of the weak points in soft ware mixers. Th e downside of analog consoles are that they require conversion from digital to analog and back again in order to be used (assuming the source is a DAW and not an analog tape recorder), and there is some loss of detail in any conversion process—the extent being determined by the quality of the conver- sion. Th e mic pre’s on the console might not be of the same high quality available in stand-alone units because of the demands of providing mic preamps (and EQ and routing, etc.) on each channel, which is expensive in the analog world. Top- end consoles provide excellent mic pre’s and processing, but they also command relatively high prices. Although there is physical fader control, these faders don’t provide access to the highly fl exible and reliable automation of the DAW. Of course, they don’t prevent you from using the DAW automation, but then the ad- vantage of the physical fader is minimized. Finally, there is disagreement about the summing issue. DAW manufacturers have sponsored shoot-outs that would indicate that digital summing in the DAW is not audibly diff erent from analog summing (and comes without the added layer of conversion), but others claim a dramatic diff erence between digital and analog summing. So, is a hardware mixer an advantage? Personally, I believe it is an advan- tage on large recording sessions if a high-end console is available. Th is gives you access to a lot of good mic preamps and EQ. Stand-alone mic pre’s are fi ne— maybe even preferable, depending on make and model—but having enough for a large session is expensive and complicates routing and operation. Analog EQ on the best mixers sounds great and has a quality that is not exactly reproduc- ible in the digital world. In smaller sessions, stand-alone mic preamps are oft en the best choice. Not all elements want or need EQ when recording (or mixing, for that matter), and there are a variety of hardware “channel strip” options that combine a mic pre with EQ and compression, so only in large sessions with many elements that might benefi t from EQ does the analog mixing console provide a substantial advantage. Personally, I love having an SSL or a Neve for a large tracking (full band or large ensemble) recording date, but outside of that,, Th e Essentials I fi nd that the overlay of duplicated functions—not to mention the expense— outweighs the advantages of a hardware mixer in most instances. U sing digital hardware mixers and control surfaces D igital mixers (as opposed to digital control surfaces, covered below) imply the presence of analog elements, most especially mic preamps. Th ey may also have hardware or soft ware processing built in (EQ, compression, reverb, etc.). Oth- 43 erwise, hardware digital mixers and control surfaces do the same thing: they provide physical, tactile control over the digital mixing functions found in the DAW. Th e advantage is in the tactile control over faders and other mixer func- tions, such as panning or plug-in parameter control. Th is is generally easier than mouse control over the same functions. Th e mixer or control surface inter- face also off ers quick access to several elements at once because of the physical faders and knobs (including two-handed operation). Th e disadvantages to these kinds of digital hardware (besides expense) is that they do not provide the graphic-based automation found in the DAW, and for many of us, this is preferable to physical control over faders. Also, most hardware mixers and control surfaces of this type off er a limited number of channel controls and require paging through diff erent screens in order to access all the channels from a larger DAW session. Personally, I fi nd a few elements in hardware mixers and control surfaces convenient for some operations, but because I prefer graphic automation, it is diffi cult to justify the expense. For many, the physical requirement of having a large console in your workspace is also problematic. Using only the DAW control surface Using what is oft en labeled as the DAW’s mixer is essentially using a control surface. You can control all routing and automation functions, but signal pro- cessing comes in the form of plug-ins and mic pre’s must be accessed from the hardware world. Th e DAW provides excellent automation (especially in graphic mode) and controls all mixer functions. Certainly in terms of cost, the DAW is the most effi cient mixer, as it requires no additional mixer or control surface hardware. (For more on maximizing your use of the DAW’s mixing capabilities, see chapter 6.) M any home and project studios have no hardware mixers or control sur- faces, using only the DAW for all mixer-type functions and using either the mic pre’s built into the DAW interface, external stand-alone mic pre’s, or some combination of the two. Manufacturers have responded to this situation with a variety of mic preamp options at a wide range of price points. Units with two, four, or eight mic preamps in a single rack space are common, and some of them have built-in digital conversion to access the DAWs digital inputs. Chan- nel strips—typically one or two channel units that oft en incorporate mic pre’s, EQ, dynamics processing, direct box functionality, and even analog to digital, THE ART OF DIGITAL AUDIO RECORDING conversion—have also proliferated to meet the needs of smaller facilities or to add variety in signal-path options at larger studios. 2 .5 EQ: General InformationEQstands for “equalization” and it has become the default name for what was 44 traditionally termed “tone controls” in consumer audio hardware. Th e term equalization comes from the original intent to “fl atten” or equalize frequency responses. Now, EQ is used to alter and reshape (and, it is hoped, to enhance) sounds, with many diff erent goals in mind. EQ is capable of altering frequency characteristics from low to high. Frequencies are expressed in Hertz (symbol Hz, named aft er the German physicist Heinrich Hertz), which is the scale used to pinpoint any particular place on the frequency continuum. EQ allows the user to shape the tonality of the sound by either boosting or dipping various frequencies. EQ is the most powerful, and most frequently used, of all the signal processors. Most hardware and soft ware mixers include the capability of apply- ing EQ in one form or another.

E Q parameters

Th ere are three primary parameters in most EQ operations, although there are a multitude of specialty EQ functions that provide somewhat diff erent tone-shap- ing capabilities. Th e human ear, operating at maximum capacity, can typically hear sounds from about 20 Hertz (or 20 Hz on the low end) to 20,000 Hertz (expressed as 20 kilohertz or 20 kHz on the high end). In typical EQ operation, you can either boost (add) or dip (reduce) the level of certain frequencies in the sound to change its sonic characteristics or frequency shape. Th e degree of boost and dip is expressed in decibels (dB), which provide a volume scale rang- ing anywhere from .1 to 15 dB in standard operation. Two of the three primary EQ parameters are pretty obvious: (1) boost- ing or dipping by a variable degree (more or less boost or dip); and (2) the frequency (from low to high) that you are boosting or dipping at. Th e number of “bands” available refers to the number of diff erent frequencies an EQ can operate on at the same time. Th e typical “tone controls” in consumer electronics provide two-band EQ, meaning you can boost or dip frequencies in two diff er- ent ranges, and those are typically labeled treble (highs) and bass (lows). I f you can fi nd the specifi cations on a typical piece of consumer electron- ics, you will fi nd the specs for the tone controls. Th ey read something like, “Tre- ble control: ± 12 dB @ 8 kHz, Bass control: ± 12 dB @ 80 Hz.” Th is means that the knob marked “treble” will allow for up to 12 dB of boost or dip at the preset frequency of 8,000 Hz (8 kHz), and the knob marked “bass” will allow for up to 12 dB of boost or dip at the preset frequency of 80 Hz. Th e third parameter of EQ manipulation involves bandwidth, which re- fers to the breadth of the EQ activity over the frequency range. You might ask,, Th e Essentials DIAGRAM 2.10 45 Bandwidth fi lter parameters “When the specs say that the treble control operates at 8 kHz, does that mean that it boosts or dips only at exactly 8 kHz?” Th e answer is—of course, not! Not only would it be very diffi cult to limit the EQ’s activity to one exceeding narrow frequency, but it wouldn’t be very helpful in shaping the sound. Th e designation of a frequency (such as 8 kHz) for a particular EQ function indicates the center frequency. In typical EQ operation, there is a bell curve spreading equally above and below the center frequency. Th e breadth of the bandwidth is expressed in a range of Hertz or in portions of musical octaves (refl ecting the musical scale’s relationship to the frequency scale). Th e bandwidth setting is also referred to as the “Q,” which is short for the “quality factor” of the signal process because changes in bandwidth aff ect the quality, meaning the characteristics, of the sound. Th e bandwidth is defi ned by the breadth of the equalizing eff ect when it falls 3 dB down from its maximum boost or dip at the center frequency. Th e current generation of soft ware EQ plug-ins is especially user friendly because they provide a graphic representation of the EQ curve along with the standard knob controls with numerical readouts. Here is a screenshot of an EQ set to boost 3 dB at 3 kHz using a two-octave bandwidth: SCREENSHOT 2.5EQboost of +5 dB at 3 kHz with Q =.667, THE ART OF DIGITAL AUDIO RECORDING Here is a screenshot of a four-band EQ with three of the bands in use. Th e fi rst band is set to boost 3 dB at 80 Hz, the second to dip 1 dB at 250 Hz, and the third to boost 5 dB at 2.5 kHz using this EQ’s preset bandwidth (Q) setting. S CREENSHOT 2.6 Four-band EQ with three active bands Th e type of EQ shown here—one that has control over all three primary EQ parameters—is called parametric EQ. Fully parametric EQ has control over boost and dip, frequency, and bandwidth. Th e consumer audio paradigm usu- ally provides two-band EQ with user control limited to boost and dip—both frequency selection and Q (or bandwidth) are preset. Because of the fl exibility of soft ware EQ (there is no more expense in providing full-range controls once the code has been written), most plug-in EQs off er full parametric control along with other EQ functions. Some of the common variations on the three standard EQ parameters described above are: S helving EQ Th is refers to a diff erent approach to setting the Q. Instead of the bandwidth’s being a bell curve as shown above, the Q setting refers to the starting frequency and (when set to shelving) the boost or dip will aff ect all frequencies above (high shelving) or below (low shelving) the frequency selected., Th e Essentials H ere is an EQ setup using +5 dB of high-frequency shelving at 8 kHz: SCREENSHOT 2.7 High-frequency shelving H ere’s another set of parameters using -5 dB of low-frequency shelving at 120 Hz: SCREENSHOT 2.8 L ow-frequency shelving S helving provides particularly smooth-sounding alterations in frequency and is commonly used when a broad increase or decrease in either high or low frequencies is desired. High- and Low-Pass Filters Th is refers to EQ that sharply limits either high- or low-frequency sounds while allowing all other sounds to “pass” through the fi lter unaff ected. Th e terminol- ogy can be a little confusing: high-pass fi lters reduce (fi lter) the low frequencies and allow the high frequencies to pass through, while low-pass fi lters reduce the high frequencies and allow the the lower frequencies to pass through. You may be able to adjust both the frequency for the passing fi lter and how steep the drop-off is on the fi ltering., THE ART OF DIGITAL AUDIO RECORDING Here’s a high-pass fi lter set to 50 Hz with a steep drop-off curve of 12 dB/ octave: SCREENSHOT 2.9 High-pass fi lter Here’s a low-pass fi lter set to 5 kHz with a gentle drop-off curve of 6 dB/ octave: SCREENSHOT 2.10 Low-pass fi lter H igh-pass and low-pass fi lters are particularly useful in clearing up prob- lems such as low-frequency rumble or high-frequency buzzes. Of course, if some of the desirable sound occupies the same frequencies as the problem sounds, the fi lters will be removing both, causing unwanted eff ects along with desirable ones. Th ese fi lters can be especially eff ective when used to clear subtle resonant, Th e Essentials SCREENSHOT 2.11 3 1-band graphic EQ and/or leakage when the problems are in frequency ranges that have no signifi - cance to the element being retained. A high-pass fi lter on a hi-hat track can help fi lter out low-frequency bass-drum leakage without aff ecting the sound of the hi-hat at all, and a low-pass fi lter on the kick drum can do the reverse, fi ltering highs that are not a signifi cant part of the kick-drum sound. Band-pass fi lters combine high-pass and low-pass fi lters to limit the frequencies on both sides of the spectrum. G raphic EQs G raphic EQs were used extensively before the more fl exible parametric EQs became common. Th ey are not frequently seen in either soft ware or hardware processors anymore. Th e “graphic” in the name refers to the fact that the layout of a graphic EQ allows you to see the EQ curve as a graphic representation. To create this graphic eff ect, a graphic EQ uses fader controls rather than ro- tary knobs for boosting and dipping frequencies. Th e frequencies are typically evenly spaced (based on their relationship to the musical octave), and the Q setting is predetermined (not user-controllable). Th e most fl exible (and some- times employed in professional settings to “tune” large monitor speakers to the acoustic anomalies of the room) are the 1/3 octave EQs that cover the range of frequencies at 1/3-octave intervals (requiring 31 bands to cover the entire audible frequency range).

Analog and digital EQs

Th e debate over whether digital EQs are as good-sounding as their analog pre- decessors continues, or perhaps I should say that it actually exists now, because up until fairly recently most professionals agreed that the high-end analog EQ sounded considerably better than any of the soft ware alternatives. As soft ware, THE ART OF DIGITAL AUDIO RECORDING developers have become more and more sophisticated in their programming (and the computers have become more and more powerful, and thus capable of running complex, CPU-demanding soft ware), a true debate about the relative merits of the two has arisen. Soft ware developers have also taken to sophisti- cated modeling of analog units (oft en either with the blessing and aid of the original manufacturers or actually developed in-house by a soft ware division of 50 the original manufacturers). Th ere is the additional matter of digital-to-analog conversion (and back again), which is necessary when using analog EQs on dig- ital audio; this raises its own questions regarding the extent to which conversion might negatively aff ect the sound. And, of course, relative cost is oft en a factor as well, with the best soft ware EQs costing much less than the high-end analog units. Th ere is the additional benefi t with soft ware that separate instances of the soft ware EQ can be used on many channels in the same session, while the analog unit is limited to a single use in any given session (unless its eff ect is recorded to a new track, so that it can be used again). Regardless of where one stands on the digital versus analog EQ debate, most agree that the newest digital EQs continue to sound better and better.

EQ and phase

I t is inherent in the nature of normal operation that applying EQ will alter the phase relationship of the sound that is being processed. Th is is because there is a certain amount of time required for the EQ to process the frequencies that it is acting on, and so those frequencies get shift ed in their time relationships to other frequencies that make up the sound. Th is time shift creates changes in the phase relationship. Developers have found ways to minimize the negative ef- fects that such phase shift ing might cause, but it is not possible to eliminate the eff ect completely. As with virtually every kind of processing, there is a some- thing sacrifi ced in exchange for what is gained. Using EQ will compromise the sound in certain ways, but it may enhance the sound in others. You need to balance the trade-off s. R ecordists may speak about EQ in terms of decibels of boost and dip, and refer to certain frequencies with various Q settings or shelving characteristics, but to many musicians this will be meaningless. Finding the right nontechnical words to communicate about EQ and other recording qualities can be a valu- able skill in managing the creative give-and-take of making recordings. (Th is is explored further in section 6.4).

Human hearing and the use of EQ

When it comes to understanding EQ, it is valuable to consider the characteris- tics of human as controlled by the capabilities of the human ear. Two research- ers defi ned these characteristics in 1933, and their description, known as the Fletcher-Munson curve, became the standard for understanding the biases of, Th e Essentials human hearing. Since then, the curve has been further analyzed and refi ned, but for our purposes it isn’t necessary to go into the details of this analysis (it is available in many other sources). Th e critical information supplied by the Fletcher-Munson curve and its successors is the nature and extent of the loss in sensitivity of human hearing to certain frequencies. Th at is, sounds become more and more diffi cult for us to hear as they get higher and higher, or lower and lower, until they pass into frequency ranges beyond our capacity to hear at 51 all. What’s more, this eff ect is compounded when sounds are played more qui- etly. Th e quieter the sound, the less capable the ear is of hearing its higher and lower frequencies, instead focusing its abilities on the upper midrange (where the primary harmonics of singing and talking reside; these are the frequencies that help us diff erentiate vowel sounds). Th is is sometimes explained as natural selection for our ability to understand the human voice without distraction— especially important when external forces threaten lives and the focus is neces- sarily on communication. A t the same time, musical sounds are, to a large extent, defi ned by their timbre (the quality of the sound), and the timbre is primarily determined by the nature of what is called the o vertone series . Th is is what explains the diff erence in sound between a piano that plays the note middle C and a guitar that plays the same note. Th e note is defi ned by its fundamental pitch (or frequency), and the fundamental frequency of these two notes is the same (middle C). Why, then, does the guitar’s note sound so diff erent from that of the piano? Th e dif- ference is in the timbre, or quality of the sound, that is a result of the particular process used to create the sound, interacting with the physical qualities of the instrument being played. Th us, the overtones of a middle C created by a piano string struck by a piano hammer and resonating within the chamber of a piano are much diff erent from those created by a guitar string struck with a plastic pick (or fi nger) and resonating inside the body of the guitar (or in the speaker of a guitar amplifi er). Th e diff erences create the timbres that make it easy for us to distinguish between a piano and guitar, even when they play the same note, such as middle C. Th e overtones series is made up of a series of higher tones than the fun- damental or root tone that gives the note its name (and its primary pitch). If we combine these two facts—(1) the human ear loses sensitivity in the higher and lower frequencies; and (2) the musical quality or timbre of a sound is largely the result of the higher frequencies created by overtones—we start to see some of the reasons for a particularly common approach to EQ-ing, which is the basis for an EQ approach sometimes referred to as the l oudness curve, or the smile curve. On some consumer-electronic units, there is a button marked “loudness,” and this button introduces EQ that is a response to the factors listed above. Th e loudness curve boosts high and low frequencies, leaving midrange frequencies unaltered. In doing so, it seeks to make up for the loss of our ability to hear these frequencies when music (or anything else) is played more soft ly (that’s why it is, THE ART OF DIGITAL AUDIO RECORDING called the “loudness” function—it is intended to increase the “loudness” during soft playback). By enhancing the lows and the highs, the loudness button is em- phasizing the tonalities that our ear starts to lose at lower volumes, and it is em- phasizing the overtones in order to maintain the musical timbres of the sound. Th e loudness curve is intended to enhance the tonalities that are lost during low-level listening, but the same approach might be applied to louder 52 sounds. Th at is to say, even at louder volumes, the qualities that are a crucial part of musical timbre—the overtone series—may be enhanced with the use of high-frequency boosting. Th is application of the loudness curve is also referred to as the s mile curve because of the shape it creates on a graphic EQ. Th e follow- ing is a typical smile curve on a graphic EQ. SCREENSHOT 2.12 Graphic EQ set to a smile curve W hile EQ remains the most powerful and frequently used tool for sig- nal processing, it can certainly create unintended eff ects. For example, EQ can enhance the natural resonances of musical sounds, it can shape the timbre of sounds to help them fi t well with other sounds (covered in section 6.2), and it can fi x problems that have specifi c frequency characteristics (covered in “Filter- ing and Fixing Problems,” below). As already noted, EQ always alters the phase relationships, and this in itself refl ects a certain compromise with every appli- cation of EQ. EQ can also diminish the quality of recorded sounds in ways not understood or necessarily noticed by the recordist. O ne of the biggest pitfalls in using EQ is that when it is used to boost selected frequencies, EQ also boosts the overall volume (gain) of a sound. Our ear tends to respond favorably to louder sounds (up to a point, of course), so when you boost the high frequencies of a sound, this can be irresistible; as you turn the boost knob up, your ear is causing you to think “Th at sounds better!” until the boost becomes obviously overdone. Unfortunately, what a recordist, Th e Essentials might perceive as “better” is too oft en simply “louder,” and in the process the recordist has created excessively thin and harsh sounds or excessively boomy and indistinct sounds. We are lulled into thinking positively about a sound because it is louder, even though a true comparison between the EQ’d and un- EQ’d sound (played at relatively the same volume) might make us chose the un-EQ’d version. Learning to EQ is a process that involves a lot of back and forth—trying 53 more and less boost and/or dip and then listening both to the isolated sound (in solo mode) and the sound within the ensemble. A/B-ing between the EQ’d sound and the fl at sound (no EQ), trying to adjust your decision-making pro- cess to the understanding that the EQ’d sound has changed in volume as well as tonality, is an essential part of using EQ. Most books on recording encourage you to take a minimalist attitude toward EQ-ing, and some promote the “all cut” approach, which makes the gain issue work in reverse (the un-EQ’d sound is louder and therefore perhaps more appealing because all the EQ-ing is cut- ting frequencies). Minimalist approaches are oft en the right way to go, and even the “all cut” approach is sometimes best, but there are times when extensive EQ-ing is called for. My own experience indicates that most recordists do quite a bit of boost EQ-ing in mixing, and that even as much as a 10 dB or more boost on certain elements may be right for the mix when and if the situation calls for it. Of course, there is really no right or wrong when it comes to EQ-ing (or any other recording practice), but there is a diff erence between making decisions based on understanding and mistakenly identifying changes in volume as im- provements in sound.

U sing EQ on input and/or output

W hen to EQ is sometimes just as important as how to EQ. Th e question arises especially in regard to whether EQ should be applied during the recording pro- cess (on the way in to the DAW) or during mixing (on the way out of the DAW). Logic argues for delaying the application of EQ until the mixing process. Th at is, signals that are EQ’d during input are permanently altered by the EQ, whereas EQ applied during mixing can be repeatedly revised without aff ecting the origi- nal recording. For the most part, this logic represents wise operating procedure and personally I apply very little EQ during recording. However, there are ex- ceptions to this rule—no blanket operating procedure will be right for every situation. Several diff erent circumstances might warrant the application of EQ dur- ing the recording practice. Th e most common is when recording sounds that typically end up being processed with a considerable amount of EQ. Th is is dic- tated by the sound itself and the musical and technical circumstances of the re- cording. An example is recording a drum set for most popular music genres. On, THE ART OF DIGITAL AUDIO RECORDING the technical side, drum sets consist of so many diff erent-sounding instruments (from kick drums to cymbals) that it is oft en desirable to EQ out frequencies that don’t relate to the specifi c drum-set element being recorded (e.g., removing low frequencies from the mic that is recording the hi-hat). On the creative side, contemporary drum sounds oft en involve highly processed sounds (very bright snare drums and/or booming kick drums). When a lot of EQ is going to be used 54 to achieve the fi nal sound, it is usually desirable to use moderate amounts when recording and moderate amounts again when mixing. Th is doesn’t overly tax the capabilities of the individual EQ, and it can help minimize phase problems by boosting at diff erent frequencies between input and output. Th e availability of hardware EQ may also dictate some EQ usage when recording. If you are working in a DAW and want to remain in the digital do- main throughout the project, then input is the one chance to use analog EQ without going through an extra stage of conversion out to analog and back to digital. If you have access to a particularly desirable-sounding analog EQ when recording, and if you’re pretty sure about some degree of EQ-ing that you’re going want on a particular element of the recording, you might take advantage of the situation and apply some of the analog EQ during input. Creative use of EQ means responding to your circumstances and planning for the likely use of each element that you are recording, while at the same time recognizing the advantages of postponing EQ-ing decisions where possible.

Filtering and fi xing problems

EQ is primarily a creative tool, but it also can be a problem solver. Buzzes, hums, fan noise, machine noise, and the like are best eliminated prior to mak- ing the recording, but this is not always possible. EQ can be used to minimize the eff ects of unwanted sounds, though there is usually some compromise in doing so. F iltering refers to the dipping of chosen frequencies—they are being fi ltered out. You can fi lter out buzzes and hums, but that oft en requires pretty broadband action (dipping across a fairly wide spectrum of the frequency range). Doing this oft en impacts the sound that you are trying to preserve. For this reason, it is usually impossible to completely fi lter out unwanted elements, and you have to make a creative decision about what point is the optimal com- promise between diminishing the unwanted sound and negatively aff ecting the parts of the sound that you want. Th ere’s no right way to make such a deci- sion—though, again, listening to a variety of options both in isolation (solo) and integrated with the rest of recording is a good way to go about coming to that decision. Occasionally, problems are completely resolved using fi ltering, such as the need to get rid of a 60-cycle hum (a hum at 60 Hz sometimes created by bad AC grounding) for a recording of the triangle. Th ere is no discernable part of the triangle sound at 60 Hz, and this particular grounding hum is restricted to a, Th e Essentials very narrow band of sound. A good fi lter can make the hum inaudible without aff ecting the sound of the triangle. Th ere are other EQ-type tools that can be valuable for fi xing problems, such as de-essers and multiband compressor (sometimes called dynamic EQs). Although these tools really function primarily as EQs, I discuss them in the following section on dynamics because you must understand the basics of com- pressors to understand how they function. 55 2.6 Dynamics (Compressors and Noise Gates) Th e most mysterious and misunderstood tool in signal processing is the com- pressor; however, it is also one of the most valuable and widely used. What does a compressor do? Why do you use compressors and limiters (and what is the diff erence between the two)? What about expanders and noise gates? What is a brickwall limiter? Th e answers to these questions, along with details of the use and operation of dynamics processors, are covered in this section.

What compressors/limiters do

Dynamics refers to the changes in volume found in almost any audio program material. Certain elements (words, notes, beats, etc.) are louder or quieter than others, and the combined eff ect of these variations in volume create audio dy- namics. Compressors and limiters are dynamics processors, which means they operate to control these changes in volume (dynamics). Compressors and lim- iters function very similarly; the fundamental diff erence is in the strength of the processing. Limiters are strong compressors. I clarify this point below, in describing the specifi cs of compressor/limiter operation. Dynamic range is described as the range between the quietest sound and the loudest sound in any particular piece of audio. Th e basic action of compres- sion is to limit the audio’s dynamic range. Th at means that a compressor reduces the range (or distance) between the quietest sound and the loudest sound. It does this by reducing the volume of the loudest sound without aff ecting the volume of the quietest sound. Below is a screenshot of a single vocal line before and aft er compression. Th e height of the waveform indicates the volume, and you can see that the louder sounds have been reduced while the quieter sounds remain at the same height (or volume). SCREENSHOT 2.13 Vocal recording before and after compression, THE ART OF DIGITAL AUDIO RECORDING

Why use compressors/limiters?

Controlling the dynamics is an important part of contemporary audio produc- tion, but compressors and limiters have a more technical function in regard to making recordings, as well. As a technical aid, compressors and limiters help prevent overload—distortion caused by audio levels above the recorder’s 56 or DAW’s capability. Because a compressor/limiter reduces the volume of the loudest sounds, it can prevent an unexpectedly loud sound from exceeding the recorder’s capacity and becoming distorted. In the studio environment, it is usually possible to do enough level checking to set recording levels within a safe range, though not always (the unexpected can still happen). In live record- ing situations, with more unknowns, compressor/limiters are especially useful in protecting against audio overload. In terms of production uses—as enhancements for audio—compressors/ limiters are used in both subtle and obvious ways. Th e most typical use of com- pression is quite subtle—unlike EQ, where eff ects are oft en obvious even when used in moderation. Compression is also somewhat counterintuitive: why re- duce the dynamic range of a musical performance? Isn’t it dynamics that pro- vides some of the most expressive and creative aspects of a performance? Yes, but reducing the dynamic range can enhance recordings, and so compression is widely used in popular music production. Consider a vocal performance on a recording with many other instruments. In the fi nal mix, many elements might be competing with the vocal for space (bandwidth) in the musical spectrum. Because we oft en wish for the vocal to be very present—for the listener to be able to hear all (or at least most) of the words and even to understand all (or at least most of the) of the lyrics—a wide dynamic range in the vocal performance can frustrate the attempts to create a satisfying blend of elements. If you make the vocal loud enough to hear the quiet words, the loud words may be too loud and seem out of balance with the band. If you balance the loud words with the band, the quiet words may be lost. Compression evens out the dynamics and al- lows you to consistently hear the vocals without passages that are either annoy- ingly loud or so quiet as to get lost among the instruments. When used in this way, recordists usually want the compressor to be as transparent as possible. Th at is to say, you don’t want to hear any audible change in the vocal sound, only a reduction in the dynamic range. Th is eff ect is oft en quite subtle, though its overall eff ect on the balance of instruments would be obvious to a trained ear. Compression may also be used more aggressively to produce much more apparent changes in the sound of certain instruments. Th e most obvious case is with the compression eff ects used on many popular music recordings of the drum set. When strong compression eff ects are used on percussive sounds, there can be a dramatic change in the tonal quality of the instruments. Per- cussive sounds have a lot of energy—complex waveforms in brief sounds that include a lot of transients (short bursts of high frequencies)—and when this, Th e Essentials energy is compressed it can produced explosive-sounding eff ects. Highly com- pressed drums have become a hallmark of certain genres of music, including a great deal of rock.

H ow to record with compression in a DAW

As a practical matter, using a compressor when recording into a DAW requires either a hardware compressor before the signal enters the DAW or proper rout- 57 ing within a DAW. Just putting a compressor plug-in on the track that you’re recording does not allow you to record with compression. Th at’s because plug- ins are inserted in the record channel aft er the audio has been recorded. A com- pressor on your recording track will apply compression to what you are hearing, but it will do so aft er the signal has been recorded, so even though you’re hear- ing the compressor working, its operation will not have been recorded along with the signal. To record with compression in a DAW, you must route your signal through the compressor before it arrives at your record track. To do this, you need to create an aux track, place a compressor plug-in on that track, and then route the signal from the aux track to the track you are recording on. Th is means that the input of the aux track will be the microphone input, and then the signal will be output from the aux track via a buss to the recording track. SCREENSHOT 2.14 Recording with compression in a DAW, THE ART OF DIGITAL AUDIO RECORDING Th e recording track’s input will match the output of the aux track (let’s say they are both set to buss 1), and then the output of the recording track will go to the stereo buss as usual. You set the compressor controls as desired by monitoring the input (see compressor operations, below). In this way, the compressor is processing the signal being recorded.

How to use compressors/limiters—basic controls

Compressors and limiters operate by detecting dynamics (volume) and then reducing the volume of louder sounds and allowing the quieter sounds to pass through unaff ected. Th e detection devices vary and will be covered in a later section that discusses types of compressors and limiters. Every compressor has two primary parameters: threshold and r atio. You always control the threshold, whereas the ratio may either be preset or user con- trollable. Th e threshold controls the compressor’s actions; it controls what ele- ments are compressed (reduced in volume) and what elements are unaff ected. Th e ratio refl ects the extent to which the elements that are compressed have their volume reduced. Th e threshold is expressed in decibels because it sets a decibel level (vol- ume) at which the compressor is activated. You can think of the threshold as a doorway to level reduction. If the audio does not achieve enough volume to get up to the door (the threshold), the audio is unaff ected. If the volume gets past the doorway—is loud enough to go over the threshold—the compressor re- duces the volume of the sound. Any portion of the sound that is louder than the threshold will have its level reduced. Once the volume drops below the thresh- old, the sound is no longer aff ected. Th e lower the threshold, the greater the amount of original audio will exceed the threshold and the more compression will take place. With a higher threshold, fewer elements of the original audio will be aff ected. Th e ratio setting on a compressor defi nes the extent to which the volume that exceeds the threshold is reduced. Two numbers describe ratios: the fi rst indicates the amount in relation to the second number, which is always 1. Th us, a ratio of 2:1 describes compressor action that will reduce the volume of any sound over the threshold by a factor of 2 to 1, meaning that for each 2 decibels that the sound exceeds the threshold, the compressor will reduce that volume to only 1 decibel. A ratio of 4:1 means that the compressor reduces each 4 decibels of volume over the threshold down to 1 decibel. In this latter case, the portion of the sound that originally exceeded the threshold by 8 dB would exceed the threshold by 2 dB when exiting the compressor. Compressor ratios can also be variable; see the following section on advanced controls and the discussion of compression knee variables. Th e following diagram shows a graphic of a waveform being processed by a compressor. Th e threshold is set to –24 dB. In the fi rst diagram, the ratio is set, Th e Essentials to 2:1. Th e left side shows that for each 2 dB above the threshold, the audio has been reduced to 1 dB above the threshold (and fractions thereof: 5 dB above the threshold will be reduced to 2.5 dB above the threshold, etc., maintaining the 2:1 ratio). Th e right side of the diagram shows the same audio with the same compressor threshold, but with a 4:1 compression ratio. DIAGRAM 2.11 Compression ratios C ompressors have a third primary control function, aft er threshold and ratio, and this is g ain control . Th is control is sometimes labeled as “make-up gain” and that describes the reason for its existence. Because a compressor re- duces the dynamic range of audio by reducing the volume of the loudest sounds, the overall eff ect is that compressed audio is quieter. It is apparent from the dia- gram above that the compressed sound has been reduced in volume because the waveforms are smaller. Th is can make the audio diffi cult to use because the volume may no longer balance with other elements, especially if aggressive compression (a high ratio and/or a low threshold) is used. For this reason, com- pressors have an output gain control, allowing you to turn up the overall gain of the signal exiting the compressor, allowing you to “make up” for the lost gain caused by the action of the compressor.

Compressor and limiter metering

Th e meter on a compressor shows the degree to which it is reducing the audio signal level and may also show both input and/or output levels. Some com- pressors show all three at once, and some have the ability to switch the meter function to allow you to view any of these three levels on a single meter. Th e metering function that indicates the amount of compression is displayed in the reverse direction of a normal meter, because it is indicating a loss of gain. Th is means that a typical VU-type meter will begin at the 0 dB designation, and as the compressor acts on the audio, the meter will show defl ection moving to the left , indicating the amount that the signal is being reduced. Th e screenshot on the following page shows a compressor meter prior to any compression activity and then with 3 dB of compression (a reduction in level of 3 dB)., THE ART OF DIGITAL AUDIO RECORDING S CREENSHOT 2.15 C ompressor metering

Th e diff erence between compressors and limiters

U p until this point, I have either used the compressor/limiter designation or just referred to compressors in the discussion. As I noted in the introduction to this section, compressors and limiters function similarly, the fundamental diff erence being in the strength of the processing. Limiters are strong compres- sors. Limiters are made into strong compressors by their use of high ratios. Th ere is no exact defi nition of what ratio turns a compressor into a limiter, but it is generally understood that ratios of 20:1 or higher may separate limiters from compressors. Brickwall limiters are a diff erent kind of processor (though they share the basic idea of a limiter) and are covered in a later section of this chapter.

How to use compressors/limiters—advanced controls

S ome compressors provide more extensive control over the compressor and limiter functions. Th e most common kinds of more advanced controls regulate the attack and release characteristics of the compressor, and what is called the “knee” function, or variable ratio control. When these functions are not con- trolled by the recordist, they are either preset in the processing unit or (more frequently now in plug-in processors) are program dependent, which is to say they vary depending on the program material (sound) that is detected by the unit, automatically adapting settings to fi t the nature of the sound. On the one hand, the detection circuitry in compressors is remarkable in its ability to detect sound levels as they approach and pass the threshold level so as to begin acting on the sound very quickly. Soft ware compressors can detect sig- nal levels in as little as .01 milliseconds (one ten-thousandth of a second!). Th is allows the compressors to control levels without any audible delay. However, it is not always desirable to “attack” a sound with compression very quickly. Many sounds contain a lot of leading-edge high-frequency components (transients) that are an essential part of the vitality of the sound. Sometimes these transients, Th e Essentials are problematic and can be tamed with a compressor set to a fast attack, but more frequently compressing these parts of the sound creates dull-sounding audio and robs the sound of its most distinctive characteristic. For this reason, using a relatively slow attack oft en produces a desirable compression on certain types of program material, such as most percussive sounds (drums, etc.) and any sound made by striking an instrument (such as piano and guitar). R elease times on compressors may also be set by the recordist, preset by 61 the unit, program dependent, or some combination of the above. Th e manner that a compressor “releases” the eff ects of compression needs to correspond to the dynamic slope of the audio in order to prevent obvious compression arti- facts (such as an audible “pumping”). If the audio decays slowly and the com- pressor releases quickly, there will be an unnatural rise in volume. If the audio decays quickly and the compressor releases slowly, the following audio may be compressed even if it is not loud enough to cross the compressor’s threshold. Because audio release characteristics oft en vary within a single musical per- formance, it is frequently desirable to use some form of program-dependent release setting, if available. Variable knee characteristics describe the ways that compressors might adjust ratio settings depending on the extent that the audio exceeds the thresh- old. H ard-knee settings maintain a constant ratio regardless of how far over the threshold a sound might be. Soft -knee settings vary the ratio so that the further the audio travels beyond the threshold, the higher the ratio and thus the stron- ger the compression. Generally speaking, soft -knee operation provides com- pression that is more consistent with musical dynamics as it scales the degree of compression to the level of dynamics. User-set ratio settings act as an overall scaling factor when in combination with soft -knee operation.

T ypes of compressor/limiters and their eff ects

Th ere are endless variations on compressor technology, especially now that they are created using computer code instead being of restricted by hardware ca- pabilities. Nonetheless, there are two basic kinds of compressor and limiters that refl ect the two most common hardware designs. Th e fi rst type is compres- sors that operate using tube technology or that simulate tube-based compres- sors. Th ese earlier compressors used optical sensors to react to dynamics and thereby apply gain control. Th e optical-type compressor has a natural variation in release times that is slower when sounds do not exceed the threshold too far (typically about 3 dB of gain reduction or less) and faster for greater levels of reduction. Th e second type is the more recent compressors that use electronic sensors (VCAs, or voltage control amplifi ers). Th ese have the reverse release characteristics—faster on smaller levels of reduction and slower on greater lev- els. Th ere are other technologies (tubes themselves; FETs, or fi eld eff ect transis- tors; and now proprietary digital processors) that may be used for “riding gain”, THE ART OF DIGITAL AUDIO RECORDING (compressing). In each case (and within each technology as well, depending on how it is implemented), there are variations in attack, release, and ratio charac- teristics that aff ect the sound. Asarule of thumb, it is the more contemporary-style compressors using VCA detection circuitry that will be the most transparent. Th at is to say, they will change the characteristics of the sound the least. Optical compressors tend 62 to have more “personality,” which means they change the sound more audibly. Th is may or may not be desirable, but it continues to fi nd widespread applica- tion and is oft en considered to enhance certain elements, especially vocals, bass, and drums in popular music production. C ompressors can also vary in how they read audio level. Th e RMS-level detection—root of the mean (value) squared—looks at average level over time, whereas peak-level detection reacts to the momentary audio peaks in level. Some compressors off er a choice between the two, and some off er control over the “window” size of the RMS readings; that is, as the RMS detection looks at a smaller and smaller window of sound for its average, it becomes more and more like a peak-detecting compressor. In general, RMS detection is better at general “leveling” compressor function, and peak compressors do better at tam- ing sounds with a lot of quick dynamic changes (like snare-drum tracks). Some recordists like to use the two in tandem, compressing peaks fi rst and then level- ing the output using a compressor that is reading the average level (RMS), or sometimes the opposite, if more dramatic leveling is desired. In general, the RMS-level detection functioning is going to producer gentler results. Variations in attack time also function similarly to peak versus RMS detection, with slower attack times producing more gentle leveling-type results and fast attack times better at taming sharp dynamic transitions. With the advent of soft ware compressors came the capability of look-ahead operation . Th is means that the compressor processes the sound with compres- sion and delay the output of the audio while it performed complex frequency and waveform analysis to provide the most transparent and musical kind of compression algorithms. Th is enables complex operations and some unique kinds of compression (see the following section on brick-wall limiters), but it may introduce signifi cant delay times that need to be accounted for, either through delay compensation or used in circumstances (such as mastering ap- plications) where delay is not a signifi cant factor.

Frequency-conscious compression: de-essers and

multiband compressors Another whole school of compressors falls into this category of f requency-con- scious compression because the compressor’s actions are also aff ected by varia- tions in frequency—the compressor is “aware,” or conscious, of changes in level within certain frequencies, as opposed to only responding to overall changes in, Th e Essentials level. Th ese compressors bear a relationship to EQ that also works on specifi c frequency ranges, and sometimes it is not clear whether it would be more ac- curate to call these processors EQs or compressors. In reality, these processors are both EQs and compressors, working in combination. Th e most common kind of frequency-conscious compressor is the de- esser. A d e-esser reduces sibilant elements in vocal performances (or in other sounds that have sibilant-like qualities). Because the s consonant is the most 63 frequent cause of sibilance, the processors are called de-essers, but they also operate on other parts of vocal performances—anything with a lot of very high- frequency information. De-essers work by using EQ’d versions of the original vocal signal to trigger a compressor. Th e technique for doing this uses a side- chain capability within the processor. Side-chain routing allows the user to send a second signal into the process and use that signal to trigger the processor’s action. DIAGRAM 2.12 De-esser plug-in routing Th e signal path used to de-ess a vocal track is as follows: the original vocal is routed to the de-esser, which is typically a plug-in that has been inserted on the vocal channel. Th e de-esser provides an EQ function that allows the pro- cessor to EQ the vocal in a way that greatly emphasizes the most prominent frequencies in s sounds and other similar sounds (very high frequencies). You don’t hear the EQ’d sound, but it is sent to the compressor within the de-esser. Because the s sounds have been so emphasized with EQ, they are the only (or at least the primary) sounds that will trigger the compression. Th e gain reduction that has been triggered by the exaggeratedssounds eff ectively turns down the original (un-EQ’d) sound. Only the elements triggered by the side-chain (the EQ’d signal) get compressed. Th e eff ect can be quite dramatic, turning down sibilant sounds considerably (depending on the threshold) and leaving every other part of the sound unaff ected. Multiband compressors work on a similar principle, but they off er side- chaining at a variety of frequencies so that frequency-dependent compression can occur at several frequencies at once. Th is is similar to a multiband EQ and,, THE ART OF DIGITAL AUDIO RECORDING SCREENSHOT 2.16 M ultiband compressor like an EQ, it can be used to either boost or dip at a variety of points along the frequency spectrum. Th ese multiband processors dip by compressing, but they can also be set to boost by expanding when triggered. Rather than side-chain- ing an EQ’d signal, multiband compressors use frequency analysis to identify frequency ranges from the original signal and use these to compress or expand at the frequencies set by the recordist. M ultiband compressors can be helpful with certain problems, especially when you are mastering program material that has already been mixed. For example, with a track that has a harsh-sounding vocal, you could use EQ to roll off some of the high mids to reduce the harshness, but that would tend to make the track sound dull all of the time. With the multiband compressor set to compress the high mids, you could probably set the threshold for the high midrange compression to trigger off the lead vocal. In this way, the high mids would be reduced only when there was vocal, leaving the track unchanged dur- ing passages with no vocal. In general, program material that has unwanted buildup in certain fre- quency ranges at certain times might best be handled with a multiband com- pressor. Th is means that mastering is the most likely place for multiband com- pressor processing, and fi xing problems is the time it is most appropriate to be put it into action. Beware of using multiband compressors too frequently—EQ and standard compression produce more consistent and predictable results.

Brickwall limiters and maximum level

D igital audio has changed the meaning of “maximum volume.” With analog, the fi nal maximum gain of any particular audio element was limited by a va-, Th e Essentials riety of factors, including the ability for a needle to track high gain on a vinyl record. In an absolute sense, there is a limit to the volume in any analog system or storage medium if the goal is to prevent distortion and other compromised audio artifacts, but it is confused by the perception that certain kinds of analog overload distortion, on certain instruments and in certain circumstances, may be considered desirable. Th ere is no such confusion in regard to maximum level in digital audio. 65 D igital audio converts gain from analog sources using a scale that cul- minates at digital zero. Digital zero represents a “brick wall,” in the sense that digital audio is unable to eff ectively process any incoming sound that exceeds digital zero. Any such sound will turn into distortion (digital noise). Unlike analog distortion, where the sound may break down gradually as distortion in- creases, all digital distortion is characterized by the same basic qualities (which are exceedingly unpleasant by any typical musical standard). I n order to prevent digital overload, and to maximize gain potential, a variety of soft ware processors known as brickwall limiters has been developed. While these are a part of the larger compressor/limiter family—they reduce the level of loud sounds based on a user-defi nable threshold—they function quite diff erently and are used for diff erent purposes than the typical compressor or limiter. A typical brickwall limiter has two basic controls: the threshold and the output ceiling. Th e output ceiling represents the loudest level that the processor will allow the audio to achieve. Th is functions as a brick wall, or infi nity to 1 (∞:1) limiting ratio, which means that it allows no overshoot beyond the ceil- ing that is set. While this ceiling might be set to digital zero to attain maximum level, it is typically set just shy of digital zero—oft en to -0.2 dB—to avoid prob- lems that processors may have trying to reproduce a lot of audio at digital zero. In order to create an absolute brick wall—no overshoot—these processors use look-ahead technology that utilizes complex algorithms to analyze and process audio prior to its output. Unlike a typical compressor, the threshold control on a brickwall limiter increases the overall volume of the incoming audio. Th e threshold increases DIAGRAM 2.13 Brickwall limiting, THE ART OF DIGITAL AUDIO RECORDING the gain linearly—for each 1 dB in lowered threshold, there isa1dB gain in overall level. As a result of the increase in gain, any elements of the incoming audio that exceed the output ceiling are limited to the absolute maximum set by that ceiling. For example, if the threshold is set to -6 dB and the output ceiling is set to -0.2 dB, the incoming audio will be boosted by 5.8 dB (the diff erence between the 6 dB threshold and the -0.2 dB ceiling), and any of that audio that 66 would exceed -0.2 dB of gain on the digital scale will be completed limited, to stay within a maximum output of -0.2 dB. Th e eff ect of this action is to make the overall level of any audio that has not reached the output ceiling louder by up to 6 dB, while any audio that would have exceeded that limit is set to -0.2 dB. Th e primary use of brickwall limiters is for mastering and their use is discussed more thoroughly in chapter 7.

Expanders/noise gates

E xpanders and noise gates are the opposite of compressors and limiters. Rather than decreasing the dynamic range of audio, they increase it. Expanders operate using the same basic control parameters as compressors/limiters. Noise gates are simply more powerful expanders (they utilize a higher ratio), following the same model as limiters, which are more powerful compressors (utilizing a higher ratio). Although expanders and noise gates fi nd a variety of applications in audio, they are much less frequently used than compressors and limiters. E xpanders allow the audio that exceeds the amplitude threshold to pass through unprocessed while it processes (reduces the gain) of the audio that does not exceed the threshold (again, the exact opposite of compressor action). Th e level of the quieter sounds is reduced based on the ratio (with high ratio settings yielding greater gain reduction). Expanders do not need to have gain make-up controls because the louder sounds have been unaff ected. Expanders and gates are useful in certain circumstances when you wish to reduce background noise or leakage from adjacent sounds. Th is is particular true in live recording situations. Th ere has been progressively less use of ex- panders in studio applications because of the capabilities of digital audio, which allow relatively easy elimination of unwanted parts of recordings. Functions such as “strip silence” work like an expander in separating sounds based on a user-defi nable threshold, but they operate as an editing tool rather than as a real- time operation, off ering more fl exible gating-type functions. Th e editing pro- cess allows you to adjust the results of expansion in many more ways than would be possible with an expander operating in real time. A typical example is a noise gate set on a snare-drum track to reduce the level of leakage from other drum-set elements when the snare drum isn’t playing. If a dynamic drum roll were played on the snare (from soft to loud), the soft hits would likely be gated by a typical real-time expander. By using “strip silence” to edit the track, you can go back and retrieve the soft snare-drum sounds that were below the gating threshold., Th e Essentials 2 .7 FX: Delay Th e most common eff ects are all delay-based, generally emulating what hap- pens to sound in diff erent environments. Th is means that these eff ects add de- layed versions of the original sound, just as acoustical environments add delays caused by the sound’s bouncing off of surfaces and returning to the listener slightly later than the direct sound that comes from the sound source (as shown 67 in Diagram 2.1 at the beginning of this chapter). Delay-based eff ects include reverb—the most complex and natural simulation of acoustic environments— and echo eff ects (delays) that provide simpler, more stylized simulations.

L ong and medium delays

A lthough single, discrete repeat delays (sometimes called “echoes”) that are nearly identical to the sound source do not actually exist in nature—any natural delay is somewhat compromised (less discrete than the original)—they are used frequently in recording to simulate the eff ect of acoustical environments. Long delays simulate larger environments where the sound travels to a distant wall or surface and the time it takes for the sound to return to the listener makes it discernable to the ear as a delayed signal. Th e most obvious example of this ef- fect is in a very large concert hall or church (or something like a rock quarry), where the listener can hear a very distinct echo of a word aft er the entire word has been spoken. In a very “live” environment (one with highly refl ective sur- faces), the delay or “echo” will repeat many times as it bounces back and forth between walls, diminishing in volume each time as the sound waves lose energy with each trip through the air. When we simulate this eff ect using a delay pro- cessor, the ear perceives the sound as having been made in a large acoustical environment. A typical delay unit has a control for the length of the delay and for feed- back. Th e length of delay is usually set in milliseconds, though there may also be settings based on musical time (e.g., one quarter note, one half note, etc.). Th e feedback controls the number of repeats, with each repeat diminishing in volume to simulate the oc- currence in nature. A setting of 0 feedback yields one dis- crete delay. L ong delays are usually about 250 ms (1/4 of a sec- ond) or longer. Long delays are usually used with some feedback to simulate the re- peating echo of large spaces. Delay times between 100 and S CREENSHOT 2.17 175 ms are medium delays S ingle delay plug-in, THE ART OF DIGITAL AUDIO RECORDING and are sometimes referred to as s lapback delays, as they provide a short but audible delay that suggests a medium to large acoustical environment. Slapback delays are typically just one discrete delay, no feedback. (In section 6.2, I explore more specifi c uses of long and medium delays when mixing.)

S hort delays—chorusing, doubling, phasing, and fl anging

Short delays, typically between 1 and 50 ms, provide a very diff erent kind of eff ect than the medium and long delays described above. Short delays are not primarily used to simulate room environments; rather, they are used to pro- vide doubling or thickening eff ects. Th e primary model for short-delay use is chorusing. C horusing refers to the typical eff ect of choral singing when no two singers are perfectly aligned with each other. Neither are any two singers per- fectly in tune with each other. Th e combination of slightly time-shift ed and slightly pitch-shift ed performances creates the thick and pleasing sound of a vocal chorus. Th is eff ect is simulated with digital signal processors by the use of a modulated short delay. Chorusing eff ects typically use delay times between 20 and 40 ms. Th ere may be only one discrete delay or multiple discrete delays with slightly diff erent delay times. Modulation is the technique used to create small changes in pitch. Typi- cally, a low-frequency oscillator (LFO) is used to oscillate (shift ) the pitch of the incoming audio. Th e waveform of the LFO nudges the pitch in a regular pattern back and forth from sharp to fl at. Th e depth setting controls the extent to which the pitch is shift ed and the rate controls the speed that the pitch is shift ed. Doubling uses one or more short delays without any modulation. Th is can thicken a sound (though it may make it more artifi cial sounding) without the regular cycling that is created by modulation. P hasing and fl anging are similar to chorusing but typically use shorter delay times. Defi nitions vary (there is no “standard”) but phasing is usually considered to use delay times in the 3 to 6 ms range and fl anging in the 1 to 3 ms range. Both use modulation, oft en deeper and faster than with a typical chorusing eff ect, and sometimes with feedback to produce even less naturalistic sounds. M any unusual sounds can be created using these kinds of delay-plus-mod- ulation eff ects. Settings can vary widely in regard to delay times, modulation depth and speed, type of waveform used for the LFO, and feedback—producing a wide variety of eff ects. Other controls such as phase reversal, EQ, fi lters, and D IAGRAM 2.14 Sine wave LFO, Th e Essentials multiple delay lines can increase the variety of these modulating eff ects. (In chapter 6, I explore more specifi c uses of short delays when mixing.) 2 .8 FX: Reverb R everb is short for reverberation and is the most realistic of the delay-based eff ects. Generally, reverbs simulate or reproduce the kinds of complex delays 69 found in an acoustic environment. Reverb consists of early refl ections, which are the quickest and most direct refl ection of sounds; and reverb tails (or late refl ections), which are the multiple refl ections that continue from the early re- fl ections. Th e large number of delays that make up the reverb tail are heard as a “cloud” of sound rather than as discrete delays. Th e early refl ections cue our ears in regard to the size and shape of the space, whereas the reverb tail cues our ears to the general “spaciousness” of the environment. Th e reverberation time, or length of the reverb, is generally defi ned by the time it takes for the delays to decay 60 dB from their original value. Th e reverb time is controlled by a combination of the size and surfaces of the room. Th e larger the room, the longer it takes for the sound to travel to the various walls and ceiling and return to the listener. Th e more refl ective (rigid) the surfaces in the room, the longer the sound will continue to bounce back to the listener. Concrete, brick, tiles, glass, and so on will provide longer reverb times, whereas carpets, drapes, and people (audiences, for example) will reduce the refl ections and the length of the reverb.

R everb devices

O ver the history of recording, many diff erent devices have been used to create the reverb eff ect. Th e most basic approach is to use a microphone to capture some of the natural reverb of the space as a part of the recording. It’s almost impossible to avoid doing this completely, but contemporary close-mic record- ing techniques do eliminate most of the natural reverberation of the record- ing space. Sometimes mics are moved some distance from the sound source to capture reverb along with the direct sound, and sometimes additional mics are used primarily to record the room ambience (reverb). Close miking became increasingly popular as techniques for adding reverb aft er the initial recording were developed. Th is gave the recordist more control over the size, quality, and amount of reverb. I n the 1950s, the e cho chamber became a popular technique for adding reverb. Th e echo chamber is a relatively small room (from the size of a closet to the size of small bedroom) that is generally all concrete and therefore very rever- berant for its size. A speaker is put in the chamber along with two microphones. Th e original signal is broadcast through the speaker and the microphones pick up the reverberated sound, which is then mixed in with the original recording., THE ART OF DIGITAL AUDIO RECORDING Th e size and refl ective characteristics of the room, along with the position of the microphones, will aff ect the length and quality of the reverb. O ther hardware reverb units are variations on the echo chamber—they feed the sound into a unit that creates reverberant delays that can then be added back in with the original sound. S pring reverbs (oft en found built into guitar amplifi ers) do this by using springs in a small metal box. Th ey tend to have a 70 somewhat crude (boingy) but distinctive sound. P late reverbs do the same thing with large metal boxes and have a much smoother sound quality, but they are large and expensive. R everb lends itself very well to digital signal processing, and digital reverbs have pretty much replaced most of the other, more cumbersome techniques. Hardware versions of digital reverbs thrived for many years, but they have been mostly replaced by the soft ware equivalent in the form of plug-ins. Digital re- verb plug-ins operate using two distinct technologies. Th e older format simu- lates reverb characteristics using complex algorithms to approximate acoustical spaces. Many of these simulations are very realistic and natural sounding, but this also provides the capabilities for creating reverberation-type eff ects that don’t occur in nature. More recent developments have allowed for the recording of acoustical spaces and the transformation of these recordings into impulse-response sam- ples that can be used in the same manner as any reverb device. Th e impulse- response reverbs require libraries of samples made from a variety of record- ings. Th ese reverb plug-ins are exceedingly natural sounding and some feature samples from famous acoustical spaces, including concert halls, auditoriums, churches, nightclubs, echo chambers, and recording studios. Digital reverbs can also either simulate or sample hardware versions of spring and plate reverbs.

Reverb control parameters

Th e most basic control for a reverb processor is the type of reverb, which is usually defi ned by the type of space being either simulated or sampled. Th us, reverbs typically have settings for concert halls, churches, rooms, plate reverbs, chamber reverbs, and so on. Because any environment can be simulated or sampled, sometimes this list is extensive and might include things like park- ing garages, stadiums, nightclubs, and bathrooms. Th e newer sampling reverbs oft en identify specifi c spaces (the Sydney Opera House, for example) that pro- vided the source samples for the reverb. Th e next basic parameter is reverb time or length. Th e reverb D IAGRAM 2.15 time is based on a combination R everberation impulse of size and degree of refl ectivity of response the surfaces. Th e confi guration of, Th e Essentials early refl ections and reverb tail, as well as the spacing of delays in the reverb tail (density), might be aff ected by the size parameter. Some reverbs allow you to balance early refl ections and reverb tail separately from the time parameter. Some split size and density into separate parameters. Th e predelay sets the amount of time before the reverb tail is heard. Th is aff ects the perception of room size. Large rooms will naturally have longer pre- delay times because of the time it takes for the sound to get to the far walls and 71 return to the listener. Predelay may also aff ect early refl ections. I n addition, reverbs may off er diff usion, decay, damping, envelope, and EQ parameters. Because of the complexities of reverbs, there are an enormous number of subtle qualities that may be user controllable. In practice, most re- cordists pick reverbs based on the space or quality of the sound that is desired. From the preset it may be desirable to adjust the time or size parameter and perhaps the predelay. It can be interesting to hear the very subtle diff erences in small parameter changes, but it can also consume a lot of time and may have negligible results. If you have to make large parameter changes to get closer to a desired sound, it is likely that you started with the wrong preset. It is gener- ally better to fi nd a preset that is close to the desired eff ect and make only small changes (unless very unusual sounds are desired). (In section 6.2, I explore more specifi c uses of reverbs when mixing.) 2 .9 Beyond Traditional DSP Th ere is a whole new world of digital signal-processing eff ects available since the advent of the DAW. Some, such as pitch correction, time compression and expansion without pitch shift ing, and sophisticated noise reduction, provide ca- pabilities never before available, and they have had a profound eff ect on music production. Others, such as guitar amplifi er simulation and analog circuitry simulation, seek to reproduce some of the capabilities from the analog world that were previously lost in the digital domain. Th e following is not meant to be exhaustive, and there are frequently all manner of new products.

Manipulating rhythm and pitch

Some of the unique new capabilities that have emerged in the era of the DAW have to do with manipulating rhythm and pitch (the fundamentals of music) in new ways. Besides the tremendous new capabilities in editing music, and thus altering rhythmic placement and even creating new rhythmic content, the DAW has brought the easy time compression and expansion (shrinking and stretch- ing) of audio. Of course, a variety of analog techniques were used for speed- ing and slowing audio, but these inevitably brought a corresponding change in pitch. Th e DAW can change tempos (speeds) without changing pitch. It does this by using algorithms to determine what to remove or add that conforms to the surround samples in a way to produce the most transparent results. Some-, THE ART OF DIGITAL AUDIO RECORDING SCREENSHOT 2.18 T ime compression or expansion times changes in speed result in audible artifacts that render the result unusable in a typical musical setting (especially with large changes in the time base), but oft en the result is not apparent. Th ere are more and less capable plug-ins that accomplish this, and it is an evolving technology. I’ve used some programs that have allowed me to speed up or slow down entire mixes by several BPMs without a change in pitch and without noticeable artifacts. Rhythm-altering soft ware has also been used to match samples of performances with diff ering tempos so as to combine elements that would not have conformed to the same musical fundamentals. In practice, this allows for the combining of beats from samples of diff ering tempos and for more complex combinations of elements as found in mashups. DAW soft ware has also been developed that allows for alterations in pitch. Th ere are relatively simple pitch-shift ing devices that can alter pitch without altering the time of the audio. Th ese plug-ins may be used to shift pitch in small ways that can be used, along with short delays, to create chorusing-type eff ects. Th e pitch shift doesn’t modulate at regular intervals as it does with a traditional chorus, but instead remains constant (perhaps between 5 and 9 cents sharp and/ or fl at). Th is technique can also be used for much larger pitch shift ing that can create standard harmonies, typically from thirds or fi ft hs, or you may choose more unusual harmony notes. Many of these plug-ins are “intelligent,” in that they will make appropriate choices for harmonies if supplied with the music’s key signature., Th e Essentials SCREENSHOT 2.19 A uto-Tune A dvances in pitch-shift ing devices have incorporated pitch-detection ca- pabilities, which then allow for the retuning of performances. Oft en referred to as a uto-tuning , these plug-ins (Auto-Tune and its competitors) allow pitch fi x- ing of vocal and instrumental performances either by automatically moving the pitch to the closest note in the scale selected or by allowing you to redraw the pitch graphically as desired. Besides being a tool for the correction of perfor- mances, Auto-Tune–type programs are being used to create new and unusual vocal eff ects that would not be possible for a singer to perform naturally.

Noise reduction

Tools for noise reduction originated with Dolby and dbx systems that were designed to reduce the tape hiss associated with analog tape playback. In the digital world, noise reduction has taken on much broader applications. Digital noise-reduction processors can reduce or eliminate broadband noise (including tape hiss and surface noise from transferred analog recordings), buzzes, clicks and pops, crackling, and so on. Th ese processors have been used extensively to “clean up” old recordings for reissue on CD. Noise reduction is accomplished through sophisticated detection algorithms and then combinations of fi ltering and compression/expansion routines that isolate and reduce the noise while having a minimal eff ect on the remaining audio., THE ART OF DIGITAL AUDIO RECORDING

A nalog simulation

F or all the problems with noise created by analog audio, there have also been many highly valued properties that are unique to analog systems. Th ese have been widely simulated in the digital realm. In fact, many of the digital signal processors available for DAWs are simulations of analog gear. Sometimes they 74 are simply modeled on a variety of analog hardware units, and sometimes they are attempts at faithful reproductions of the eff ects of a specifi c piece of gear. I say “attempts” because it is not possible for digital reproductions to create ex- actly the same eff ect as their analog counterparts. Nonetheless, a lot of research and development has gone into making as accurate reproductions of classic an- alog processing units as possible. Th is includes all of the processors discussed above, including EQs, dynamic processors, delays, and reverbs. Th e same is true for other analog gear, including guitar amplifi ers, tube processors, and tape recorders. Th e distinctive distortion provided by guitar amplifi er circuitry has been extensively modeled, as has the harmonic distor- tion created by tube processing of audio and saturation eff ects of analog tape compression. Elaborate soft ware that models the many possible eff ects of these various kinds of analog processing is available. For many recordists, it has be- come standard practice to record electric guitars directly (with no amplifi er or external processing) and to create the fi nal guitar sound using these soft ware simulations. Guitar amp simulators have also been used extensively on other instruments, and even vocals, to create distinctive eff ects. Other analog simula- tions of tube or tape recorder eff ects are routinely used on instruments and over entire mixes to subtly enhance the sonic character of recordings. Th ere is end- less debate in pro audio forums about the accuracy of these reproductions, but for most users the point is not whether the soft ware is an accurate reproduction of the original but simply whether the soft ware is producing a desirable eff ect. As always, it comes down to “What does it sound like?”

Vibrato and tremolo

A couple of standard eff ects that have been around for a long time, but they don’t fi t neatly into any of the above categories. Vibrato is a periodic shift ing of pitch (frequency) and tremolo is a periodic shift ing of volume (amplitude). Although these are the proper defi nitions, the two terms sometimes get confused, such as the tremolo bar on an electric guitar, which actually produces a vibrato eff ect, and the vibrato settings on some guitar amps, which actually produce tremolo. In practice, when produced by singers or on stringed instruments using fi nger and/or bowing techniques, there is oft en a certain amount of both eff ects being created at the same time. V ibrato is related to the modulating eff ect of chorusing, but it tends to be more pronounced. It is generally produced by the musician, as opposed to being controlled electronically. Th e periodic pitch shift ing adds interest to sus-, Th e Essentials tained notes, provides a thickening eff ect, and allows for a more forgiving rela- tionship to the center pitch of the note. A deep and wide vibrato is associated with certain musical styles and with various historical periods (older operatic singing, for example). Th e use of fi nger vibrato on the guitar is associated with certain seminal electric guitar players, including BB King and Eric Clapton. T remolo is most frequently heard on electric guitar and as part of certain keyboard eff ects. Guitar tremolo is associated with certain styles of country and 75 American roots music, and the spinning action of a Leslie speaker gives the traditional Hammond B3 organ sound a kind of tremolo eff ect.,

Chapter 3 Recording Sessions A Practical Guide

3.1 Setup S etting up for a recording sets the tone for the entire session. Careful and com- plete setup makes for smooth running sessions. If you are at a commercial stu- dio you may have help with your setup, but you will need to direct the assistants. Setups may range from the very simple to the very complex, but in any event it is best to do as much of the setting up before the session, and before the musi- cians arrive, if possible. Th is means having a good session plan. It’s best if you’ve been able to consult with those involved beforehand so you know what they are planning and expecting. If the plan calls for a variety of recordings that require separate setups, you should consider what you think is a realistic goal for the time allotted. You don’t want to set up for a bunch of things that you may well not get to, but you do want to do as much of the setup as you can in advance. Th is section divides setup into microphones, headphones, consoles, patching, DAWs, and then testing and troubleshooting. Careful and complete setup pro- cedures will save time and foster a creative working environment. Microphone setup S etting up the mics also means choosing the mics and having a plan for the num- ber and positions of mics for the elements being recorded. For complex setups, a written m ic plot (or input list ) is essential. Many studios have preprinted forms for mic plots that allow you to list the mics and the associated inputs. Except for simple setups involving three or fewer inputs, it is a good idea to write down the mic, the instrument, and the input points to avoid confusion in setup and patch-, Recording Sessions D IAGRAM 3.1 M ic plot ing. For very complex setups, you may also want to make a diagram showing where all the musicians and instruments will go (a stage plot ). O nce the mic plot is established, the best mic stand and mic cable available for the job need to be selected. Th e cable should be attached to both the mic and its point of input as dictated by the mic plot so that it’s ready to be tested. It should be properly positioned for recording, but if the musician hasn’t arrived or gotten his or her instrument set up yet, the mic should be place close to where it will be used. For instruments held by the musicians (horns, strings, acoustic guitar, etc.) the fi nal mic setup needs to be done in conjunction with the musi- cian so the individual can show you exactly how he or she holds the instrument when playing. Y ou will want to consult with all the musicians to make sure that the positions of the mics and the stands are not interfering with their playing in any way . Th e survey of instrument recording techniques later in this chapter has recommendations for specifi c mic and, in some cases, mic stand selection.

Headphone setup

Along with the mic setup, the headphones for each musician need to be set up and positioned. I devote the following section of this chapter to headphone, THE ART OF DIGITAL AUDIO RECORDING (also called cue or monitor) mixes, so as far as the basics of setup go, you just need to make sure that each musician has a properly working set of headphones located for easy access. Closed headphone designs that are made to limit leak- age from around the ear are necessary for studio work near an open micro- phone. H eadphones vary widely in terms of power requirements to achieve equiv- alent volume levels. Th is is why it is essential that either all of the headphones are of the same make and model or each musician has individual volume con- 78 trol for his or her own headphone. Th e overall headphone amplifi cation system is also important; you need to ensure that there is adequate power for every musician. Each set of headphones that is added to a system increases the load on the amplifi cation, so more power is required to drive more headphones. Headphones with higher ohm ratings require less power (and some models of headphones are available in diff erent ohm ratings), so this should be considered when purchasing headphones.

Console setup

By console setup I am referring to a hardware console or mixer; setup for the internal or virtual mixer within the DAW is covered in a later section. Th ere may not be any console setup for your session if you are simply plugging mics or line inputs (synths, etc.) directly into an interface and all processing and rout- ing is done within the computer. Of course, all microphones require preampli- fi cation before going into a DAW, so this must be supplied by the interface, a stand-alone mic preamp, or a console with mic preamps. For this reason, many studios with larger session requirements have hardware consoles with mic pre- amps and routing capabilities to send audio to the DAW. Sometimes the console is used for headphone routing as well. I nput setup on a console A hardware console is usually essential for large sessions, though this has been replaced in some studios with numerous stand-alone mic preamps and a patch bay. Th e advantage of a console is the ease of centralized operation, along with headphone mix and output monitoring capabilities. In a typical studio environ- ment, the console is interfaced with the wall panels from the studio for input and the DAW for output. Th is means that if you plug a mic into input number 1 on the wall panel in the studio, it is hardwired to input number 1 on the con- sole. Th e output of buss 1 on the console is hardwired to input number 1 on the DAW interface. More complex studio setups require that a patch be made in order to route the signal from the wall panel in the studio to the console and/or from the console to an input on the DAW interface. I f the console is acting as a series of mic preamps, then each channel strip will provide preamplifi cation and phantom power, if needed. Th e preamp con-, Recording Sessions D IAGRAM 3.2 79 Mic preamp to buss output to DAW trols the input level into the channel strip and the output fader controls the level from the console into the DAW via a buss. Setting the appropriate record level requires balancing the mic preamp input with the buss output, and reading the fi nal record level as shown on the channel meter in the DAW. For initial setup, you simply want to verify that all the connections have been made and that you are getting signal from the mic into the DAW. Levels should be kept low until the musician is available and fi nal record levels can be set. Monitor mix and Headphone mix setup: Console or DAW? Th e proper setup for control-room monitoring and headphone mixes depends on many factors, but the fi rst question that needs to be answered is whether the mixes should be created at the console or within the DAW. Th ere are advantages and disadvantages to each method. S etting up all of your monitoring functions in the DAW means that all of your setup will be retained from one session to the next—simply recalled as part of your fi le. It also makes creating rough mixes in the computer (for burn- ing to CDs or sending as mp3s) much quicker because what you are hearing is ready to be bounced down, all within the digital realm. I f you take all your DAW outputs and return them to separate channels on your mixer, and run your headphone mixes from the cue sends on each chan- nel, you have the ease of using hardware controls rather than struggling with the mouse and the virtual mixer controls in the DAW. However, your setup will not be saved from one session to the next, and taking rough mixes to put on a CD requires some extras steps. You will need to either record from the console’s stereo output back into the computer and then make the appropriate fi les or record into some other system such as a stand-alone CD recorder. However, if you are using a digital control surface you get the advantages of both systems: the DAW controls are mirrored in the control surface hardware, giving you the ease of using hardware controls, while all your level and processing functions are still retained within the DAW., THE ART OF DIGITAL AUDIO RECORDING Overdub situations usually involve considerably fewer inputs than initial tracking sessions and they oft en require editing as a part of the process. Levels generally stay fairly constant for large periods of time as well, so it is easiest to control everything from a control surface or within the DAW when doing over- dubs. If the DAW is interfaced to a console, this means simply monitoring the stereo buss from the DAW through two channels (or some other stereo return) on the console. Headphone mixes can also be routed to two channels of the console and those can be sent to the headphone mix as a stereo pair.

Patching setup

Patching, or interconnecting, all of the elements for a session can range from the very simple to the extremely complex, depending on the number of elements involved and the studios patching system. Studios have a variety of patching strategies and patch bays can vary widely in how they are wired. Problems with patching—whether incorrect patches, bad cables, or bad patch points—are some of the most common problems that slow down sessions, so an under- standing of patching and attention to patching detail is critical. Patching strategy Th ere is one simple rule for the best patching strategy: always patch from the source to the destination, following the signal path. In a typical patching situ- ation, this means starting by plugging the cable into the microphone and fi n- ishing by patching into the DAW or other recorder. Sometimes some of this patching is already done with dedicated patches, such as console outputs that that are wired to feed DAW inputs. Each patch follows the intended signal path from the source through whatever series of outputs and inputs needed to record the signal. Simple patching A simple patch might be plugging a mic cable into a microphone and then plug- ging that cable into a DAW interface that includes a mic preamp. Th is completes the chain from source to destination. A slightly more complex patch might start with the mic cable into a microphone, from there into a stand-alone mic pre or channel strip, and then the output of the mic pre would be patched into the line input of a DAW interface or a console that is already interfaced with the DAW and requires no further patching. Even with simple patches like this, it is always best to patch from source to destination—from output to input, following the signal path. Patch bay use As patching gets more complex and studios wish to streamline the process of interconnecting a variety of elements from a variety of sources, patch bays be-, Recording Sessions P HOTO 3.1 A console scribble strip indicating the stereo buss and two stereo cue mix returns come an essential part of the studio. Many consoles have built-in patch bays to simplify access to all the patch points needed to get in and out of the various con- sole functions. Patch bays can take on many diff erent shapes and sizes and use a variety of types of connectors. Single-point patch bays may use 1/4-inch, RCA, or TT (tiny telephone) connectors and multiple-point patch bays may use a wide variety of D-subconnectors that have anywhere from 9 to 50 patch points at each connection point (though not all patch points may be wired for use). P atch bays are centralized patching stations that facilitate the patching process. If a studio has a variety of recorders (DAW and/or tape based), out- board processors, mic preamps, and recording spaces, then patch bays become an essential element in functionality. Besides the fundamental in-and-out com- ponent of a patch bay, the use of normaled (and half-normaled) pairs of patch points allow patch bays to pass signal when connections are in their “normal” use but still allow the user to “break” the normal in order to create patches for alternative uses. Th e “logic” of normaled and half-normaled patch points is as follows: N ormaled T wo patch points are considered normaled when nothing is plugged into either jack and the signal is wired to pass from the top jack to the bottom (typically confi gured as one patch point above the other in the patch bay). For example, an external mic pre is wired to one jack and below that is a jack that is wired to line input number 1 on your console. With no plug in either jack, the mic pre goes right to input number 1 of the console. But plug a patch cord into either jack and the connection to the console is broken. When a patch cord is plugged into either jack, it separates the “v” part, breaking the connection between the, THE ART OF DIGITAL AUDIO RECORDING two patch points. If, for instance, you want to send the mic pre to a compres- sor before it comes into the console, you would break the normal by plugging a patch cord into the mic preamp jack and routing it to the compressor. In a schematic normaled patch points look like this: D IAGRAM 3.3 Normaled patching Half-normaled When two jacks are wired to be half-normaled, the connection is not broken un- less there is a cable into both connections. Th e mic pre’s input could be tapped at the top jack, but it would still go to the console. Plugging something else into the console’s output, however, breaks the connection from the mic pre. Th is kind of patch is useful if you want to send the signal from the mic pre to two diff erent recorders (that weren’t both accessible from the console). D IAGRAM 3.4 Half-normaled patching While patch points that are half-normaled can be used to eff ectively split a signal, sending it to two places at once, many patch bays also have mults , which are used to split signals. Wired in parallel, mults provide multiple patch points that off er as many outputs as there are patch points in the mult—excluding one of the mult patch points, which serves as the input. Because mults are wired horizontally, any patch point in a mult can be used for the input. Mults are com- monly used to send signals to auxiliary recorders (in which case, for stereo you will need two mults—one each for the left and right feed). Complex patching A complex patching situation might go as follows: a cable is plugged into a mic and from there connected to a wall panel in the recording room, the wall panel output has been wired to a patch point on a D-subconnector (D-sub) in the machine room, from there it is patched to another D-sub in the machine room, Recording Sessions PHOTO 3.2 A TT (tiny telephone) console patch bay that feeds the console inputs in the control room, the buss output of the console feeds a D-sub in the machine room, and from there it is patched into another D-sub in the machine room that feeds a wall panel D-sub in the control room, the D-sub in the control room is patched into a D-sub that feeds the inputs into the DAW. Th is signal path would be described by a series of outputs and inputs: • M ic out to wall panel in • Wall panel out to machine room D-sub in • Machine room D-sub out to machine room D-sub console in • Console buss out to ma- chine room D-sub in • Machine room D-sub out to machine room D-sub DAW in • Machine room D-sub DAW out to control room wall panel D-sub in • C ontrol room wall panel D-sub out to DAW in W hile patching can become very complex, as in the above example, if you adhere to the rule of patching by following the signal fl ow from PHOTO 3.3 beginning to end, it can be straight- A machine-room patch forward and you can have consis- bay with Elco and other connections tently good results., THE ART OF DIGITAL AUDIO RECORDING

DAW setup

U nlike a hardware mixer whose capacity is fi xed, a DAW’s mixer confi guration can be set up for individual projects as needed. You can build your mixer as you work and you can also create templates that make complex setups much faster and easier. Although the specifi cs of each DAW will vary, the basics of DAW setup include creating the number of tracks needed for a recording ses- sion, naming the tracks, and assigning the appropriate inputs and outputs for 84 each track. Some basic level and panning settings, creating sends for headphone feeds, and some eff ects such as a reverb that might be used for monitoring can also be set in advance. Many fi les or one big fi le? When starting a project that involves many songs (a typical CD project, for ex- ample) you will need to decide how you are going to manage the song fi les. It may be tempting to record all the songs into one fi le, as that does not require using a template and setting up a new fi le each time a new song is going to be started. It makes things easier at the basic session to have all the songs in one fi le and it can make mixing easier as well, but it is generally only a good idea for projects that are going to be very limited in the amount and variety of recording to be done. I f the project is solo piano, or acoustic guitar and voice, then one big fi le will be easier to manage and will save time. Th e same is true for live recordings, even if there are many tracks involved, because there are usually no overdubs (or very few), and a very consistent sounding mix for all songs is oft en appropriate. Of course, there isn’t time to switch fi les during most live recordings, anyway. For projects where there are going to be a lot of overdubs and a fair vari- ety of instruments and/or arrangement elements (background singers on some, PHOTO 3.4 A studio-room wall- mounted patch panel, Recording Sessions horns on others, etc.), then it is best to create a new fi le for each song. Ulti- mately, this makes the recording and mixing process simpler and more focused because there are not a lot of extraneous elements that don’t relate to the song you are working on at any given moment. By using a template at the basic ses- sion, it doesn’t take much more time to set up a new fi le for each song;, and in the long run this makes for more effi cient work and better fi le management. R egardless of how you organize your fi les, it is a good idea to periodi- cally remove recorded and edited elements that you are not using. Th is includes multiple regions that may have been created in the editing process. Because 85 each DAW fi le needs to keep track of all the elements recorded into that fi le, too many recorded elements can slow or even stop the operation of a fi le. DAWs have diff erent ways of naming and identifying unused bits of recordings or ed- ited pieces, so you will have to explore your own DAW to fi nd the way to elimi- nate these elements; but it is important to do so, especially in large and complex projects. Simply remove these elements from the current fi le; don’t erase them from your hard drive (two diff erent choices in the “fi le management” function). Remember, if you maintain multiple fi les for each song or project, you can al- ways return to an earlier fi le to retrieve elements that you may have removed in a later fi le. I name my fi les using ascending numbers, creating a new fi le at least once each day that I work on a song. For example, a song titled “Swing the Hammer” will be saved as Swing the Hammer 2 the second day it’s worked on and saved as Swing the Hammer 3 the third day, and so on. WHAT NOT TO DO Don’t record more than one song into an individual DAW fi le if you expect there to be a lot of recording (multiple takes and/or multiple elements) for that song. Too many recorded elements in one fi le will slow down the DAW’s ability to function, and can even prevent it from functioning at all if the fi le gets too large. This is one of the most frequent causes of poor fi le performance and can often be fi xed by removing unused audio fi les and regions from your session. You don’t have to eliminate the audio from your hard drive to do this. Remember that keeping a separate session fi le for each day of work (or even more frequently, if a lot is done in a day) will allow you to recover previous material easily if needed. Managing multiple takes Th ere are two basic options for managing multiple takes of the same piece of music (e.g., diff erent takes of the same song). You can (1) place each take one aft er the other on the DAW’s timeline as you would on a tape recorder, or you can (2) use virtual tracks and place each take “on top” of the other so that only, THE ART OF DIGITAL AUDIO RECORDING one take at a time is visible in the DAW. Th e advantages to technique 1 are that you can see all of the takes at once and create markers for each individually. Th e advantage to take 2 is that your timeline is less cluttered and if you are working to a fi xed tempo or click track, you can line all your takes up and more easily edit between various takes. Many DAWs are developing new working protocols for handling multiple takes. Some are providing ways to manage virtual tracks so that they can all be seen at the same time and you can establish a hierarchy to automatically take 86 care of muting when moving from one to the other. I have seen the various techniques debated in user groups, and it’s clear that no one approach is best— use whatever approach seems most comfortable to you.

L ine testing, setting levels, and troubleshooting

O nce the setup for a session is complete, it is important to test your signal path, set rough levels, and, if necessary, troubleshoot before the recording begins. You can do most of this yourself, but if you have an assistant, it makes the process easier. It is always important to work as effi ciently as possible, but if you have to involve the musician in the testing process it is doubly important. Line testing Th e fi rst test is a line test, in which you verify that signal is passing as expected from the source to the recorder and then out to the monitoring system. Th is is easiest to do with an assistant lightly tapping each mic. If the mics are close enough, you can clap to see if they are active. You can also turn up the gain on the mic pre and see if you detect signal, but be careful as this can easily cause feedback. Asapart of the line-testing process, you also want to check to see if the headphones are working properly, both for talkback and playback. You might be able to hook headphones up to the cue system in the control room and check that way. If you already have something recorded, you can play that back and go out to the studio to see that the headphone playback is working and to check for WHAT NOT TO DO Do not ask a musician to put on headphones and proceed to playback audio for the individual without knowing that the level of that audio is not too loud. There are few things worse that blasting audio into a musician’s ears at a recording session. Not only is it unpleasant and unnerving, it can actually affect the person’s hearing for a period of time and make it more diffi cult for the musician to perform. Always check the headphone level before the initial playback for musicians., Recording Sessions level. You will want to check your talkback level, as well. You can do this if you have headphones in the control room or use an assistant or one of the musicians to check. It is always a good idea to start with the headphone level at a relatively low volume and to turn it up slowly to meet the musician’s requirements. Many studios now have headphone boxes with volume control so each musician can control his or her own volume. Small units are available for home and project studios, and this feature is highly recommended. Setting levels 87 Setting input levels for each instrument requires the participation of the musi- cian being recorded. Once you have confi rmed the signal path and the head- phone operation (ideally before the musician arrives), you can ask the musician to play for you. Besides determining the quality of the sound, dependant on mic selection and mic placement (as explored in section 3.3 on recording various instruments), you will need to set the input level. Proper level setting requires discovering something close to the loudest volume the musician will be playing so that you can get a good amount of signal for your recording without over- load and distortion. Th is can be a challenging process, but here are some rules of thumb. B egin by explaining to the musician that you need to hear the person’s loudest playing level in order to set a recording level. Ask the musician to play the loudest part from the piece that you’re going to be recording, as diff erent pieces will have diff erent dynamic ranges. It’s quite common for musicians to play their part louder when the recording actually begins, so always leave some headroom when initially setting levels. Some times musicians end up playing somewhat soft er than they did when they were testing, so level adjustments may be necessary in either direction. N onetheless, it is most desirable to not change levels once recording has begun—especially not during an actual recording pass. With the heavy reliance on editing in many contemporary recordings, a consistent level makes it much easier to piece together performances from many diff erent takes. Nonetheless, level does matter. Th ere is the obvious problem of distorted audio if the level is too loud. If the level is too low, there is some sacrifi cing of resolution, as fewer bits are available to describe the audio’s timbre. Th ere may well be a confl ict between the desire to record at the maximum level without overload and the advantages of not changing level once record- ing has begun. Keep in mind that even the fi rst run-through—sometimes the musicians aren’t even aware that you’re recording (you should a lways be record- ing)—may produce the best music of the day. Levels can be adjusted to com- pensate for level changes made during recording passes, but it can be diffi cult and time-consuming. Knowing when it is necessary to change your input level and when it’s best to leave it alone, even if it’s a little louder or quieter than opti- mal (without distortion of course), is part of the recordist’s skill set., THE ART OF DIGITAL AUDIO RECORDING T roubleshooting Of course, you hope that there won’t be any troubleshooting required at any session, but the reality is that with so many cables and knobs, and so many com- puter and soft ware issues, there are likely to be some problems at many record- ing sessions. Fast, effi cient troubleshooting is one of the primary ways topnotch recordists distinguish themselves from those with less experience. Th e key to effi cient troubleshooting is the ability to think logically through the signal fl ow to determine the most likely cause of the problem. 88 Th e most common problem is no signal and the cause can be anywhere in the signal chain. Some consoles show input level, and that means you can determine if signal is getting from the mic to the console. If there is input level, then the problem is somewhere between the console and the DAW; if there is no input level at the console, then the problem is before the console. Problems can be anywhere in the signal path—bad mics; bad cables; bad con- nection points in the wall panel, patch bay, or DAW interface; or they can be computer related, such as soft ware glitches that require program or computer restarts (or worse). Th ere are other typical problems, such as buzzes or hums. Th ese can have multiple possible causes, from electrical to electronic to cell phone interfer- ence. Th ere can be intermittent problems that can be almost impossible to track down until you can fi nd the cause and reproduce the problem without having to wait for it to occur on its own. Th ere can be dropouts. Th ere can be computer freezes. Th ere can be polarity problems from inconsistent wiring. Th e list is al- most endless. Some problems can be easily solved, and some cannot be solved without sending gear out for repair, requiring sessions to be canceled in the meantime (the most dreadful outcome, of course). Following the signal path and using logical procedures to determine the most likely reason for the prob- lem are the best companions to experience in troubleshooting. O ne of the most valuable ways of correcting problems is the w orkaround . Th at means fi nding a way to eliminate the problem without necessarily identi- fying what is causing the problem. If there is a complex patching situation like the one described in the previous section, and you fi nd that audio is not passing through to the DAW, you might start by plugging into a diff erent patch point at the wall panel, which is going to bring the audio in to a diff erent channel on the console. If that solves the problem, you don’t necessarily know if it was a problem with the wall patch point, the patch point into the console, the channel or buss in the console, or the patch point at the computer interface. You make a note to track down the problem later (there is a trouble report form at most commercial studios) and simply move on. Workarounds are oft en quicker than identifying the specifi c thing that is causing the problem, and speed is the num- ber one priority when it comes to troubleshooting—especially when people are waiting to start or continue recording., Recording Sessions 3.2 Headphone Mixes I have allocated a whole section of this chapter to headphone mixes (sometimes called cue or monitor mixes) because of how important they are to making successful recordings. However, before examining the process of making tradi- tional headphone mixes, I explore some important alternatives. WHAT NOT TO DO 89 Do not use headphones if the situation doesn’t demand them. Following are some circumstances where headphones are not needed. For most musicians, playing while monitoring through headphones is not as comfortable or familiar as playing without them. If you are recording a solo musician or an ensemble that plays together and balances their sound without the use of amplifi cation (a string quartet or an acoustic duo, for example), then do not use headphones. Or, if you are able to bring the musician into the control room and work with the monitor speakers, this is almost always preferable to using headphones. W orking with the monitor speakers rather than headphones is easy to do with synthesizers and other instruments that do not require microphones (such as a bass guitar recorded direct), but it is also usually pretty easy to do with am- plifi ed instruments such as electric guitars. Place the guitar amp or speakers in a separate room, run a guitar cable to the amp, and then mic the speaker. If the amp head is separate from the speakers, the head can be in the control room and then run a cable from the amp to the speakers in a separate room. Th e gui- tarist can monitor his or her sound along with the rest of the recording in the control room with you. I f you are “sharing” the monitoring (through the speakers in the control room) with a performing musician, the performer should dictate the mix. Be sure to keep checking with the musician to see if the individual is hearing as de- sired, in terms of both his or her own volume relative to any other instruments and the overall volume of playback. S ome people even like to do vocals—and other recording that requires a microphone—in the control room. Of course, feedback and leakage can create problems if you’re using speakers rather than headphones for monitoring, but there have been many great “live” recordings done with speaker monitors, so it certainly can be done. You can set up fl oor monitors in the recording room, as you would at a live gig, or you can use the control-room monitors. If you are monitoring in the control room, one trick is to put the control-room monitors, THE ART OF DIGITAL AUDIO RECORDING out of phase so that there is some phase cancellation when the sound reaches the microphone. Th is can help reduce leakage. U sing your control-room mix for the headphones An alternative to the traditional, independent headphone mix is using the same mix as you have for control-room monitoring for the headphones. Th e advan- tages of doing this are the ease of setup and the easy control of all the elements. Just as you are oft en making small adjustments in the control-room monitoring 90 as performance dynamics change and the focus shift s to diff erent elements, so might headphone mixes benefi t from continuous monitoring and subtle shift s in balance. Sharing mixes with the performer also means that you are continu- ously monitoring the headphone mix so you will be much quicker to correct imbalances, such as oft en happens when a new instrument enters (a solo, for example) that hadn’t necessarily been balanced in the initial headphone mix. I almost always use my control-room mix for the headphones when recording a single musician doing overdubs—especially vocalists. A variety of circumstances will prevent your using the control-room mix for the headphone mix. When the musicians need to hear something that you don’t want to hear (such as a click track), or when there is signifi cant leakage in the studio but not in the control room (such as live drums), you need to adjust the headphone mix to account for the room sound. Oft en, when musicians are in the studio with live drums, they will need almost no drums in their head- phones, as they get enough just from the sound in the rooms. Of course, you still need to hear the drums well in your monitor mix, as you aren’t hearing the live drums in the room. Th e other disadvantage to sharing mixes is that you can’t change your mix to hear something diff erently while the musician is re- cording. For example, if you decided you wanted to hear the background vocals as loud as the lead vocalist you are recording, to see how in tune they are, you wouldn’t be able to do that because it would disrupt the headphone mix for the performing vocalist. Nonetheless, in many instances, the advantage of sharing mixes outweighs the disadvantages. Creating separate headphone mixes Th e typical situation, especially in larger sessions, requires a separate head- phone mix for the musicians. Oft en two or more mixes are needed, especially if musicians are in diff erent rooms. A classic example is a band recording in the main room and a vocalist in an isolation booth. Th e band needs a separate mix to account for the live drums in their room, whereas the vocalist needs suffi cient drums and usually a lot more vocal level in order to sing. You might also be using a click track that needs to go to the drummer, but not to the other musicians. In that situation, three separate headphone mixes would be best: one for the drummer that includes the click; one for the other musicians in the, Recording Sessions main studio room that does not include the click or much drums, because they would be getting most of their drums from leakage; and one for the vocalist with a normal amount of drums and enough vocal to allow the vocalist to sing comfortably. Once you have decided whether to use a DAW or your console for your separate headphone mixes (discussed above, under “Setup”), there are certain technical details that are essential to all headphone mixes. First is that aux sends will be used to control the levels for the separate headphone mix, and second is that all the aux sends must be set to pre-fader. P re-fader means that the auxil- 91 iary send is tapping the signal before it gets to the main output fader (what you use for your control-room monitor mix) and is therefore independent of that fader. With pre-fader aux sends, you can create an independent headphone mix that doesn’t change when you change levels on the channel’s main output (post- fader sends follow the main channel output and are used primarily as eff ects sends in the send and return model, covered in section 5.2). B esides separate pre-fader sends for each of your headphone mixes, you need separate amplifi cation for each cue mix. Th ere are a variety of small head- phone amplifi er and mixer options that provide from one to six separate head- phone amps, and there are modular systems that allow you to add as many as you need (with some limits, of course). Some things to keep in mind are that diff erent kinds of headphones have diff erent power requirements and the number of headphones in use will also aff ect the ability of any given ampli- fi er to supply suffi cient level to all of the headphones. Some of the professional headphones come in diff erent ohm ratings, meaning they have diff erent power requirements. If you know that you’ll be driving a lot of headphones at the same time, you can chose the model with a higher ohm rating that requires less power to drive each pair of headphones. Consider the needs of your studio situation and research the amplifi er and headphone options that fi t your needs and your budget. Musicians make their own headphone mixes It has become increasingly common for studios to have systems that allow each musician to make his or her own headphone mix. Th ese systems consist of small mixers with headphone amplifi ers that can easily be stationed near the musi- cian. By feeding separate elements to the mixer (pre-fader, of course), the musi- cian can then adjust the level and panning of each element to meet his or her own needs. Th ere are several commercial systems available that provide varying features, including 4-, 8-, or 16-channel mixers. Some have built-in limiters to help guard against accidental overload. Depending on the number of channels, it is likely that you will need to make some submixes for the headphone mixer boxes. With an eight-channel system, for example, you might make a stereo submix of the drums and then, THE ART OF DIGITAL AUDIO RECORDING have individual channels for the bass, the guitar, the keyboard, the vocalist, and the click track. Th at would be a total of seven channels. As a result, you would still need to make adjustments depending on musicians’ needs (more kick drum, less hi-hat, for example), but the bulk of the headphone mixing can be done by each musician. Th e value of these personal monitoring systems is that they allow in- dividual musicians to craft their own headphone mixes in the way that suits them best, and it allows them to adjust overall volume, as well as individual 92 elements, instantly as needed. Th e disadvantage is that they do not allow the recordist to hear what the musicians are hearing, and as a result, they don’t necessarily get the benefi t from your experience. In the following section on the creative side of headphone mixes, I explore ways that headphone mixes might aff ect performances; what I have learned is that when the situation is appropriate for sharing mixes (control-room monitor mix and headphone mix), I do that, even at studios where there is the option for the musicians to control their own mixes. Th e creative side of headphone mixes Headphone mixes aff ect performance, and with experience you can help musi- cians create and alter their headphone mixes during the course of a recording in order to improve their performance. One example is working with a drum- mer who is playing to a click track. If the drummer is having diffi culty staying with the click, it may be that the click isn’t loud enough. Drummers without much experience playing to a click oft en don’t realize how loud it needs to be in order to maintain the groove to the click. On the other hand, if the drummer is staying with the click but having trouble making appropriate transitions—as in changing his or her part for the chorus—then the drummer probably doesn’t have enough guide vocal in his or her headphones so the musician is losing track of where he or she is in the song. W ith singers, getting the headphone mix right is an essential part in help- ing them sing in tune. If their own vocal level is too loud in the phones, they will not have enough pitch reference from other instruments; and if the voice is too quiet, they won’t be getting enough pitch reference back from their own voice. If extraneous elements are too loud (percussion or horn section, per- haps) and fundamental instruments are too low (bass and rhythm guitar or keyboard), then the singer will have trouble fi nding the pitch. It is oft en valu- able to keep working with vocalists on their headphone mix over the course of a session. An appropriate balance of elements in the headphones will aff ect the de- tails of a musician’s performance. A subtle shift in headphone balance can in- spire a musician who has been overplaying to lay back more, and it can encour- age a musician who has been struggling to fi nd a part to come up with just the, Recording Sessions right thing. It isn’t always possible to know what is going to work, which is why communication is such an important part of headphone mixes. Talking about headphone mixesAtevery session that involves multiple musicians, and each time I work with someone I haven’t worked with before, I have a discussion about headphone mixes before we start working. What I say is essentially this: “It’s hard enough to play music; it’s much too hard to do so when you’re not hearing well. So, please, let me know if you’re not happy with your headphone mix. I don’t care if you’ve 93 already complained ten times and you’re feeling like you’re bringing everybody down; you must have a good headphone mix and I want to work with you until you do. Th e worst thing for me at the end of a session is to have someone say, ‘I could have played better if I were hearing better.’ P lease , keep complaining about the headphone mix until it’s right!” Even aft er off ering this advice, it is still important to continue to ask the musicians if they are hearing okay. Over the course the session you want to ask “Are you hearing okay” every so oft en, just to remind the musicians to speak up about anything that might make them more comfortable with their headphone mix. WHAT NOT TO DO If you are using a click track or a loop for tempo control in a band recording, only allow the drummer to hear the click or loop. You want all the musicians to play to each other, especially to the drums—and not to the click. After all, the click will not appear on the fi nal recording, so the groove that matters is the drummer’s groove, even if it is being guided by a click track. Sometimes musicians need the click for a break when the drums don’t play. In that case, print a track that has the clicks in the break and feed that to everyone in their headphones. Musicians will often ask for the click if the drummer is getting one, thinking that it will help them with the groove. Try to talk them out of it, if you can. 3.3 Survey of Recording Techniques for Instruments and Voice Probably the most important elements in the fi nal sound of any instrumental recording are (1) the way the musician plays the instrument; and (2) the sound of the instrument itself, including how it’s set up and tuned. Th at said, the recordist’s job is to capture the sound in the best way possible for the intended purpose. Th is survey is not intended to be exhaustive—that would be impossible—but the fol- lowing represents many years of personal experience and research., THE ART OF DIGITAL AUDIO RECORDING

U sing EQ, compression, and limiting when recording

R ather than addressing the use of EQ and/or compression for each instrumen- tal recording technique, I am going to discuss this topic in more general terms. Th e problem with advising on EQ and compression usage is that it varies in every situation, depending on the sound of the instrument, the room, the mu- sical genre, and the ultimate instrument confi guration (e.g., solo instrument, small band, large band, etc.). Nonetheless, there are some general things I ad- 94 vise in regard to using EQ and compression when recording. Asageneral rule, elements that will ultimately need a considerable amount of EQ and compression in the fi nal mix should have some applied when record- ing; and elements that will need a small amount of EQ and compression, or none at all, should have none when recorded. For me, this translates roughly as follows: drums get EQ but no compression; bass, vocals, horns, electronic keyboards, and most acoustic instruments get compression but no EQ; electric guitars get no EQ or compression. Th ese rules of thumb can easily be over- turned by circumstances, but what really varies greatly is the amount of EQ or compression that might be applied. Th ere is no substitute for experience in this regard, but again, as little as seems obviously benefi cial is the best guideline. L imiting can also be used as a guard against overload when recording, and this may be especially valuable in live recording situations. With studio record- ing, it is usually possible to do enough testing to be confi dent that overload is unlikely, but in situations where unexpected changes in level seem likely, a limiter in the recording chain that is set to limit near the top of the acceptable record level can be a worthwhile addition.

D irect boxes, reamp boxes, etc.

D irect boxes (or DIs, for “direct inputs”) are an important part of many of the following descriptions of recording techniques. A direct box converts instru- ment level (and impedance) into microphone level (and impedance), and as a result provides a cleaner signal path that can be run for longer distances. Elec- tric guitars and basses, synthesizers, samplers, drum machines, and so on put out various amplitudes of line-level signal that can benefi t from a DI for record- ing. Most DIs provide two outputs so that the unprocessed signal can continue out of the DI to an amplifi er while the converted signal goes to the mic preamp for recording. Th ey also provide ground-lift capabilities that can help prevent hums and buzzes caused by improper AC grounding. Although direct boxes can be bypassed in many situations by plugging line-level sources directly into mixers or interfaces, DIs will generally provide better results. P assive DIs require no external power whereas active DIs need to be pow- ered, either by batteries or by phantom power from a console or mic preamp. Some active DIs require either batteries or phantom power and some are ca- pable of using either. Passive DIs are less expensive, but they may introduce, Recording Sessions some high-frequency signal loss. Many contemporary mic preamps include a DI function so that instruments can be plugged directly into the preamp for conversion to mic level. Some direct boxes and mic preamps are tube based, and these provide a diff erent tonality. A reamp box is a relatively new device that converts the output signal from your DAW back to a typical output level from a guitar or bass. Th is allows for easy reamping, which means putting the recorded signal back through an amplifi er, miking the amp’s speaker, and rerecording the sound. Th is can be convenient for unsatisfactory guitar sounds or for situations where the desired 95 guitar amp isn’t available for the initial recording. Reamping works best when the initial recorded signal is the direct signal from the guitar, so some record- ists record guitars both directly and through amps, just in case they decide they want to do some reamping later. By the e tc . in the title of this subsection I am referring to other boxes that can be valuable aids in recording, such as splitters and other level or impedance conversion boxes. Splitters that allow a guitar signal to be split out to two sepa- rate amplifi ers without losing gain can be a useful tool, as can other conversion boxes, such as those that convert -10 dBV output level (consumer gear) to +4 dBu input (professional gear).

Drum set

Drum sets can vary enormously in their specifi cs. Here, I cover the basic types and principles for recording drums. Th e section in chapter 6 on mixing drums might provide some ideas about how these recording tactics play into various mixing strategies for a fi nal drum sound. Recording a drum set can be one of the most challenging jobs for a re- cordist. On the other hand, I read an interview with Mick Jagger in the 1980s in which he was asked what had changed most about making recordings between the ’60s and the ’80s. He answered that it was recording the drums. In the ’60s, they used to spend an enormous number of hours—sometimes days—trying to get a decent drum sound, but by the ’80s it would take less than an hour. Ex- perimenting with drum-set recording techniques can be fun and can yield great results, but there is oft en a considerable limitation on time available, so the tried-and-true techniques that have been developed, and that prompted Jagger’s response, are good starting points. Th ere are numerous potential strategies for recording drums, but there is a basic technique that has become pretty well standardized. Th is involves using separate microphones for almost every element in the drum set, and frequently two mics on the critical bass drum (more frequently called the kick drum) and the snare drum. Mics are also used on each individual tom-tom (usually re- ferred to as either rack toms if mounted on the bass drum or fl oor toms). Th e hi-hat is miked and then a stereo pair of mics is used for “overheads” that cover, THE ART OF DIGITAL AUDIO RECORDING all the cymbals and provide something of an overall drum sound. Frequently, a separate mic is used on the ride cymbal and, if the room the drums are in has an appealing sound, then a stereo pair of room mics is also used. On a typi- cal drum set (confi gured with two rack and one fl oor toms), this could easily amount to 13 microphones as follows: ( 2) K ick drum: a mic inside the drum and one outside in front of the drum 96 (2) Snare drum: a mic above and a mic below the drum ( 1) Hi-hat ( 3) Tom-toms: one for each tom, two rack toms and one fl oor tom (2) Overhead: a stereo pair ( 1) Ride cymbal (2) Room: a stereo pair A more thorough explanation of the tactic for each drum follows: Bass drum (also called kick drum) D rums have two basic elements to their sound. Th e initial attack portion of the sound created when the stick (or bass drum beater, or hand, or whatever) strikes the drum and then the resonant vibrations of the drumhead and shell aft er the drum has been struck. While a mic can easily capture both of these ele- ments, with an instrument as important to popular music as the kick drum, we oft en use two mics, each one optimized to capture the two diff erent elements of the sound. On the bass drum, this means having either a hole in the front head of the drum, no front head at all, or an internal mounting that allows us to place a mic within a few inches of the front head (from behind the beater side, inside P HOTO 3.5 An AKG D112 microphone positioned inside a kick drum, Recording Sessions the drum) to best capture the sound of the attack when the drum is struck, and another mic—oft en a couple feet away from the front head—to capture the resonance of the drum. Th e close mic can be either directly across from the beater or set off -center for a slightly soft er attack sound. It can be anywhere from 2 to 10 inches inches away from the beater, the closer positioning providing a more pronounced at- tack. Experimenting with this mic position can be productive, though a simple standard (across from the beater and about 4 inches away—or some variation on this that you prefer) can provide excellent results very quickly. A dynamic 97 mic is almost always the best choice, and certain mics have become industry standards (Electrovoice RE-20, Sennheiser 421, AKG D-112), but there are many new mics coming onto the market all the time that have been specifi cally created for recording kick drums and these can also do a great job. Th e outside mic can be a large-diaphragm condenser mic as long as it isn’t too sensitive to loud sounds (most contemporary non-tube mics will hold up fi ne). Th e classic mic to use is the Neumann U-47 FET, but it is expensive (though it can also be used for many other things, including vocals). Because the mic is outside the drum, in front of the drum set, it is subject to picking up a lot of leakage from the other drums and cymbals. It is a good idea to isolate this mic by creating a tunnel that eff ectively extends the shell of the kick drum. Th is is most commonly done with mic stands and a bunch of packing blankets, but it can also be done with a rolled-up carpet. I like to refer to this structure as the “tunnel of love” because the tunnel creates such a lovely kick-drum resonance. A s an alternative or in addition to the mic outside the drum, you can use the speaker “trick” for capturing the very low end of the kick drum. Th is in- volves placing a speaker very close to the bass drum so that the speaker cone is vibrated when the drum is struck and then wiring the speaker with an XLR connection and taking its output as though it were a microphone. Remember, a microphone and a speaker are at the two ends of the same process—one captur- DIAGRAM 3.5 Kick-drum miking with a “tunnel of love”, THE ART OF DIGITAL AUDIO RECORDING ing and one reproducing sound—and as a result they use very similar technolo- gies (a vibrating membrane). A Yamaha NS10 speaker, at one time the standard for small monitor speakers in studios, is oft en used as a “faux” microphone to capture the low end of the kick. S nare drum Th e snare drum is oft en the most prominent drum in a fi nal mix, and it is fre- quently one of the loudest overall elements as well. Although many other mics 98 and techniques have been tried and are used sometimes, the standard is a Shure SM57 placed a few inches in from the rim and a few inches above the drum. Th e mic can be placed at varying degrees off -axis, and this will aff ect the sound slightly. Many people also use a second mic underneath the snare drum, point- ing up at the snares. A small-diaphragm condenser is a good choice for this mic, but many diff erent mics, including a second SM 57, will work fi ne. Th is mic allows you to add more of the rattling “snare” sound if you want it. Because it is facing above the snare mic (180 degrees out of phase), it needs to have its polarity switched for the two mics to be in phase. Although I sometimes record a second mic under the snare, I fi nd that I rarely end up using it in the mix. If miking a snare is so easy, why is it so diffi cult to get a great snare sound? Th e key to the snare drum is in the way it is hit and the way it is tuned. Th ese two elements can vary so greatly as to completely alter the sound of the drum, no mat- ter how it is miked. Th e drum itself is important as well, but unless it’s really poor quality or in really bad shape, almost any snare drum can sound great if it is struck and tuned well. How the drum is struck really changes the sound (normal hit, rim shots, hit in the center, etc.—consult a drummer!). Snare drums are also compli- cated to tune because of the interaction between the two heads and the snares. Damping is oft en used on the top head, and this can alter the sound dramatically. PHOTO 3.6 Miking the top of a snare drum (Shure SM57), Recording Sessions Anything from small bits of duct or gaff er’s tape, a small square of folded paper towel, a wallet, or some “moon gel” (a gel-like substance that sticks to the drum and is sold at music stores) can be used for dampening the snare drum. If you are not happy with the snare drum sound, there is a good chance that the mic or mic placement is not at fault; it’s more likely to be a combination of how the drum is being hit, how the heads are tuned, and what, if any, dampening is used. Using EQ and compression on the snare drum in the fi nal mix (as discussed in section 6.2) will also play a signifi cant role in the fi nal sound of the drum. WHAT NOT TO DO Don’t assume anything about the sound of a snare drum. Shallow snare drums don’t necessarily sound higher in pitch and metal snare drums are not necessarily brighter than wooden ones. It’s a good idea to have several snare drums available at a session and to audition each—but remember, tuning and dampening can really change the sound of the drum. H i-hat R ecording the hi-hat is generally accomplished best with a small-diaphragm condenser mic placed a few inches in from the outer edge and above the hi-hat cymbals. You will want to check how far up the top cymbal is when the hi-hat is open and how loose the two cymbals are to make sure that the mic has a few inches of clearance. You can aim the mic slightly away from the drum set to minimize leakage from the other drums. P HOTO 3.7 Miking a hi-hat (AKG 452 EB), THE ART OF DIGITAL AUDIO RECORDING Tom-toms T om-toms are recorded much like the snare drum, generally using dynamic mics with similar placement above and in a bit from the rim. Th e Sennheiser 421 has become the default microphone for the tom-toms and it does a great job, but there are many equally good alternatives on the market these days. Some people prefer condenser mics on the toms, and it is a diff erent sound— more detail but less of the woody warmth. If you use a condenser, make sure it can withstand the level or be sure that the drummer isn’t going to hit the toms 100 very hard. Positioning the tom mics can be a challenge, depending on where the drummer’s cymbals are placed. It’s important that you fi nd a spot that doesn’t interfere with the drummer’s stick movement and won’t get hit by a swinging cymbal. As discussed in section 8.1, do not ask the drummer to move any of his or her drums or cymbals to accommodate the mike positioning! A lthough I list drums as one of the few elements that I tend to EQ while recording, the tom-toms are the one part of the drum set that I oft en do not EQ until the mix stage. I’ve found that when tuned and recorded properly, tom- toms require very little EQ, so it is best to reserve it for the mix stage. PHOTO 3.8 M iking tom-toms (Sennheiser 421s) Overheads Th ere are many ways to approach overhead miking. Th e size of the drum set, the sound of the room, whether or not you are using room mics, and the kind of drum sound you prefer will all be factors in choosing a strategy for your over- head mics. In most cases, my preferred overhead setup is with small-diaphragm, Recording Sessions P HOTO 3.9 Overhead drum miking using the ORTF stereo confi guration—insert shows mics confi guration in close-up (Neumann KM-84s) condenser mics centered two to three feet over the drums in an ORTF stereo confi guration. For larger drum sets you may want to use a spaced-pair con- fi guration to capture all the elements of the set more evenly. Th is may produce slightly more phase problems, but it will give a good stereo spread. You can also use a classic X/Y stereo pair for a tighter, virtually phase-free sound but with a narrower stereo fi eld. If you have room mics, then the overhead mics may be closer to the drums, really focusing the attention on capturing the cymbals. If you don’t have room mics and like the sound of your room, you may want to consider using large-diaphragm condenser mics for overheads and putting them another foot or so higher above the drums. In this way, you capture some of the room acous- tics along with the overall drum sound. If the room sound is problematic, then you’ll want to keep the overheads pretty close to the drums. Ride cymbal E ven with overhead and room mics, it is a good idea to put a separate mic on the ride cymbal. Generally, a small-diaphragm condenser mic positioned a few inches above the ride cymbal is best. Although the overheads will pick up a lot of ride cymbal, there may be times in mixing that you want more ride cymbal relative to the crash cymbals, and a separate ride track allows you to balance the two. It used to be that the limited number of tracks made it hard to accom- modate a separate track for the ride cymbal when it isn’t always needed, but the expanded track count of most DAWs has eliminated that problem. Th e question may be whether you have enough mics and mic inputs, and if you do, I recom- mend a separate ride cymbal track., THE ART OF DIGITAL AUDIO RECORDING PHOTO 3.10 Miking a ride cymbal (AKG 452 EB) R oom mics Room mics can be a wonderful addition to an overall drum sound if the drums are in a nice-sounding room. Placement may vary depending on the room and the amount of ambience desired. A pair of large-diaphragm condenser mics works well, and typical placement is eight to ten feet from the drum set, point- ing down from above. Some recordists swear by a mic placed a few feet above the drummer’s head pointing at the set. PHOTO 3.11 S tereo room mics for drum recording (Neumann U-87s)


Th ere are literally thousands of percussion instruments (including the drum set), so it is impossible to cover them all. Instead, I have divided percussion into three basic “families” and will cover the general principles for each., Recording Sessions WHAT NOT TO DO Don’t skimp on drum mics. There seem to be endless stories about how great drum sounds have been captured using minimal miking setups. The stories are no doubt true, and you may indeed want to use the three-mic drum sound (kick and two overheads) or the fi ve-mic drum sound (kick, snare, two overheads, and room) or whatever. But none of these tactics precludes having many more mics to choose from. I recently made a record, and during the tracking the 103 artist (who favored the sound of blues records from the ’30s and ’40s) said, “That’s too many drum mics.” I said, “We don’t have to use them all.” And indeed, in the fi nal mix we often only used a few of the drum mics to get the best sound for the record. However, at one point during one mix, the artist said “Can I get more hi-hat?” and at that point we were both glad that I had used a mic on the hi-hat. D rum percussion H ere, I include congas, bon- gos, timbales, djembe, taiko, and other drum-based per- cussion instruments. Also in this family are the drums in the drum set, and a similar strategy for recording may be employed for all of these instruments. Th is means that a dynamic mic, placed a few inches in from the rim and above the drum, is a good starting point. When placing mics, the recordist needs to be sensitive to how the drum PHOTO 3.12 is played so as to not interfere M iking conga drums (Sennheiser 421s) with the musician’s technique. H igh-pitched percussion H ere, I include cymbals, tambourine, chimes, triangle, bells, and other percus- sion that produces primarily high-frequency sounds. Because of the fast-mov- ing transient frequencies of these instruments, their sound is very eff ectively captured by large-diaphragm condenser mics. Th e mic should be reasonably close to the instrument, but take care not to interfere with the playing., THE ART OF DIGITAL AUDIO RECORDING P HOTO 3.13 Miking a tambourine (AKG C-414 ULS) Clacking-type percussion H ere, I include cowbell, woodblock, castanets, guiro, and other percussion that is struck and that produces sharp clacking or scraping sounds. Th ese thick, strong sounds are generally best captured by a dynamic mic. You will notice slight variations in tonality depending on which part of the instrument the mic is facing.

B ass

Th e low frequencies of bass instru- ments are easily compromised by amplifi ers and room acoustics, so take care if you wish to capture the purest sound possible (certain genres may encourage all kinds of experimentation that does not value a “clean” bass sound). For obvious reasons, electric and acoustic basses require diff erent recording tactics. E lectric bass P HOTO 3.14 Electric bass is oft en recorded very Miking a cowbell (Sennheiser 421) simply, using a direct box (DI) to transform the output from the in- strument into a mic level output that is fed into a mic preamp and then di- rectly to the DAW (or through a mixer and then to the DAW). Th e advantage of direct recording of bass guitar is that it bypasses the various problems that amplifi ers and speakers can cause (low-level distortion and unwanted eff ects, Recording Sessions caused by room acoustics). Th ere are a variety of direct boxes avail- able and they will aff ect the sound of the bass, as well. Some record- ists swear by the sound of tube DIs for bass. I oft en include some com- pression in the input chain. Th e Empirical Labs Distressor and the Urei 1176 are frequent choices for 105 bass compression, though many compressors—including plug-in versions—will compress the bass without noticeable artifacts. Some bassists like to record the amplifi ed sound along with the direct sound and combine the two P HOTO 3.15 when mixing. If doing so, I like to Miking a bass guitar speaker (AKG D112) place the mic about 12 inches away from the bass speaker to allow for some greater contrast to the direct sound. Sometimes it is best to just take the direct out from a bass amplifi er without using the speaker. Th is allows you to record the eff ect of the amplifi er’s preamplifi cation, as well as any onboard ef- fects that you might want from the amplifi er without the additional diff usion created by speaker reproduction. Although I usually simply take the bass DI, I am happy to record the preamp or mic the speaker as well if the bassist feels that it is an important part of his or her sound. Acoustic bass A coustic bass can be a challenge to record eff ectively, especially if it isn’t isolated from the drums or other sounds. A small-diaphragm condenser mic oft en works best for acoustic bass. It should be placed about 12 inches from the front of the instrument facing one of the f-holes—this usually allows the mu- sician suffi cient freedom of move- ment when playing. Most bassists P HOTO 3.16 have a pickup that they use to am- Miking an acoustic bass (Neumann KM-84) plify the bass, and this can be ex-, THE ART OF DIGITAL AUDIO RECORDING ceedingly valuable when there are leakage issues, but it never has as good a sound as the instrument when properly miked. Unfortunately, where there is a lot of leakage, you sometimes have to use the pickup sound primarily. I’ve found the acoustic bass pickups are sometimes wired in reverse from the typical mic cable, and so the pickup signal and the mic signal are 180 de- grees out of phase. Th is is easily fi xed by switching the polarity of either one of the signals. It’s a good practice to check the phase and polarity any time you are getting two distinct signals from the same sound source .


R ecording guitar has become an elaborately studied art, as guitar has occupied such a central role in so much popular music. Th e most widely practiced basics are covered here. Electric guitar Th e sound of the electric guitar is intimately tied to the sound of the amplifi er and speakers used to reproduce the sound before it is recorded. Th e elements in the chain, from the guitar itself to any stomp boxes in the chain, to every setting on the amp, to the type and size of the speaker used, to the mic, mic placement and mic preamp, all combine to create the fi nal sound of the electric guitar when recorded. All kinds of mics, combinations of mics, and mic placement strategies have been used to record electric guitars. Th e classic approach—a Shure SM57 placed halfway between the center of the speaker and its edge, slightly off -axis (the plane of the mic’s diaphragm at a slight angle to the plane of the speaker cone), and an inch or two from the speaker grill cloth—still produces excellent results and is sometimes the best ap- proach to capture the desired sound. O ther frequently employed strategies include using a “far” mic (or a stereo pair of “far” mics) in con- junction with the close mic, placed anywhere from 2 to 20 feet away from the amplifi er (even mics just 2 feet away from the amp will produce a much diff erent sound than mics 2  inches away). Far mics are oft en PHOTO 3.17 either small- or large-diaphragm con- Miking an electric guitar densers. Other dynamic mic models speaker (Shure SM57) are sometimes used, and ribbon mics, Recording Sessions have become very popular as close or far mics, either in combination with a close dynamic mic or as a replacement. O n-axis placement, varying degrees of off -axis placement, angling toward or away from the center of the speaker, up against the speaker grill cloth or any- where between 1 and 3 inches away, closer to the center or closer to the edge of the speaker—all of these represent variations on close-mic strategies for captur- ing the sound from guitar amplifi er speakers, and each will make a small but audible diff erence. When there is time, it can be valuable to explore any or all of these variations and/or additional miking options, but sometimes it is neces- 107 sary to simply “throw a 57 up to the speaker and go!” A s mentioned above, recording the direct sound from the guitar has be- come more popular so as to allow for either reamping or using one of the many amp modeling plug-ins now available for DAWs. Acoustic guitar Recording acoustic guitar has also been explored extensively, and there are many possible tactics. However, the one that many of the most experienced recordists have settled on involves using two small-diaphragm condenser mics. Th e primary mic is placed across from the 12t h fret (one octave) on the guitar and aimed toward the sound hole. Th is placement coincides with the most reso- nant spot on the neck. Th e second mic is aimed from the other side of the guitar and can be positioned the same distance as the fi rst mic or a bit farther away if you want to capture more of the guitar’s fullness. Many alternative approaches may also produce great results with acoustic guitar, including using large-diaphragm condenser mics, ribbon mics, alterna- tive mic placement, and so on. However, if you are using only one mic, I recom- mend the positioning across from the 12 th fret as the starting point. PHOTO 3.18 Two-mic technique for recording acoustic guitar (Neumann KM-84s), THE ART OF DIGITAL AUDIO RECORDING


Recording vocals is one of the most complex of studio activities, and there is information at other points in this text regarding headphone mixes and talk- back techniques that are essential parts of the vocal recording process (sections 3.1 and 8.2). From the technical standpoint, it can be pretty straightforward: a large-diaphragm condenser mic set to the cardioid pattern, with a pop fi lter in front of it and the mic about 8 to 10 inches from the vocalist is the stan- 108 dard. However, within that context there are many subtle variations. Th e type of mic, the distance from the mic, and the exact placement relative to the singer’s mouth are all elements that can be adjusted depending on the musical genre, the volume of the singer’s voice, and the style of his or her delivery. A lthough large-diaphragm condensers are generally the fi rst choice, there are other mics, especially dynamics such as the classic Shure SM57 or the Shure SM7, that may be right for your particular singer. Within the ranks of large- diaphragm condenser mics, there is a broad choice, including tube-based mics. It is likely that a good-quality large-diaphragm condenser will sound good for pretty much every vocalist, but when you are getting down to the subtleties, cer- tain mics will sound better for certain singers, and it can be diffi cult to predict. One can go with a warmer mic on a male vocal to capture the generally lower tonalities or a brighter mic to provide more clarity. You can choose a warmer mic to soft en a female vocal or a brighter mic to accentuate the presence. If you have more than one mic for vocals, and if you have the time (both pretty big “ifs” in many cases), it can be valuable to test to see which one is most appeal- ing. It has been very interesting for me to discover that, when there has been an opportunity to compare vocal mics, there has almost always been an immediate agreement among all involved as to which mic sounds best. PHOTO 3.19 Miking a vocalist— Michael Moorhead (Neumann M49), Recording Sessions Large-diaphragm condensers are very sensitive and can be overloaded by a loud vocalist who is too close to the mic. However, a more intimate and de- tailed sound can be captured when the vocalist is very close to the mic (a couple of inches away) as long as the mic doesn’t overload. All directional (cardioid) mics exhibit the proximity eff ect (a bass boost when a singer get very close to the mic), but the large-diaphragm condensers are smoother and richer in prox- imity so it can be desirable. Ideally, the vocalist works the mic—coming in for quiet passages and leaning back when belting—but even if the vocalist doesn’t, you might play with his or her distance from the mic to get the most detail with- 109 out overload or unwanted proximity eff ect. F inally, I off er a note about the mic position relative to the mouth. I prefer to have the mic very slightly above the singer’s lips, so as to encourage the vocalist to tilt the head just slightly up and thereby keep his or her throat open. Some singers are not comfortable with this and prefer the mic directly across from the mouth, and some singers prefer to tilt slightly downward. As in all things, the desire of the musician comes fi rst unless you’re convinced that it is truly detrimental to the performance and then you can discuss it. Some recordists like to place the mic well above the singer’s head, angled down at the mouth. Again, experimentation is helpful in determining the best approach for any individual singer.

Piano and keyboard percussion

Th e piano is oft en considered a percussion instrument because of the hammer actions in striking the strings. Other keyboard-based instruments, such as the vibraphone and marimba, are also considered part of the percussion family. Th e key to recording these instruments is to achieve a good balance between the percussive attack and the resonant sustain. Because these instruments are rather large, and their sounds cover the entire frequency spectrum from low to high, they are usually recorded in stereo (at least two microphones). Grand piano Th e piano is a wonderfully complex instrument, with very rich sonorities cov- ering a huge spectrum of fundamental frequencies and overtones. It is also used in a wide variety of setting so there are many strategies for recording the grand or baby grand piano. Although you can record the piano with one mic, it is typi- cal to use at least two mics to capture a stereo image of the piano. A stereo pair of small-diaphragm condenser mics is most commonly set in one of the standard stereo confi gurations, such as ORTF or a coincident pair. Th e mics are usually placed 6 to 8 inches above the strings and can be put parallel to the hammers or the bridge. I prefer the ORTF confi guration and an over-the-bridge placement. I also use a third “centering” mic to capture a little more ambience and to fi ll the “hole” that can be created with a stereo pair. I typically use a large-diaphragm condenser, placed above the lip of the piano casing and aimed to capture the, THE ART OF DIGITAL AUDIO RECORDING PHOTO 3.20 Three-mic technique for recording a grand piano (Schoeps CM-5s and Neumann U-87) e ntire piano. Th e centering mic is used to stabilize the stereo image and to bal- ance the percussive sound of the stereo pair with additional ambience. O ther strategies for recording piano vary from the best isolation techniques to the most elaborate miking plans. Th e best strategy for isolation—when the piano has to be recorded in the same room as the drums, for example—is to use a stereo pair up over the strings, with the piano lid in its lowest position (using the short stick to hold the lid up). Th is takes some careful placement in order to get the mics as far from the piano strings as possible while still being able to lower the lid (short pencil condensers such as Neumann’s KM-184s are help- ful for doing this). Once the mics are set and the lid lowered, you completely cover all of the openings around the lid, using as many as 20 packing blankets (or other blankets—though packing blankets are relatively cheap and a great asset in a variety of studio setup applications). Th is does deaden the sound of the piano a bit, but a remarkably good recording is still possible while achieving enough isolation to be able to adjust piano tone and level independently. F or solo or small ensemble recordings where the piano is central, and where there is isolation from other instruments, it is best to remove the piano lid altogether (the hinges have removable pins to make this a relatively simple task). You can start with the same three-mic setup described above and add mics as desired. I have used as many as nine mics on a grand piano by adding a stereo pair 3 feet above the piano and another stereo pair 3 feet or so above that—both in the coincident pair confi guration to minimize phase issues. Addi- tional mics can be place at the foot of the piano, facing the player, and above the player’s head, facing the piano (because these two mics are facing each other, whichever is most out of phase with the other mics will need to have its polarity, Recording Sessions switched). You can use either large- or small-diaphragm condensers for these mics, though the stereo pairs should be matched models. When mixing, you may not use all of these mics, and some of them may be used in very small amounts, but tremendously rich recordings are possible by using multiple mic confi gurations such as this. Upright piano A similar three-microphone technique as described above for grand pianos can be used for upright pianos. It is necessary to remove the covers over the 111 strings and sounding board, both above and below the piano keyboard, and to open the top of the upright. A stereo pair of small-diaphragm condenser mics is placed near the bottom of the keyboard box, facing the strings in an ORTF or coincident pair confi guration. A third, centering mic, is place near the top of the piano facing down toward the strings and sounding board. Many other strategies can be used for recording an upright piano, includ- ing miking from the back and miking only from the top or bottom. Problems with leakage and access will aff ect the technique used to make the best record- ing, under the circumstances. PHOTO 3.21 Three-mic technique for recording an upright piano (Shure SM81s and Neumann U-87) O ther keyboard percussion Th ere are a multitude of instruments that are laid out like a keyboard and struck with mallets, but the two most common are vibraphone (vibes) and marimba. Th ese larger instruments can be miked very similarly to a grand piano, using the three-mic technique. Placing the mics over the instrument and leaving enough room for the musician to play comfortably requires careful placement and consultation with the musician. Smaller instruments, such as glockenspiel, THE ART OF DIGITAL AUDIO RECORDING or orchestra bells, can be miked with a single mic eff ectively, or with a stereo pair, probably best in the coincident-pair confi guration.

Brass, reeds, and horn

sections 112 H orns of all types can be recorded using a variety of techniques and a wide variety of microphone types, depending on the desired sound. Of all the instrument groups, they P HOTO 3.22 probably receive the broadest treat- T hree-mic technique for ment—dynamics, ribbons, small- recording marimba—Beth Wiesendanger (Shure and large-diaphragm condensers all SM81s and Neumann have valuable roles in possible mik- U-87) ing strategies for horns. B rass Th e brass instruments include trumpet, fl ugelhorn, trombone, and tuba, as well as many instruments that are less common in popular-music settings, such as the bugle, French horn, and sousaphone. Although these instruments cover a wide spectrum of frequencies, ranging from the trumpet to the tuba, a basic miking technique will serve well for all brass: the mic is placed opposite the bell (the large opening at the end of the instrument). Th e mic can be placed closer or farther from the horn, depending on its ability to withstand high SPLs (horns can be loud!) and the degree of detail you wish to capture. Keep in mind that some of the “detail” of horn playing includes the sound of saliva in the instru- ment, so too close a placement may capture more undesirable elements, but too much distance may lose too much detail. I fi nd 10 to 14 inches a good rule for the higher pitched brass (trumpet and fl ugelhorn) and 18 to 24 inches good for lower pitched instruments (trombone and tuba). Th e mics can be placed on- axis (pointed straight at the bell) for a brighter, clearer sound or off -axis for a soft er, more diff used sound. S electing the type of microphone provides a variety of sonic options. Con- denser mics, both small and large diaphragm, capture the truest sound of the instrument (and are generally preferred by the player), but they can yield an overly bright sound in an ensemble. Dynamic mics off er a warmer, rounder sound that may blend better with other horns and instruments. Dynamics also have an easier time handling the high levels that brass can put out, though many condensers (especially if they have a pad) can withstand the levels as well. Rib- bon mics have also become popular for recording brass, especially with many, Recording Sessions P HOTO 3.23a and 3.23b. M iking a trumpet on-axis and off-axis—Brandon Takahashi (Shure SM81) of the newer models able to withstand much greater levels than earlier versions. Ribbons provide a clearer high end than do dynamics (closer, though not equal to a nice condenser) and still provide the warmth typical of dynamics. Make sure the ribbon mic you are using can withstand the SPLs. M y preference for brass is generally the ribbon mics, though I don’t always have one available (or one that is capable of handling the level). I will generally go for a small-diaphragm condenser if there isn’t a ribbon option, and place it just slightly off -axis on trumpet and on-axis for most any other brass. If the, THE ART OF DIGITAL AUDIO RECORDING horn is being used as a solo instrument, I will usually go for the condenser. For horn sections I will sometimes use dynamics to get a better blend. Ultimately, the nature of the player, the instrument, and the way the horn is used in the ensemble will all play a role in determining the best choice. W oodwinds H orns classifi ed as woodwinds include the saxophone family (baritone, tenor, alto, and soprano) and clarinet, as well as the fl ute and the double reeds, such as 114 the oboe and bassoon. Although classifi ed as woodwinds, many of these instru- ments (saxes and fl utes, most notably) are made from metal. Recording strategies will vary from instrument to instrument. Woodwind recording is not as straight- forward as brass because the sound isn’t necessarily coming primarily from the bell of the instrument (the fl ute doesn’t have a pronounced bell). As with brass, a wide variety of microphones may be appropriate, depending on the goal. Th e instruments also vary greatly in frequency range, from the lows of the baritone sax to the highs of the piccolo fl ute, and this will aff ect your recording strategy. Saxes do have a bell, and a strategy similar to that described above for brass is oft en a good tactic. A mic 10 to 24 inches from the bell (on the closer side of things for the alto and farther away for the tenor and baritone) captures most of the sound. Dynamics, ribbons, and small- and large-diaphragm con- densers can all produce excellent results, with the dynamics and ribbons being warmer (or duller, depending on your point of view) with less high-frequency detail and the condensers being clearer and brighter, but with the potential to reveal too much of the harshness of the instrument. Generally on-axis position- ing (directly facing the bell) will be best, but an angled, off -axis approach can be tried if you feel the need to soft en the sound a bit. PHOTO 3.24a & 3.24b Miking a saxophone, one- and two-mic techniques— Joe Del Chiaro (Neumann U-87s), Recording Sessions Soprano saxes usually benefi t from a diff erent approach. With all of the saxes, a good deal of the sound emanates from the sound holes, where the keys are used to change pitches by closing certain of the openings. Because soprano sax is so bright sounding and has a relatively small bell, the more appealing sound tends to come from the sound holes rather than out of the bell. For this reason, I oft en simply mic from the side of the instrument, primarily captur- ing the sound that comes out of the sound holes. You can also use this aspect of the sax in a two-mic technique, capturing both the sound out of the bell and the sound from the sound holes. A tenor sax that is used as a solo instrument, 115 especially in a small ensemble, can benefi t from this recording tactic. C larinet is similar to a soprano sax and is usually captured best with a mic at the side. Th e lower notes come primarily from the keys, but the higher notes and high overtones come increasingly out of the bell. Placing the mic to the side, but down closer to the bell, can allow for a good balance through the frequency range. Because the sound emanates from diff erent places in diff erent frequencies, it is best if you can get some distance on the mic—preferably at least a foot and up to 3 or even 4 feet might yield the best results, depending on the room and the desired eff ect. Again, dynamic, ribbon, and condenser mics can all yield excellent, though tonally pretty diff erent results. Ensemble playing oft en benefi ts from the warmer mics and soloing will benefi t from the greater detail provided by the condensers. Th e fl ute is generally captured from the side of the instrument. Most of the sound of the fl ute comes from the mouthpiece, so if you are using one mic, it should be across from the mouthpiece. In order to capture the instrument more evenly, and because the fl ute is so bright and benefi ts from more interac- tion with room acoustics to soft en the sound, it is usually recorded from a dis- tance of at least a foot away and usually more successfully from a couple of feet away. Interaction with other instruments playing in the same room may dictate a closer mic positioning. Th e less common alto, bass, and baritone fl utes can be captured in the same way. D ouble reeds, such as oboe and bassoon, are fairly rare in popular music, but you still need to be prepared if one happens to show up at a session. Th ese woodwinds are related to the fl ute and the clarinet, in that the sound comes from diff erent places depending on the frequency, so getting some distance on the mic is defi nitely recommended. Th e double reeds also produce a lot of hi- mid transients, sometimes heard as a nasal quality, so room ambience helps to soft en the sound in a pleasing way. Again, all mic types can produce excellent results. Horn sections From 2- to 20- piece horn sections can be recorded with individual mics on each horn player, with mics capturing the section together (stereo pair(s), Decca Trees, etc.), with mics covering each section (brass, reeds, etc.), or some, THE ART OF DIGITAL AUDIO RECORDING combination of the above. Th e horn players in a small section (two to six play- ers) are usually miked individually. Mic placement may be a bit closer than with individual horn recordings in order to prevent too much bleed from the adjacent horn players, and you will probably need mics that can withstand the level that horns can produce when played aggressively. I fi nd that dynamic mics oft en work well for section recordings because they tend to help the horns blend and to occupy less frequency space when mixed with other instruments. Unfortunately, I have found that experienced horn players are sometimes un- 116 happy when they see a mic they associate with live gigs being set up in a studio situation. Th ey know that condensers record a more “true” and detailed sound, and they prefer them, even for section work. I might suggest to them the logic behind using dynamics, but I will always go along with the players’ wishes if I can tell they are unhappy about using a dynamic mic. Happy musicians trump subtle recording preferences every time! Strings and string sections By string instruments here I am referring to those instruments that are primar- ily bowed, including double bass, cello, viola, and violin, as well as numerous less common stringed instruments from other cultures, such as the Chinese erhu. Th e double, or acoustic, bass has already been covered in the popular music context, where it is much more frequently plucked (pizzicato) than bowed (arco). One important guideline for successful recording of bowed instruments is to make sure that the mic has suffi cient distance from the sound source. A bow on strings produces strong transients that can be very harsh sounding if not allowed to soft en. Th e mic should be at least 18 inches from the in- strument, and usually a distance of 3 to 4 feet produces the best results. Small- diaphragm condensers are preferred in ensemble situations because of the e xcellent off -axis response, but large-diaphragm condensers will yield out- standing results in solo recording situations and some recordists prefer the warmer sound of the ribbon mics. Th e mic is usually placed in front of and above the instrument. String quartets and other string ensembles are usually best captured with a stereo pair or some form of Decca Tree mic confi guration (as described in section 2.3). Because the mics need to have some distance anyway, it usually doesn’t make sense to try to mic each instrument separately. Exact placement in terms of distance from the ensemble and height off the fl oor will vary with the size of the ensemble and the room acoustics. Th e musicians will balance them- selves, so the mics should be placed in such a way as to best capture a balanced version of the entire ensemble. However, you may consider using a spot mic on the cello (probably a small-diaphragm condenser or a ribbon)—18 to 24 inches away. If the room has decent acoustics, you may not need to use any of the extra, Recording Sessions P HOTO 3.25 Miking a violin—Reiko Kubota (Neumann KM- 184) cello mic, but sometimes the room mics will be a little light in the low end. If the room is of a decent size, then a Decca Tree with at least two stereo pairs will allow you to adjust room ambience by balancing the closer and farther of the stereo pairs. WHAT NOT TO DO Do not close-mic bowed string instruments. A colleague tells the story of recording a violin for the soundtrack to a horror movie. He wanted a very harsh, frightening tone, so he put the mic a few inches from the violinist’s instrument. After the recording, he invited the musician to hear the playback. When she heard the sound of her violin, she cried! Unless you’re after a very special (and particularly annoying) effect, do not close-mic bowed string instruments. 3.4 Beyond Of course, there are many other instruments not mentioned, but the above should provide enough guidelines to get you started with almost any rec ording. Exotic instruments usually fall into one or another of the categories covered;, THE ART OF DIGITAL AUDIO RECORDING pay attention to how the sound is created (struck, picked, plucked, bowed, blown, etc.) and where the sound is coming from (it may be more than one spot), and then mic according to a similar and more familiar instrument. When it comes to electronic instruments—synthesizers, samplers, and so on—see the section on direct boxes and use those guidelines for direct-input recording.,

C hapter 4 Editing 119 Th e New Frontier

I am calling editing “the new frontier” because of the tremendously expanded editing capabilities in the DAW. Not only is editing a much more important part of almost all recording projects than it was in the past, a lot of today’s music is primarily c reated through editing within the DAW. Although all the major DAWs contain similar editing capabilities—and I use screenshot examples from several of them—the terminology for some specifi c editing functions in some DAWs does vary. For the sake of simplicity, where there are diff erences in ter- minology, I use the Pro Tools terms for these functions. It should be reasonably simple to determine which tools provide the same function in other DAWs. 4 .1 Editing Features Th e operating tools of editing begin with basic functions—such as cut, copy, and paste—that are familiar to anyone who uses computer programs. Func- tions that are somewhat more specifi c to audio, but still easily understood, in- clude duplicate, repeat, loop, clear, and mute. Th e expanded editing capabilities within a DAW really take advantage of the computer environment. Becoming a capable audio editor who can work quickly requires a lot of experience mak- ing all kinds of edits and a familiarity with the idiosyncrasies of your particular DAW. A udio regions Th e building blocks for all editing features are audio regions. Regions are either a complete piece of audio as it was recorded from start to stop or some smaller, THE ART OF DIGITAL AUDIO RECORDING segment of that initial recording that you have subdivided intoasub-region . Th is screenshot shows a region of a complete recording (beginning to end) and then, duplicated on the channel below, that region divided into sub-regions). SCREENSHOT 4.1 R egions and sub-regions T ypically, all regions and sub-regions are simply referred to as regions, but the distinction may be important when editing. Th e region created by each full recording pass is a complete entity, whereas sub-regions created from smaller elements of these regions can be restored to include the entire region. Th e full region created from each complete recording pass is what is stored on the hard drive. Th e sub-regions are simply an instruction by the DAW program to play only a part of the original recording. In Pro Tools, the initial region is indicated in bold type in the regions list and the sub-regions are listed below it in regular type. Th ere may be several diff erent ways to create sub-regions from the initial recording. Th ese are basic editing operations that diff er within diff erent DAWs, but the principle— the ability to create very accurately timed sub-regions—is essential to much of the ed- iting process. Th e segment later in this sec- S CREENSHOT 4.2 tion on edit modes will defi ne the ways that regions can be created and controlled before Region and sub-region list they are edited. C ut, copy, paste Th e most basic kind of audio editing is just like editing with a word processor or just about any other computer program, and it begins with the ability to cut, copy, or paste audio regions. Cutting, copying, and pasting is made possible by the DAW’s use of a clipboard, which is a temporary holding place for data. When any piece of data is either cut or copied, it is placed on the clipboard and available for pasting, but only one unit of data can be put on the clipboard at a time. It remains there until another bit of data has been either cut or copied. A whole universe of editing can be done with these most basic tools—cut, copy, and paste combined with the clipboard function that keeps data available to you as you work., Editing Duplicate, repeat, loop, mute, clear Th e next set of edit functions expand on the basic cut, copy, and paste concept. As with many edit functions, these are oft en simpler and quicker ways of doing something that could be accomplished with more labor using the basic func- tions. Th e ability to work quickly and effi ciently becomes very important when literally hundreds of edit functions need to be accomplished at a single session. D uplicate allows you to duplicate an audio region with one step rather than copying and pasting (two steps). Once a region has been selected (usually by highlighting it), the duplicate function creates a duplicate region adjacent to the original region—that is, the beginning of the region is butted up against the end of the region being duplicated. 121 S CREENSHOT 4.3 Top track is a duplicated region; middle track is a repeated region; bottom track is a muted region Repeat allows you to duplicate a selected region many times, one aft er the other. Th e repeat function requires that you enter a number of repeats into a dialog box. Repeat functions very much like a loop, continuously repeating an audio region, but it does so by actually creating new regions, stretched along the DAW’s timeline (see screenshot 4.3). Loop allows you to continuously repeat a section of the timeline. Th is can be valuable to check the ability of a region to loop seamlessly before creating multiple repeats of that region that stretches along the timeline. Some DAWs also have a loop record function that allows you record multiple takes while looping over one passage (e.g., you could take several guitar solos in a row while the audio looped over the solo passage and the DAW would keep each take as a separate virtual track). M ute is a way of accomplishing the same thing as cut, but without com- pletely eliminating the audio region from your timeline. When a region is muted, it no longer plays—just as though it had been cut—but a grayed-out image of the region remains on your editing timeline. Th is can be helpful when you’re not sure what you want to do with a particular piece of audio. A classic example is editing guitar fi lls. You may be uncertain as to whether a particular fi ll should be included or not. If you mute the fi ll, you can audition the song without the fi ll, but it remains immediately available if you decide you want to use it aft er all. Sometimes there are many elements that you’re not sure about, and by muting them they remain easily accessible and you are, THE ART OF DIGITAL AUDIO RECORDING reminded of their presence by the grayed-out waveform. I’ve frequently had the experience of seeing a piece of audio that I had muted much earlier in the editing process and realized that it was now an element that would be a valu- able addition to the music. Although too many muted regions can clutter the editing screen, a good philosophy is “When it doubt, mute, don’t cut!” (see screenshot 4.3). C lear is a form of cutting that can be useful in certain editing situations. It operates exactly the same as the cut command, except that it does not place whatever has been cut on the clipboard. Here’s an example of how the clear command might be used: Let’s say you’ve copied a piece of audio and are past- 122 ing it into many diff erent places (a snare-drum hit or a sound eff ect, perhaps). As you navigate through the timeline and locate places to paste this sound, you run across a separate piece of audio that you want to cut. If you use the cut command, you will lose the item on the clipboard that you still need to paste in more places. By using the clear command you retain whatever was on the clipboard. Oft en, using the delete key accomplishes the same thing—cutting without placing on the clipboard.

Edit modes

A n edit mode (with one exception) represents a way of restricting our ability to move and place audio regions. Th is may seem odd. Why would you want to restrict your ability to edit? It turns out that restricting the editing capabili- ties allows the recordist to perform some editing functions that would be very time-consuming and tedious without those limitations. Although I am using the terminology from the Pro Tools soft ware, many of these same terms, and most all of these same functions, are applicable in every DAW. First, the one exception: unrestricted editing mode. In the unrestricted mode (called slip mode in Pro Tools) you can place an audio region anywhere on the timeline, down to the smallest possible increment, which would be one sample. Th is will probably be the most frequently used mode, though it de- pends on what kinds of editing you are doing. Grid mode is the restricted editing mode that is probably the most com- monly used. In grid mode your ability to move or place audio is limited to a user-defi ned grid. Th is is frequently used when dealing with music that has been played or constructed to a regulated pulse by using either a click track to guide the musicians or loops set to a specifi c BPM (beats per minute), or both. Th e grid is then set up based on musical time, meaning a grid limited to quarter notes, eighth notes, or some other basic musical division of time. In grid mode you are limited to placing the beginning of an audio region at a grid point along whichever musical grid you have selected. Th e following screenshot shows a quarter-note grid with various regions, all starting at one of the quarter-note subdivisions., Editing S CREENSHOT 4.4 Regions on a quarter-note grid Grid mode is very useful in placing and moving musical events in a way 123 that sets up or maintains an accurate relationship to musical time (beats and bars). Moving any part and maintaining its relationship to the beat, using loops, and repeating parts in various places (like copying and pasting a background vocal part into several diff erent choruses) are all done much more quickly, ac- curately, and eff ectively in grid mode than with unrestricted editing. Shuffl e mode (as it is called in Pro Tools) restricts all editing movement to sliding (or “shuffl ing”) an audio region from its current position to a position butting up against the end of any audio region that precedes it. Th is placing of audio from end to end can be very useful in doing things such as editing of nar- ration, where you are oft en sliding cut up pieces of audio together and you want to be sure to have seamless transitions from one region to the next. SCREENSHOT 4.5 Before and after “shuffl ing” together Spot mode represents the most restrictive of the edit modes and has very limited but very valuable functions. When you have selected a region of audio to edit in spot mode, you are presented with a dialog box asking where you wish to place (or “spot”) the region. Th is comes from the fi lm world, where audio fre- quently has to be placed at an exact location based on the corresponding frame of visuals. In this case, the fi lm frame is identifi ed by its SMPTE timecode—the timing code used to maintain and mark location along the fi lm timeline—and the audio can be placed by inserting the SMPTE timecode location number in the spot mode dialog box start-time fi eld. Th e beginning of the audio region se- lected is then placed at the timecode location indicated. Spot mode is essential for placing music, sound eff ects, and dialog in fi lm and video work. S pot mode can also be set to clock time or musical time (bars and beats), and audio can be placed anywhere on these grids in the same manner. Th is may be useful in placing audio events in certain circumstances, though outside of, THE ART OF DIGITAL AUDIO RECORDING timing to visuals, there are usually simpler ways to place audio than using spot mode. Spot mode does have one other valuable function and that is returning audio to the place that it was originally r ecorded on the timeline (iden- tifi ed as its t ime stamp) . Some- times, audio gets moved acci- S CREENSHOT 4.6 dentally and it can be diffi cult (or impossible!) to return it S pot mode dialog box to its original location without help. When audio is recorded, it is time-stamped with its start time and when additional regions are created they are similarly time-stamped. If there is a discrepancy between the original time stamp (where the audio was recorded on the timeline) and the user time stamp (where the audio is currently sitting on the timeline), you can use the spot mode dialog box to reload the original time-stamp time into the start fi eld, returning the region to its originally recorded position. (See the lower portion of Screenshot 4.6.) O ne way to avoid the above problem is to l ock audio in place. Th is is the ul- timate editing restriction. When an audio region is locked, it cannot be moved or recorded over. Th is can be very useful, especially if more than one person is SCREENSHOT 4.7 L ocked audio regions, Editing working on a project. Th e user can always unlock the audio if needed, but the lock function prevents certain accidental or careless errors.

E dit tools

Th e edit tools represent the heart of an editing system. Th ese tools are used to manipulate audio regions. As with edit functions, there are some edit tools that are familiar from almost any computer application. Again, although I am using the terminology from the Pro Tools soft ware, many of these same terms, and most all of these same functions, are applicable in every DAW. Edit tools may also serve double duty and are revisited in section 6.3, where automation is covered. 125 Th e selector is the tool used to select portions along the timeline. Selected areas are highlighted. Th is tool is represented by a cursor like the one used in most word processors to select text. By positioning the cursor at any point along the timeline, engaging the primary mouse button, and sliding the mouse in either direction, the user can select any region along the timeline. Th is may encompass many audio regions, parts of one audio region, and/or areas of the timeline with no audio. If a portion of an audio region is selected, it may be made into a separate region and then cut, copied, pasted, moved, muted, and so on. A lot of editing begins by selecting an audio region. SCREENSHOT 4.8 A selected region SCREENSHOT 4.9 A separated region Th e grabber tool allows the user to “grab” an audio region and move it along the timeline. Using the grabber and engaging the primary mouse button allow the audio region to be slid in any direction by any amount in slip mode, or the movement may be restricted by the selected edit mode, as described above. Selecting, grabbing, and sliding an audio region to a new location is one of the most basic and common editing functions. Th e trimmer tool enables the trimming of either the front or the back of any audio region. Th is is a convenient way of cutting unwanted material from, THE ART OF DIGITAL AUDIO RECORDING the beginning or end of any audio region. Th e trimmer tool also allows you to restore all or part of a sub-region that has been trimmed (or cut or deleted). SCREENSHOT 4.10 A region before and after trimming Th e pencil tool allows for a very specifi c (and generally fairly rare) editing function, but the tool becomes much more useful in its role in automation (cov- 126 ered in section 6.2). As an edit tool, its only function is to redraw waveforms. In order to use this function, the waveform has to be viewed in a small enough region to be represented by a line (rather than a fi lled-in waveform). You will need to magnify to smaller than 50 ms (milliseconds) on the timeline in order to use the pencil tool in this way. With the pencil tool selected, you can activate the primary mouse button and redraw small parts of a waveform by moving the mouse. Th is yields practical and desirable results in only a very few circum- stances. Sometimes very short glitches in audio that are caused by timing errors or other kinds of interference can be corrected by redrawing the waveform. At- tempts to redraw longer unwanted elements (such as an unwanted click, buzz, or other noise) will at best diminish the unwanted sound but not eliminate it and will oft en cause something worse than the original problem. It is best to du- plicate your audio before attempting to use the pencil for redrawing waveforms to make fi xes because you may be permanently altering the audio fi le. S CREENSHOT 4.11 A short glitch corrected by redrawing the waveform Th e ability to nudge audio in user-defi ned increments is another very use- ful editing function. Th e nudge menu is identical to the grid menu, and it al- lows you to enter values in a variety of formats, including clock time, musical time (bars and beats), SMPTE timecode, or samples. Nudging can be very useful in fi ne-tuning the placement of audio events—for example, moving a guitar fi ll from one place in a song to another. If the recording was done to a click or loop, you can probably just use grid mode and move the fi ll by maintaining the rela-, Editing tionship to the grid. However, if the recording was not done to a click, or even if the performance has wandered a bit from the grid, you may fi nd that the fi ll does not feel like it’s placed quite right against the existing rhythm. In this case, you might set your nudge value to 10 ms, highlight the audio region, and then nudge it earlier or later in increments of 10 ms until you fi nd the place where it sounds like it is sitting right. You can do this by sliding the region with the mouse, but this is not as accurate and not repeatable. You can try several small diff erences in location using nudge, keeping track of the amount and direction nudged, and when you’ve settled on a location, you can go back to it easily and accurately.

F ades and cross-fades 127

Fades and cross-fades are essential editing tools. Fades and cross-fades can be accomplished by defi ning the desired fade graphically and then creating a new piece of audio that follows the fade instructions that you have defi ned (see screenshots 4.12, 4.13, and 4.14 for various examples of this). Fades can also be created by moving (or automating) the output channel fader. Small fades and cross-fades are almost always done using the fi rst method, whereas longer fades, such as song fadeouts, are almost always done using the second method, which is explored in section 6.3, where automation is covered. Short fades S hort fades can be very helpful in smoothing edits. One technique for creating seamless edits is through using zero crossing points. Th e zero crossing point represents the place in an audio waveform where the waveform crosses from positive to negative and the amplitude is zero. Whenever there is any audio right at the beginning or end of an audio region that isn’t set right at the zero crossing point in the waveform, there may be a clicking or popping sound when playing through that region. You can locate a zero crossing point and trim to it, but it’s usually faster and easier to avoid these clicks and pops by creating a very short fade-into or fade-out of the audio region. In most DAWs, there is a way to select many audio regions that you may have created in edit- ing and apply a very short fade-in and fade-out of all of them. If short enough (5 ms is safe), this will be inaudible as a fade but will create a smooth transition in and out of all the regions. S CREENSHOT 4.12 A short fade-in and a short fade-out, THE ART OF DIGITAL AUDIO RECORDING S hort fades are also useful when starting or stopping elements that are part of a continuous audio event. Ending a guitar lick early or starting a vocal line in the middle oft en means that you will need to create a short fade, not only to prevent a click or pop but also to make the new start or ending sound natural. Th e length of these fades will vary depending on the program material, and you will oft en have to experiment to fi nd the most natural fade-in or fade-out of the edited audio event. Th e following screenshot shows a fade-out, a fade-in, and a region set to end at the zero crossing point. Cross-fades 128 C ross-fades can be used on two adjacent audio regions. Cross-fades create fade- ins and fade-outs that intersect the two regions. Short cross-fades can be used to smooth the transition between regions and to avoid clicks and pops. Longer cross-fades can be used to make smooth transitions between sustained sounds— the crossfading action is like morphing—slowly transforming one sound into the other. Th is can be fun for special eff ects, but it also can be useful for certain, diffi cult edits. Editing in the middle of sustained vocal sounds where the idio- syncrasies in sound from one performance to another would make a normal edit obvious can sometimes sound very realistic by using long cross-fades. O bserving waveforms and editing with some visual aids can enhance the use of cross-fades. Using small cross-fades is the quick and easy way to make edits between audio elements that have only low-level sound or silence between them. Some more complicated edits, however, may require more than a simple cross- fade. Using the zero crossing for both sides of an edit will avoid many problems, and sometimes that is all that is required for a seamless edit. A zero crossing edit point plus a small cross-fade is even more likely to produce inaudible results. SCREENSHOT 4.13 A zero crossing edit, a zero crossing edit with a cross-fade Choosing the edit point that is most likely to produce the best results can also be made easier using visual cues. Editing together sections that have equal gain at the point of the edit usually makes for smoother results. In most cases, you have some leeway as to exactly where the edit can be made, and you can, Editing search for matching adjacent regions with similar gain. In the following screen- shot, I show two audio regions with two possible edit points. Th e edit point where the gain (height of the waveform) is roughly equivalent is much more likely to produce the best results. SCREENSHOT 4.14 129 T wo possible edit points, the fi rst being most likely to produce good results

F ade and cross-fade shapes and styles

Th e recordist can select from a variety of shapes for fades and cross-fades. Fades can be linear (straight line), have varying degrees of curvature, or even be “S” curves. A linear fade (a consistent change in volume characterized by SCREENSHOT 4.15 A fades menu showing various options, THE ART OF DIGITAL AUDIO RECORDING S CREENSHOT 4.16 An equal-power cross-fade and then the same cross- fade using equal gain S CREENSHOT 4.17 Editing percussion recordings to avoid unwanted elements a straight line) works for most situations. Cross-fades can also utilize various volume curves. Cross-fades can be created to maintain equal power or equal gain. Equal- power cross-fades means that the overall volume is maintained throughout the cross-fade. Equal-gain cross-fades maintain the gain relationship regardless of overall volume. Equal-power cross-fades work best in most situations, though equal-gain cross-fades might work best for looping the same sound to avoid a spike in level. It is important to keep in mind that with longer cross-fades, more elements from both audio regions will be heard. Cross-fades oft en need to be short in order for you to avoid remnants of unwanted material from one side of the fade or the other. Th e example above shows two recordings of percussive sounds— the top track (track 1) shows the material before the edit point from the second (right-hand) track, while track 2 shows the material past the edit point from the fi rst (left -hand) track. Track 3 completes the edit and shows that with a short cross-fade there would be no extraneous material from either track included. Track 4 shows that a long cross-fade would include bits of earlier or later ele- ments from each track—probably creating undesirable results. 4.2 Screen “Real Estate” Eff ective editing requires careful management of what is showing on your com- puter screen at any given moment. Many editing functions are impossible if there is too much or too little showing on the screen. I think of the computer, Editing monitor as “real estate”—the territory that I have available to work on. Large screens are great for working with DAWs, but you can be eff ective on any size screen if you have good real-estate management techniques. Dual screens can be nice for spreading out, but I’ve done a fair amount of DAW work on my 12- inch laptop, and with good screen management it’s not too bad.

R eal-estate tools

Some of the edit tools are simply real-estate tools. Th ey don’t do any actual edit- ing, but they help you manage what’s on the screen and that allows you to edit properly. Th ese tools, along with the strategies for using them to manage your 131 workspace, are key elements in eff ective editing. M anaging the timeline Th e fundamental real-estate issue for editing is how much of the timeline is showing. You need to see enough of the timeline for the editing function that you’re doing, but you don’t want to be seeing too much more than necessary, so that you can select and manipulate the relevant regions easily. I n most DAWS, there are many ways to manage the amount of timeline showing and being familiar with all or most of them will help speed the editing work. Th ere is usually a magnifying glass tool that allows you to select part of the timeline by holding down the primary mouse button and sliding it across the region that you wish to occupy the screen. When you let go of the mouse button, the portion of the timeline you selected will occupy the entire screen. Th is is a great way to focus on the area you want to work on. Th ere are also usually some q uick key (shortcut) methods for adjusting the amount of timeline that you are viewing. Th ere may be a shortcut for expand- ing or contracting the timeline in increments, allowing you to zoom in or out without having to access the magnifying glass. Th ere may also be presets that allow you to defi ne distinct areas of the timeline with quick key commands. Th is is particularly useful, because you can create easily accessible view areas for each fi le. You might have one command to show the entire song on the timeline, one to show approximately one verse or chorus, one to show approximately one vocal line, and one to show approximately one word. On a diff erent fi le with a much longer timeline—a suite of songs, mastering fi le, or audio book record- ing, for example—you can have diff erent preset regions that are appropriate to that fi le’s timeline. Managing your overall workspace Th ere are many DAW features besides the timeline that require active screen management. You may be able to control the track height for editing; control the size of the waveform regardless of the level it was recorded at; pick which, THE ART OF DIGITAL AUDIO RECORDING tracks are showing and which are still available but hidden; pick from a variety of time- line rulers to show or hide; decide if various submenus, such as a regions list, or sec- ondary windows, such as a system usage window, are in view; confi gure some of the virtual mixing consoles fea- 132 tures, such as number of sends to be visible or hidden; and so on. Th ere are too many op- S CREENSHOT 4.18 tions to detail here and they R ecall options for screen vary among DAWs. setups S ome of viewing op- tions, such as adjusting the height and size of the waveform, are key to the effi ciency of your editing; and some of them, such as extraneous windows that are open, may be small annoy- ances. What is important is that you take an active roll in managing your real estate and trying to optimize your DAW working environment. Most DAWs also have an elaborate recall system for storing and recalling a variety of real-estate setup features. You may be able to recall which windows are in view, which tracks are showing, track heights, timeline selections, and other features. Th is can be particularly useful for large-scale projects, such as movie or video soundtracks, where you may have multiple setups within one fi le, one dedicated to music, one to sound eff ects, and one to dialog. One day you may be editing music and the next day dialog, and the screen setups for the two jobs may be complex and very diff erent. Th e ability to store and recall these screen setups can save you a tremendous amount of time. 4 .3 Virtual Tracks (Playlists) Virtual tracks are an essential part of the vastly expanded capabilities that com- puter recording provides over tape-based recording. Diff erent DAWs use dif- ferent names for virtual tracks, such as “playlists,” “takes,” or “comps.” Beyond a basic understanding of virtual tracks, I cover the working models for using this capability in creating composite (“comp”) performances. I n the analog world, each audio track was limited to one recording—in order to use a track for a new recording, whatever was already recorded had to be erased. But in the DAW, each track may contain many diff erent recordings, each one represented by a virtual track. Th ey are called virtual tracks because each track in the DAW is still limited to one track of playback at a time, though, Editing there may be many recordings to choose from on each track. Th e list of virtual tracks shows all the tracks that have been recorded S CREENSHOTS separately using this one individual track. 4.19a, 4.19b, 4.19c. You can select any of the recordings fromaVirtual track pop-up virtual track for playback, or you can dupli- menus cate the current virtual track for editing or to rerecord a portion of the track, or you can create a new virtual track to record on.

Duplicate virtual tracks

A duplicate virtual track can be made of a track that has already been recorded. Th ese duplicates provide extra fl exibility in the editing process. It is a good idea to make a duplicate virtual track before you begin any editing. Th is allows easy access to the original track at any point. Sometimes you can work yourself into, THE ART OF DIGITAL AUDIO RECORDING a corner with editing and want to just return to the original track and start again. If the original track was created with multiple takes (punched in), it can be almost impossible to recreate the original aft er a lot of editing. With a dupli- cate playlist, however, the original is immediately accessible and you can make another duplicate and start editing again from the beginning D uplicate virtual tracks for material that has already been edited can also be very valuable. Sometimes you may have edited a track or multiple takes onto a virtual track and think that the job may be complete. However, you want to try some diff erent edits and see about diff erent possibilities—perhaps you used a more conservative approach to picking performance elements and you want 134 to see what happens if you select more adventurous performance elements. By duplicating an already edited version, you can create a revised edit without los- ing your previous work. Many edited versions can be stored on diff erent virtual tracks.

New virtual tracks

A new virtual track is a completely blank track available to record on. Whatever has already been recorded on other virtual tracks is still available for playback if selected, but a new virtual track is an empty track. Although each virtual track acts as a new track, it doesn’t make sense to treat it as a completely separate track. If you were to record a rhythm guitar on one virtual track and a lead gui- tar on another virtual track that is a part of the same original track, you would only be able to play either the rhythm or the lead guitar. Virtual tracks are typi- cally used to record many versions of the same thing. It makes sense to record ten tracks of lead vocal on separate virtual tracks if you will only be using one lead vocal on the fi nal recording. SCREENSHOT 4.20 Three tracks and two edited versions using virtual tracks, Editing V irtual tracks versus many tracks Th ere are two ways to approach multiple recordings of the same part (such as a lead vocal) in a DAW: as virtual tracks or as many individual tracks. Virtual tracks off er some advantages over multiple individual tracks. It is simpler to select an individual track for playback using virtual tracks than it is to mute and unmute individual playback tracks. Th ere’s less screen clutter with virtual tracks playing back on a single track than there is with individual playback tracks. Virtual tracks provide a convenient way to store old takes and retain easy ac- cess to them— again, without cluttering up your screen. Virtual tracks may also reduce the load on your CPU (depending on your DAW) by demanding less in terms of track count and audio fi le accessibility. For these reasons, virtual 135 tracks are a great resource when recording multiple takes of a single part. Th ere is more on how to edit multiple takes on virtual tracks in the following section on composite editing.

Composite editing (comping) using virtual tracks

Making composite versions (comping) of multiple performances to create one complete performance has become standard practice. DAWs use diff erent strat- egies for how to manage virtual tracks when comping, but the principle is the same. Elements from many recordings of the same part are pieced together onto a “master,” or “comp,” track of that part. Comp tracks may be made for just about any performance, from drums to solos to string sections, and in many cases, multiple tracks are grouped together and comped (such as with a typical drum-set recording). Probably the most frequently comped performance is the lead vocal, and I use that as the model for how comped tracks are made. Th ere are two common tactics for creating a fi nal lead vocal. One is for the singer to sing the song through until the vocalist and/or the producer feels the performance is either complete or mostly complete. If complete, the task is done. If mostly complete, then punch-in replacement parts are sung for any line SCREENSHOT 4.21 Three vocal takes as virtual tracks and three vocal takes on three separate tracks, THE ART OF DIGITAL AUDIO RECORDING or section to be replaced until a satisfactory replacement is sung. In this case, either there is only one vocal performance or there may be earlier vocal takes as well, saved as virtual tracks, but they have not been used as part of the fi nal vocal. A more common technique is for the singer to sing multiple takes and to then, oft en in collaboration with the recordist, make a composite performance by picking the best lines, or phrases, or words, or sometimes even syllables, from the available takes and editing them into the fi nal vocal performance. Comping tracks can be a relatively simple procedure, but it does require good management of virtual tracks and good editing practices to be done ef- fi ciently. Diff erent DAWs manage virtual tracks diff erently, and this is also an 136 area of further soft ware development. Newer versions of DAWs have been in- tegrating new ways to view and access multiple virtual tracks. Quick and easy access to many virtual tracks makes comping tracks faster and simpler. None- theless, it is a good idea to limit the number of takes you are managing when making a comp. I’ve found that between three and fi ve takes is a reasonable number to work from. Th is may be the fi rst three to fi ve takes done, or you may fi nd that you have recorded a few passes before the singing or playing really starts getting consistently good, and you end up comping by starting with Take 3 or Take 4 and including the following three or four takes. Some will want to use more takes when comping, but it becomes increasingly diffi cult to keep track of which parts you liked the best when listening to so many versions, and it can dramatically increase the overall time it takes to complete a comp. Th e “It Could Have Happened” approach to making comps Oft en, the primary objective in making comps is a fi nal version that still sounds like “it could have happened.” Th at is, it could actually have been played or sung in a single performance. Sometimes this is not important; sometimes we strive to create a performance that obviously could never have been sung or played as it is heard. But most recordings, even when put together from many diff erent performances, still conform to the “it could have happened” ethic. A djusting pitch, timing, and gain while comping Details about adjusting pitch and time are covered in the following section on advanced editing. Th is may or may not be part of the process of comping, though I recommend that you do integrate these adjustments as you work through a comp. Adjusting pitch is not always successful, so if you think you want to use a particular performance, but it has pitch issues that you want to correct, you had best try to do so while you’re making the comp. If you comp together a performance and then go back and try to fi x pitch concerns, you may fi nd that some of the parts selected don’t meet your particular standard of performance, even aft er being adjusted. Where there are issues regarding the timing of a performance, these should also be addressed as a part of the comping process. Sometimes phrases, Editing or words are simply performed slightly earlier or later than desired, and a slight shift in timing makes this piece of the performance sound just right. Of course, pitch and timing are highly subjective. Even though there are theoretical stan- dards for correct pitch, perfect pitch is neither possible nor desired for musical performances. “Fixing” pitch and time in performances should be something that is done within creative guidelines, and these vary greatly among artists and recordists. A djusting gain may also be a part of the comping process. Even though performances may have all been recorded using the same input level, some- times when pieces of diff erent performances are comped together, there are unnatural-sounding changes in volume from one element to the next. While 137 these adjustments can be made with automation, you may want to avoid au- tomation until the mixing stage. Once there is volume automation on a track, you cannot simply move the fader up or down for level changes without dis- abling the automation. For this reason, I oft en make gain (or level) changes by actually processing the piece of audio and “gaining” it up or down using an off - line plug-in. Th is gives me a new piece of audio with the gain more correctly balanced for the context and it allows me to avoid automation until I’m ready to start mixing. Level of detail in comping Comping can be done in large sections or down to the smallest level of detail. With vocal comping, I have made comps in a few minutes by taking whole verses or choruses from a couple of diff erent takes and I’ve spend days comping a single vocal by taking lines, words, and even syllables from many diff erent takes. Using the editing procedures described elsewhere in this chapter, you can edit very small elements—I have constructed three-syllable words using syllables from three diff erent takes!—but this kind of work can be very tedious and time-consuming. Surprisingly, some of the most complicated edits end up sounding very natural—certainly like they “could have happened.” B esides the basic process of comping from diff erent takes, and the ad- ditional possibility of adjusting pitch, timing, or gain, there are other comping procedures that can be used eff ectively. You can take bits of performances from diff erent locations on the timeline and place them where they are needed. Th is is oft en done in the case of background vocal parts, which may be sung in one chorus and then copied and pasted into all the choruses. Th is works easily when the track was recorded to a click or a loop, but can be a challenge (or almost impossible) if the track is not referenced to a consistent tempo. Repositioning audio can be done with almost any part. Pieces of vocals can be taken from one spot and placed in another—either because it is a recurring part or, in the case of an ad lib, simply because you think it sounds better in a diff erent location. Working with a grid when editing a recording that is referenced to a consistent tempo makes this kind of relocation work very easy., THE ART OF DIGITAL AUDIO RECORDING E ven more complex maneuvers are possible when comping. Elements can be recombined in ways that create completely new parts and pieces of music. I have taken bits from solos and rearranged them in ways such that new solos were created that were unlike anything that had been originally played. I have constructed “ad-libbed vocal vamps” from elements that had been sung at com- pletely diff erent locations, sometimes constructing lyric content by stringing words together from diff erent contexts (you can make someone say almost anything this way!). Comping can be endlessly creative—and endlessly time- consuming, if you’re not careful. 4.4 Advanced Editing Our ability to manipulate sound through editing has expanded enormously in the age of the DAW. Although I list the following techniques as advanced edit- ing, most of them really need to be part of every recordist’s arsenal of capa- bilities. Implementation varies in diff erent DAWs and new editing features are added on a regular basis. Th e following represent some of the most common and useful editing techniques beyond the basics already covered in this chapter.

Global edits

G lobal editing is used when you want to remove or add whole sections of a piece of music. Editing all the tracks in any fi le requires its own technique. One of the most remarkable capabilities of nondestructive editing is the ability to ad- just the edit point of each track individually when making a global edit. Th is al- lows us to make edits that would have been impossible with analog tape, where all tracks were necessarily edited at the same point. Global cuts To discuss the process of making global cuts, I will consider one possible exam- ple. Perhaps you have decided that the song should go straight from the bridge to the outro without having a third verse in between, so you need to remove the third verse. Th e process for making such an edit is as follows: 1. S tart by making a copy of all the tracks onto new virtual tracks so that you don’t lose the song construction you had before making the edit. In most DAWs, this can be done with one keystroke by holding down one of the command, control, or option keys and selecting duplicate virtual tracks (or playlists or takes or whatever name your DAW uses to identify virtual tracks). 2. Group all of the tracks together so that you can cut, copy, and paste them all as a unit. Some DAWs have a default “all” group- ing mode for all tracks., Editing 3. S elect the area that you wish to delete (in the example, this would be the third verse). If your music has been played to a click track or a loop, you can do this in grid mode, selecting the area from the downbeat of the verse to the downbeat of the outro—proba- bly something like exactly 8 or 16 bars. If the music was not done to a grid, fi nd some element that plays the downbeat of both sections (the kick drum oft en works well for this if drums are a part of your recording). Carefully select the entire verse from the beginning of the kick drum (or whatever sound) that starts the verse to the beginning of the kick drum that starts the outro. Separate all the tracks into regions that conform to this editing 139 selection and then select them all—all the tracks for the part of the song that is to be cut should be highlighted. 4. Place the DAW editor into shuffl e mode. Th is is the mode that automatically moves the material from later on the timeline and butts it up against the earlier material when audio is deleted (this mode may have a diff erent name in your DAW). Th is edit mode can be seen at Screenshot 4.5 earlier in this chapter. 5. H it the delete key. Th is will cause the selected verse to be deleted and the beginning of the outro to butt up against the end of the bridge. 6. A udition your edit. At this point, you should be able to tell if the timing sounds okay. 7. If you think the edit might work, you can then start going through each track in solo to “massage” each edit point indi- vidually. 8. Each track may require some trimming before or aft er your initial edit point to make a smooth transition. Each track will probably require a cross-fade, the length of which will depend on the nature of the material. Some edit points may need to be slid several beats. For example, the vocal at the end of the bridge may have hung over several beats into the following verse and now needs to be extended into the outro. 9. Th e screenshots on the following page show a global edit of this type both before and aft er the edit, with each track’s edit point adjusted and cross-faded for a smooth transition. Global additions Adding material globally requires a similar approach as global cuts. To explore the process of making global additions, consider one possible example. Let’s say you’ve decided that an already recorded song needs a verse added between the bridge and the outro. In order to do this you will need to copy an earlier verse,


(a) Global edit, before the removal of a verse; (b) Global edit, after the removal of a verse, Editing from the song and insert it in the spot where you want to add a verse. Th e pro- cess for making such an edit is as follows: 1. Start by making a copy of all the tracks onto new virtual tracks so that you don’t lose the song construction you had before making the edit. In most DAWs, this can be done with one keystroke by holding down one of the command, control, or option keys and selecting duplicate virtual tracks (or playlists, or whatever name your DAW uses to identify virtual tracks). 2 . G roup all of the tracks together so that you can cut, copy, and paste them all as a unit. Some DAWs have a default grouping 141 mode for all tracks. 3 . Find the edit point for where you wish to place the new verse. In our example, this would be the downbeat of the outro. Separate all the audio regions at this point (this should require just one editing move, as all your tracks are grouped together). Grab all of the ma- terial aft er the edit point and move it farther along the timeline. You can either drag the material or cut and paste it. At this point, it doesn’t matter exactly where the material is put; you just want to make sure that you leave a large enough gap between the end of the bridge and the beginning of the outro to insert the new verse. 4. Select the verse that you want to repeat for the new verse. Th is could be either the fi rst or the second verse, and once the initial edit has been made, you may be able to use elements from either of the existing verses. If your music has been played to a click track or a loop, you can do this in grid mode, selecting the area from the downbeat of the verse to the downbeat of whatever section follows the verse, oft en exactly 8 or 16 bars. If the music was not done to a grid, fi nd some element that plays the down- beat of both sections (the kick drum oft en works well for this if drums are a part of your recording). Carefully select the entire verse from the beginning of the kick drum (or whatever sound) that starts the verse to the beginning of the kick drum that starts whatever section follows the verse. Copy that verse. 5. Paste the verse into the area you created between the bridge and the outro (screenshot 4.32 on the following page). 6 . Place the DAW editor into shuffl e mode. Th is is the mode that automatically moves the material from later on the timeline and butts it up against the earlier material when audio is deleted (this mode may have a diff erent name in your DAW). Th is edit mode can be seen at Screenshot 4.5 earlier in this chapter. 7. Grab the verse you pasted into the space and move it so the beginning of it butts up against the end of the bridge, then grab, THE ART OF DIGITAL AUDIO RECORDING 142 SCREENSHOT 4.23 Global edit, new verse inserted but transitions not yet made the outro and butt it up against the end of the verse you have just pasted in. 8 . Audition your edits. At this point, it may sound very rough, with a big pop at both edit points, but you should be able to tell if the timing sounds okay. 9. If you think the edit might work, you can then start going through each track in solo to “massage” the edit into shape. Each track may require some sliding before or aft er each of your initial edit point to make a smooth transition. Each track will prob- ably require a cross-fade, the length of which will depend on the nature of the material. 1 0. Th e screenshot below shows a global edit of this type with each track’s edit point adjusted and cross-faded for a smooth transition.

Adjusting timing and pitch

Th e capability to make adjustments in timing and pitch in a DAW has rev- olutionized the creation of music. We can quickly and easily make extensive changes in musical performances in regard to both rhythm and melody. For better or worse, we now oft en alter musical performances as a part of the edit- ing process. I say “for better or worse” because there is considerable debate over the wisdom and value of altering timing and pitch. I believe that both sides have valid points. On the one hand, performances can be sapped of life when pitch and timing are fl attened into very close adherence to absolute standards of in- tonation and rhythm. On the other hand, the ability to make adjustments that satisfy those responsible for the recording (artist and/or recordist) allows some great performances to be salvaged that would otherwise not be used. Before the DAW era I had to erase some spectacular musical moments because they, Editing S CREENSHOT 4.24 G lobal edit, with 143 transitions into and out of the added verse were slightly fl awed in one way or another that was unacceptable to the artist. Th e ability to fi x the timing or pitch of one note (or two, or whatever) within a performance has allowed me to save some wonderful bits of music . Th e following discusses these capabilities without further comment on the advisability of their use. Suffi ce it to say that most people agree that musical performances are not meant to be as close to perfectly in time or perfectly in tune as possible. Th e extent to which they do conform to theoretical perfection varies widely, and all the more so now that we can make adjustments that can bring virtually any performance close to theoretical perfection. A lot of con- temporary music employs constructed drum parts that play with metronomi- cal accuracy, but many of the other instrumental and vocal performances may have a much more “humanly” typical kind of variation in beat accuracy. Many vocal performances are now more accurate in regard to pitch, but they aren’t necessarily made “perfect” throughout. Th ese tools are also used in obvious and creative ways, such as the machine perfection of the bass-drum hook in Chris- tina Aguilera’s breakout song “Genie in a Bottle,” or the obvious pitch-adjusted vocals used for eff ect in Cher’s song “Believe,” and taken to new heights more recently by T-Pain and others. Adjusting timing Th ere are many ways to adjust the timing of a performance within a DAW. Th e simplest is to move a portion of audio by selecting and then dragging it earlier or later along the timeline. Perhaps a guitar fi ll or a vocal line feels rhythmically a little early or late. You can select the piece of audio, slide it, and then audition the results. Th e screen view is important because the smaller the amount of overall time showing on the timeline, the smaller the increment you can com- fortably slide the audio. While the sliding technique works fi ne in many in- stances, there are aids in making these kinds of adjustments. In some DAWs,, THE ART OF DIGITAL AUDIO RECORDING you can “nudge” audio by a user-defi ned amount. Th e advantage to nudging over simply sliding is that you can repeat your actions exactly, trying a variety of positions for the audio and then returning to an exact placement when you’ve settled on a new location. I have found that 10 ms represents a good nudge factor when trying to re- position audio that doesn’t feel like it’s sitting comfortably in the rhythm. Th is is a small enough increment to fi ne-tune location but large enough to hear the dif- ference in one nudge (if you have an ear that is sensitive to rhythm). Sometimes the audio will feel out of place, but you can’t be sure if it’s early or late. You can select the audio and nudge it three or four times (30 or 40 ms.) in one direction 144 and then three or four times in the other, audition the results each time, and by then it’s usually clear which direction is solving the problem. You can place the audio back at its original location and nudge in 10 ms. increments, usually going beyond the proper point and then going back and forth among options until you settle on the one that sounds the best to you. With practice this can be a pretty quick process. Th ere are many variations on this basic nudging technique. Sometimes smaller or larger increments will work better. Sometimes you have to adjust diff erent elements by diff erent amounts. For example, you might fi nd that an awkward-sounding vocal line seems to sit best when the fi rst three words are nudged 30 ms later and the rest of the line only 10 ms later. Sometimes you nudge a piece of audio around for a while and decide that it sounds best where it was originally played! Altering timing based on the relationship to a musical grid is another com- mon way of adjusting performances. If the music was played to a click or a loop, then the DAW will provide a grid that shows the metronomic timing locations. You may fi nd that in some instances moving something onto the grid yields easy and desirable results. For example, you hear one snare drum hit from the drum- mer that sounds late. You locate the snare beat, and you can see that it is beyond the gridline for the beat where it should be located. You separate the snare-drum beat in question from the start of its attack to the end of it—actually you have to move all the drum tracks (as a group) where that snare drum was played because the snare sound leaks into all the other mics as well. You then select that piece of audio and move it so that the approximate center of the snare-drum attack (tran- sients) is centered on the gridline for the appropriate beat. You then audition the results. You may fi nd that you still need to nudge the snare beat earlier or later from the grid to get it to sit right with the rest of the drums. Th ere are many more elaborate techniques and tools for adjusting audio timing based on a grid. Th ese techniques evolved out of the MIDI quantizing function that allowed recordists to align the timing of a MIDI performance to a grid automatically. Because each MIDI event was a completely distinct object, and because MIDI data were very simple to manipulate compared to audio recordings, this was quick and easy to do. Today’s DAWs and computers, Editing S CREENSHOT 4.25 Snare-drum hit lined up with the appropriate beat gridline allow us to quantize audio much in the same ways that we do with MIDI. To do this, we must fi rst divide the audio into individual events. Th ere are programs that analyze audio based on transient patterns; these are the leading-edge high-frequency sounds created when an instrument is struck, plucked, bowed, or blown. Th e program then divides the audio into segments (regions) based on what it has analyzed, intending to yield the beginning of each individual event. Th is can be relatively simple with recordings with very strong transients (drums and percussion) and almost impossible with slow and/or weak tran- sients (vocals and strings). Th ere are user parameters that can be helpful in diff erentiating events based on transient qualities. As these programs have be- come more sophisticated and users more adept, it is oft en possible to create audio that can be quantized to a grid relatively quickly. Many DAWs come with timing analysis and adjustment programs as a feature (such as Beat Detective, Beat Mapping, etc.). W hen using grids, you have a variety of options that may produce more natural-sounding results. As with MIDI quantizing, it is possible to adjust audio elements to a grid but allow for varying degrees of less than perfect placement. You may move the audio a defi ned percentage closer to its grid location (e.g., a quantizing “strength” of 75 percent moves events 75 percent of the way toward the exact grid location). You can also work with groove templates that establish various rhythmic “feels” that are based on variations from perfect timing. You can use groove templates designed by others, use those derived from other performances, or create a groove template based on one of the performances in your recording, and use it to adjust the timing of other performances. For example, you can use one of the timing programs to analyze your drum track and map the variations in performance from the metronomic grid. You can then use this tempo map to “groove” other per-, THE ART OF DIGITAL AUDIO RECORDING SCREENSHOT 4.26 B its of a drum performance quantized 146 using Beat Detective formances (e.g., the bass track) to the idiosyncrasies of the drum track. If you choose to go down these roads, the possibilities are endless—endlessly creative and endlessly time-consuming, too! Adjusting pitch M ore recent innovations in DAW functionality have greatly expanded the re- cordist’s ability to adjust pitch. Th ese are used most frequently to adjust intona- tion—understood as the degree of pitch accuracy—but they can also be used to shift in half and whole steps and to change key signature (transposing). Th e ability to pitch-shift a musical sound, without aff ecting its speed, came early in the development of digital audio and was a great advancement from the direct connection between pitch and speed in analog audio (to raise pitch, the record- ing had to be speeded up—the “chipmunk eff ect”—and vice versa; lower pitch could only be created using a slower playback speed). Now, we can adjust small changes in pitch to improve intonation by selecting a piece of audio and using a transpose function to raise or lower the pitch by a user- defi nable amount without aff ecting the playback speed of the audio. Th is moves pitch in much smaller increments than true note or key transpositions—oft en shift ing only a few cents (there are 100 cents in a musical half step). Th is can be diffi cult for fi xing performances, especially because pitch oft en wavers relative to the note rather than being consistently sharp or fl at throughout. Even when theoretically possible, pitch adjustment requires a very good ear and a lot of experience or a lot of tries to successfully correct intonation this way. Th e development of the now-famous Auto-Tune program changed all this by providing a means of adjusting pitch that was completely variable, allowing for diff erent adjustments in pitch over the course of one performance or even one note. Auto-Tune provides a graphic readout to show the user how the perfor- mance diff ered from “correct” or “perfect” pitch, and the audio can be adjusted graphically or in an automatic mode. In the graphic mode, you redefi ne the pitch of the performance by redrawing the graphic representation of pitch. Th e auto-, Editing matic mode adjusts pitch according to various parameters and can (sometimes) correct whole performances in real time (this may or may not work all of the time, depending on the exact nature of the original performance). Auto-Tune— and now its many successors with similar capabilities—allows detailed pitch ad- justing far beyond anything previously available. Its primary limitation is that it can only adjust single-note performances—it can’t diff erentiate between two or more notes played simultaneously and adjust them independently (though at the time of this writing this polyphonic capability has been released in a new version of Melodyne—an alternative pitch-fi xing program). T o use the graphic mode of pitch correction, you have several routing pos- sibilities. You can copy the piece of audio you wish to adjust onto a new audio 147 track, put Auto-Tune or whatever program you’re using in as a plug-in (insert) on that track, and route it back to the original track using an internal buss. Place the original track in record ready and make sure you are in “input only” mode rather than auto-switching (covered in section 5.3). When you play the seg- ment, the plug-in will read the pitch of the performance and you can make ad- justments within the program as desired, hearing the results through the origi- nal channel. When you are happy with the adjustments you’ve made, simply activate recording and the adjusted material will be recorded onto your original track. You may want to adjust the timing of the new recording to correct the small amount of delay (latency) created by the pitch-correction soft ware. You can also make these corrections off -line, auditioning the material in the plug-in directly from the track and then processing it directly back onto the original track. Th is requires an off -line processing capability (such as audio- S CREENSHOT 4.27 Melodyne pitch and time correcting software, THE ART OF DIGITAL AUDIO RECORDING suite processing in Pro Tools). You will need to check to see if this function also causes a small amount of delay in your DAW. N ew techniques for pitch correction are being added to programs regu- larly and also off ered by third parties. Elastic pitch in Pro Tools now off ers the ability to adjust pitch in real time without having to render new audio fi les (you can render them later if you want to save on computer-processing power). Th e ability to adjust intonation using one or more of these programs has become a part of the recording process. While some artists use it extensively and some very rarely (and a few never), operation of pitch-correcting tools is a necessary skill for virtually every professional recordist. Adjusting by ear or by sight A long with these expanded capabilities of adjusting timing and pitch has come a basic confl ict concerning whether adjustments should be made and the extent to which things should be adjusted. Part of the confl ict comes from the visual nature of the tools we use to make these adjustments versus the aural (sound) nature of the material being adjusted. When looking at waveforms of two diff er- ent performances in the same piece of music, you can see how well they line up with each other—at least it seems that you can see that relationship. Waveforms can be deceptive in appearance, depending on attack and frequency character- istics, but the visual cue as to rhythmic relationships is generally pretty reliable. Th is is similar with pitch correction in a plug-in’s graphic pitch mode. You can see how far the note strays from the “correct” pitch and you can adjust it by sight. Again, there may be some problems with this depending on the nature of the program material, but it is generally quite reliable. We can adjust thousands of elements in performances, both timing and pitch, completely by sight. We can also make these adjustments automatically using auto mode for pitch shift - ing and various forms of quantizing (automatic rhythm alignment)—a whole other creative approach, or can of worms, depending on the circumstances and your point of view. B ut should we use the visual cues for making decisions about adjusting timing or pitch? Th e simple answer is no; adjustments should be made and ac- cepted based on the way things sound, not on the way they look. I oft en tell the artists I’m working with, “If you can’t hear it, don’t fi x it” though the advice is not always followed. Th at said, sometimes visual cues can make the process of time and pitch adjustment much faster and simpler. If you feel something isn’t right and you want to adjust it, using the visual aid of waveform position or graphic pitch readout can make the task easier.

T ime compression and expansion

Another function that is used frequently in contemporary editing is time com- pression and expansion. Th is is the reverse of the new pitch-adjusting soft ware, Editing that changes pitch without changing length. With time compression and expan- sion, you change the length of a piece of audio, making it either shorter or lon- ger, without changing the pitch. Th is has become especially valuable and useful in music that uses preexisting audio and when the recordist wishes to conform the timing of these various elements to each other. Compressing or expanding loops Th e most common use for time compression and expansion is to adjust drum and percussion loops to conform to a particular tempo. Th e use of loops in contemporary music has become very common, and this is partly due to the fact that it is now quite easy to make various loops play at the same tempo. 149 Th e basic operating procedure for setting any loop to a specifi c tempo varies in diff erent DAWs. Many DAWs now have multiple working procedures for time compression and expansion, and there are a variety of third-party plug-ins and programs that make working with loops very simple. Th e fi rst thing to be sure of is that the loop you’re planning to use is loop- ing “correctly.” Th at is to say, when looping back from the end point to the be- ginning, make sure the transition sounds seamless and rhythmically comfort- able. You will certainly need to do this if you are creating the loops yourself, but even commercial loops sometimes require adjustment (adding or subtracting time and or small fade-ins/fade-outs) in order for them to loop comfortably. Once you have created an audio region that is looping properly, you can then adjust its tempo in a variety of ways. Th ere are programs—some within certain DAWs and some from third parties—that allow you to select a region and assign a new tempo to it. Th e program then creates a new piece of audio that has been either compressed or expanded to the tempo you have entered. In order to do this you must know (or be able to fi gure out) the tempo of the original audio so you can instruct the program to change from x tempo to y tempo. While pretty straightfor- ward, there are several programs that do this same thing using shortcuts that make the process even faster and simpler. You may be able to set the tempo for your fi le, go into grid mode (using bars and beats as your time basis), and then take any piece of audio and place the beginning at the start of a bar and use a trimming tool to place the ending at the desired end point (perhaps exactly one bar long, or two bars or whatever). When you release the trim- ming tool, the program will create a new piece of audio that has been either compressed or expanded to fi t exactly the selected amount of time. F urther advancements in time compression and expansion now provide these capabilities without having to render new audio fi les. Th e program ana- lyzes the material and compresses or stretches it as directed. Because it doesn’t actually render a new fi le, it works much more quickly than the traditional techniques. You may have the option of rendering your work later, once you’ve settled on the various expansion and compression adjustments that you want, THE ART OF DIGITAL AUDIO RECORDING S CREENSHOT 4.28 A time compression/ expansion menu to make, in order to relieve the computer’s processing power of analyzing each element during playback. Compressing or expanding melodic material A ll of the techniques described above in regard to drum and percussion loops also applies to melodic material. Oft en, melodic loops, such as a two-bar bass line, are used along with rhythmic loops and their times can be adjusted to a tempo using any of the techniques described above. Sometimes melodic mate- rial is expanded for diff erent reasons. You may fi nd a particular note to have been played shorter than you wish—perhaps it stops a quarter note short of the next note and you want it to sustain to the next note. You could expand the note in question by the small amount needed for it to last until the next note. Quality issues may be more pronounced in compressing or expanding melodic material. Th e delicate timbres of acoustic instruments may be most noticeably altered when compressed or expanded. Th e algorithms used for SCREENSHOT 4.29 Before and after expansion to fi ll a space, Editing compression and expansion have become increasingly sophisticated. Revised and newer programs are capable of greater alterations in even the most delicate audio material with fewer and fewer undesirable artifacts. G lobal compression or expansion Th e ability to easily adjust the overall tempo of the many elements in a typical recording project is one of the very few things that were lost in the move from analog to digital. While it is true that both time and pitch are shift ed when a tape recorder is speeded up or slowed down, it was a welcome capability in many situations. Sometimes things just sounded better a little faster and a little higher in pitch (or slower and lower) than where they were recorded. Th e va- 151 rispeed control on an analog tape recorder has still not been completely repli- cated by DAW developers, but they are getting much closer to similar capabili- ties. they can also separate the alterations in time from those of pitch, providing even greater fl exibility. Compression and expansion programs sometimes off er diff erent settings for diff erent kinds of material, and in some instances they are capable of changing the length of all of the individual tracks within a project or altering fi nal mixes, without noticeable side eff ects.

Strip silence

Another innovative editing tool within many DAWs is a function called s trip silence . Th e strip silence function is akin to what is traditionally done with a noise gate, but it does so with much greater control and operates off -line rather than in real time. Like a noise gate, strip silence provides a means of diff eren- tiating between desired material and noise or leakage by detecting the louder elements. Strip silence “strips,” or removes, the quieter elements rather than “gating” them, but the eff ect is the same as a hard gate. Th e term strip silence is a little confusing because you’re not usually stripping silence; you’re stripping low-level noise and turning it into silence. S trip silence provides a set of parameters, with the primary one being the threshold (just like a noise gate). Th e threshold is set in dB and determines the level above which material is retained and below which material is stripped into silence. Along with the threshold, you set the minimum length of time for each element (to avoid very short spikes or random noise elements, if desired). You can also set an attack buff er (region start pad), which allows you to retain the rising transients at the beginning of a sound whose level exceeds the threshold rather than simply starting the sound once it has reached the threshold. A re- lease time (region end pad) can also be set to allow for the natural decay of a sound, even though it falls below the threshold. In practice, strip silence is most useful on drums, percussion, and other material made of relatively short percussive sounds, although it can also be used on material of mixed sustain such as guitar tracks, where you might want to, THE ART OF DIGITAL AUDIO RECORDING eliminate a lot of amp noise between the guitar parts. Th e most typical usage would be on kick, snare, or tom-tom tracks within a recording of a full drum set. Sometimes leakage from other drums onto these tracks may add undesir- able elements to the overall drum sound. Th e great advantage of strip silence over real-time gating is its ability to adjust for anomalies, such as a snare-drum hit that might fall below the general threshold of noise on the snare-drum track (a roll that starts quietly and cre- scendos in volume, for example). Th e following screenshot shows a snare track that is ready to be stripped, then stripped, then adjusted to include the low-level snare elements in the roll. It is shown on three diff erent tracks here to follow 152 the process, though, of course, only the fi nal, processed snare drum would be playing. SCREENSHOT 4.30 A snare-drum track before processing with strip silence, after, and fi nally adjusted for low-level elements In regard to tom-tom tracks, it is almost always desirable to strip silence. Th is is because the positioning of the toms means that there is likely to be a lot of leakage into those microphones, and yet the toms themselves may be played rather rarely. However, it is oft en easier to manually “clean” the tom tracks rather than bothering with the strip silence function. To do this you would simply navigate to each place the toms are played, select the region, and sepa- rate it, leaving a little buff er before and some release time aft er the tom hit. Th e material before and aft er the tom hit can be muted or cut. Th e eff ect is the same as what would happen with strip silence, but for a limited number of tom hits this would be faster.

M iracle edits

Using combinations of the above-mentioned techniques (and others that will no doubt be arriving soon to various DAWs), it is possible to do some rather miraculous things when it comes to editing. By exploring diff erent possible edit points and diff erent sizes and shapes of cross-fades, you can massage into S CREENSHOT 4.31 M anually “cleaned” tom tracks, Editing shape the edits that might at fi rst seem problematic. Adding the ability to gain change, pitch shift , and/or time compress or expand can sometimes allow for seemingly “miracle” edits. Aft er doing a lot editing, using all the tools at hand, you will learn to foresee your options and choose the most likely tactic for successful edits. Unless you’re really pressed for time, don’t give up too quickly on a diffi cult edit—you may fi nd that you will create some miracle edits of your own!,

C hapter 5 T hree Fundamentals

154 T echniques Every Recordist

Needs to Know

Th is chapter covers essential technical and procedural practices that are part of almost every recording setup: inserts, sends and returns, and auto-switching. Th ese are signal-path basics that oft en pose the greatest problems for many be- ginning to intermediate recordists. Th ese three fundamentals are not obvious parts of the mixer or recorder, but are crucial signal-path operations used in almost every recording session. Proper routing for digital-signal processing (DSP, such as EQ, compres- sion, delay, reverb) is one of the most basic practices in audio recording, yet it is oft en done improperly and in ways that make the recordist’s job more confus- ing and complex than need be. Th ere are two fundamental techniques for using signal processing: the insert model; and the send and return model. I cover them here as the fi rst two of these three fundamentals. Th e third fundamental— auto-switching—describes an important option in the monitoring capabilities of every DAW and is discussed later in the chapter. 5 .1 Inserts/Plug-ins P roper routing begins by employing the most effi cient signal-path model for each type of signal processor. In chapter 2, I covered all of the basic types of signal processors and I introduced the idea of plug-ins, which is the format for using signal-processing tools in the DAW. You’ll remember that the plug-in is oft en the digital equivalent of the eff ects box from the analog world of hardware processing. Th e simplest way to use a plug-in is as an insert, and it is the proper, Th ree Fundamentals way to use many, but not all, of the signal-pro- cessing tools.

I ndividual channel inserts

A n insert is a means of making a particular processing tool a part of the audio channel. Th e easiest way to understand an insert might be to go back to the way we access EQ in an analog mixer. In almost every analog mixer, there is EQ circuitry built into each channel. Th us, as the signal fl ows through the channel path, it passes through the EQ. Th at is to say, equal- D IAGRAM 5.1 155 ization circuitry is physically inserted into and I nsert routing for an EQ plug-in made a part of the channel in the mixer. You probably just thought of the EQ as part of the channel, not as at insert, butitis the insert model of routing that makes the EQ part of the channel. In the virtual mixer of a DAW, there are multiple positions to insert plug-ins, each one making any inserted processor part of the channel. J ust as EQ is normally put into use as an insert, the same is true for the general category of signal-processor tools called d y- namics. Processors that control dynamics generally reshape the audio of each individual sound, creating new contours in the fun- damental gain structure of the processed audio. Th e insert model allows dynamics processors such as compressors, limiters, expand- ers, and gates to completely integrate their eff ect into the chan- nel output. Using EQ and/or compression on inserts of individual channels is a very common method of shaping sound as part of the mixing process, as described in section 6.2. B ecause inserts are eff ectively a part of the channel, when you are using more than one processor on a single sound source, the specifi cs of the insert signal path become an important con- cern. For example, consider the common technique of using an EQ and a compressor on a single channel. Two processors in- serted on the same channel must be placed in sequence—that is, one insert must follow the other in the signal path. Th erefore, there is the potential for interaction between the two processors. For example, when EQ and compression are in sequence, the ac- SCREENSHOT 5.1 tion of one of the processors may aff ect the other. What happensAnEQ plug-in inserted if the compressor follows the EQ in the signal path? Th e com- directly into the fi rst- pressor’s functions are dependent on the threshold setting that position insert of a virtual mixing console controls the actions of the compressor based on the level of the, THE ART OF DIGITAL AUDIO RECORDING DIAGRAM 5.2 EQ feeding a compressor incoming signal. If the signal level that feeds the compressor is increased or decreased (while the threshold is constant), then more or less compression will 156 be applied to the signal. But boosting or dipping frequencies using EQ aff ects the signal level. So, if the compressor follows the EQ in the insert path, changes in the EQ settings will aff ect the actions of the compressor. Th is may cause un- wanted eff ects. D espite the possible problems created by placing a compressor aft er an EQ, there are times where you may choose this signal path. You may want the EQ to infl uence how the compressor responds. For example, if you are adding a lot of low end to a particular sound, and you want the compressor to control the dynamics based on this added low-end content, then the compressor needs to follow the EQ. If the EQ follows the compressor in the signal path, the com- pressor is unaff ected by changes in the EQ setting. Changes in the compressor settings won’t aff ect the EQ because the eff ects of EQ are constant, regardless of changes in input level. For this reason, placing the EQ aft er the compressor is the more common routing for using these two processors on a single channel. It is simple to reorder the insert sequence in most DAWs by dragging the insert into a new position. As you build a sequence of inserts on an individual track, it may be necessary to change the order so as to control the interaction between the processors.

Inserts on groups and on the stereo buss

I nserts are also used on groups and stereo buss (master) channels. In these cases, you are applying DSP on multiple tracks; but again, you are integrating the processing directly, using insert routing. S CREENSHOT 5.2 You may want the same EQ or the same compressor on a group An EQ follows the of tracks—drums, backing vocals, or whatever. While the eff ect is compressor in a typical diff erent from individual track processing, it may be desirable (for insert series example, group compression tends to blend elements). It is also, Th ree Fundamentals S CREENSHOT 5.3 Multiple drum channels grouped and bussed to a stereo auxiliary channel with a compressor on the insert of the aux channel more effi cient in terms of computer-processing power to use groups to do this rather than using individual processors on each channel. Group compression on drums is quite common. Similarly, compression or EQ (or other eff ects) may be used on the over- all mix by placing them on the master or sub-master fader insert. In chapter 6 (mixing) and chapter 7 (mastering), I explore the specifi cs of using DSP on groups and on the mix buss. Th e graphic interface used to control processors takes advantage of the computer’s fl exibility and helps to maintain the best use of the monitor screen’s real estate. Th is means that when a processor is placed on an insert, it is gen- erally shown as a small box in the insert section of the virtual mixing board. When the box is clicked with the mouse, the full control panel of the processor is revealed and parameter changes can be made. Th e parameter control panel can be put away when not being used, so as to maintain a clear working space for other functions. A typical EQ control panel is shown on the following page., THE ART OF DIGITAL AUDIO RECORDING SCREENSHOT 5.4 A six-band EQ control panel WHAT NOT TO DO Don’t always follow the “rules!” This is a reminder that audio recording is a creative endeavor, and as with all creative endeavors, rules are made to be broken (sometimes). There are good reasons for the standard operating procedures, and most of the time we are best off if we follow them, but there are always instances where breaking the rules might yield desirable results. That’s why I keep using the words typically or generally in my explanations; these suggestions are not meant to be absolute. For example, using EQ or compression on the send and return model may be worth trying, and it provides a different kind of fl exibility that you may like. In general, the model that puts EQ and dynamics on channel inserts is going to work best—but nothing is to be considered sacred. You never know what unorthodox experimentation might yield! However, any experimentation is going to work best if it is done with knowledge of what rules are being broken and what kind of effect is being sought. Random experimentation that isn’t based on a solid understanding of fundamentals tends to waste a lot of time and yield random results. Th e input and output connections for a soft ware plug-in is handled au- tomatically when it is inserted into one of the insert positions on the virtual mixer. Hardware inserts (on mixing consoles, guitar amps, etc.) require a physi- cal input and output to get to and from the piece of hardware (processing unit, reverb, EQ, etc.) that’s being inserted. Th ese connections need to be made with, Th ree Fundamentals cables. Th ese I/Os are sometimes labeled “insert in” and “insert out” and some- times “insert send” and “insert return.” In this case, the terms “send” and “re- turn” mean the same as “output” and “input.” Th is is a slightly diff erent use of the terms from the send and return model for the use of soft ware DSP plug-ins that I describe in the following section. 5 .2 Sends and Returns One of the most common and most challenging signal-path models for using signal-processing tools (plug-ins) is the send and return model. I introduced the idea of sends and auxiliary inputs (used for returns) in the previous section on mixing boards (section 2.4), and delved further into the use of sends in the discussion of headphone mixes (section 3.2). Th e complete send and return 159 model is probably the most complex kind of signal routing that is still a fun- damental part of basic engineering practice. Sends and returns form the rout- ing model that is the alternative to using direct-channel inserts for plug-ins, which was covered in the previous section. In brief, rather than inserting the plug-in directly into the channel, you use an auxiliary input channel and insert the plug-in there, accessing it through the mixer’s send capabilities. I show this routing model in detail, but fi rst it helps to understand the most frequent uses for the send and return model.

S ends and returns rather than direct-channel inserts

Generally, it is reverbs and delay—the time-based eff ects fi rst discussed in sec- tion 2.7—that are accessed using sends and returns. Th e key reasons are that the send and return model allows you the share these eff ects among many audio channels and provides greater fl exibility in operation. Th e logic of this is pretty straightforward: time-based eff ects simulate environments (rooms, concert halls, the Grand Canyon, etc.) and you may well want diff erent audio elements to share acoustic environments (you may want all guitars to sound like they were played in the same room, for example). Because the EQ and dynamics ef- fects are specifi c to the sound (this guitar brighter, another guitar warmer, etc.) the direct-channel insert on each individual instrument is generally the best approach for EQ and dynamics processing.

Send and return signal path

It makes it easier to remember how to set up a send and return system if you keep the logic of what you’re trying to accomplish in mind. You want to be able to access one eff ect (a reverb, for example) from many diff erent audio channels. In order to do this, you must place the eff ect on an auxiliary input channel (or aux track). Th e aux track is always accessible for input, unlike an audio chan- nel that only receives signal when it is in the record mode. Once you’ve created, THE ART OF DIGITAL AUDIO RECORDING an aux track and inserted a reverb (or other signal-processing plug-in), you want to be able to send audio from any audio track to that reverb. Each audio track has many auxiliary sends, so you need to make sure that you are using the correct send to get the audio signal over to the reverb on the aux track. As previously discussed regarding aux sends (section 2.4), you have the choice be- tween external routing (through the hardware interface) and internal routing (through the internal bussing system). Because you are operating with internal processors (plug-ins), you want to use the internal buss system for your routing. So, you start by setting the input of the aux track to a buss, and if this is the fi rst buss you are using in this particular session, you might as well start with buss 1. Th is means that any audio sent over buss 1 will arrive at the input of the aux track and get fed into the reverb. 160Inorder to send audio from a given audio track, you must create a send for that track and assign it a buss routing. To access the eff ect that is receiving signal on buss 1, you must create a send that is sending on buss 1. Having done that, you set the level of the send for each individual track and you have created a send and return model that allows you to access that particular eff ect from any audio track via buss 1. Besides providing access to the eff ect (plug-in) from any audio track, the send and return model provides considerable fl exibility. Th e overall return level of the eff ect is variable (more or less total reverb, for example), but by adjusting the level of the send from each track, you can vary the amount of eff ect on each track independently. Th e routing model for two audio channels being sent to the same reverb, but with diff erent amounts of reverb added to each track, may be described like this. • Audio track 1 with a send routed to buss 1 and the send level set to 0.0 dB (unity gain). • A udio track 2 with a send routed to buss 1 and the send level set to -5.0 dB. • Aux track with a reverb inserted and set to receive on buss 1 . A s the output of the aux track is raised or lowered, there will be more or less overall reverb added to both the audio channels, but there will always be 5 dB less reverb on audio track 2 than on audio track 1 (unless the send levels are altered). B y using the send and return model, you have balanced the original audio track (the “dry” signal) with the eff ect (the “wet” signal). Th is process of balanc- ing audio and eff ect is sometimes accomplished by inserting the eff ect directly on the audio channel (the insert model) and then balancing the two using a dry/wet control on the plug-in. Using a send and return model allows for easier balancing of dry and wet (the two are controlled separately with faders rather that interacting with the single dry/wet control), while at the same time provid- ing the added fl exibility of use on multiple tracks., Th ree Fundamentals D IAGRAM 5.3 S end to reverb

Send and return model and panning

Th e send and return model also provides more fl exibility for changes in pan- ning between the dry signal and eff ect. In the example below, one audio channel is shown with a send going (via buss 1) to an auxiliary track, which is receiving the send on buss 1. Th ere is a reverb inserted on the aux channel. Th e original audio is panned 50 percent left and the reverb is panned 75 percent left . Th is can be helpful in creating the subtle panning relationships that combine to pro- duce the overall stereo fi eld. Expanding the send and return model using stereo eff ects requires man- aging stereo sends and returns. One frequent model maintains the mono send going from the audio channel, but a stereo eff ect is created by using a mono in/ stereo out (mono/stereo) plug-in. Th is was discussed in section 2.7, when re- verbs and delays were introduced and I noted the common practice of feeding a mono signal into a reverb and letting the DSP create a stereo reverb eff ect. Th is is much like what occurs in nature when a single sound source (voice, guitar, horn, or whatever) is aff ected by the room acoustics and received by our two ears. It’s our two ears, receiving slightly diff erent versions of the eff ects of room acoustics, that create our sense of stereo, even though the original sound was mono (single sound source). Reverb plug-ins simulate this eff ect in their mono- to-stereo mode. T rue stereo eff ects require stereo sends, and they allow the panning from the original-source audio to be refl ected in the eff ect (a stereo reverb, for ex- ample). As in the example below, the original audio (which is a single track and thus a mono source) is sent via a stereo send and the send is panned 75 percent to the left ). Th e stereo reverb receives the panning information and the amount,


A send and return setup showing a variation in panning between the audio channel and the reverb return


Mono aux to stereo reverb, Th ree Fundamentals DIAGRAM 5.5 Stereo aux send to stereo reverb of reverb is balanced according to the panning of the send. Th is is particu- larly useful in instances such as sending an entire stereo program to a reverb (a completely mixed song containing many elements, for example), and you want the reverb to retain as much of the left /right panning fi eld as possible. I t should be noted that not all re- verbs or other time-based delays off er true stereo functionality even in their “stereo” mode. You can check this by setting up a send and return model such as the one in Screenshot 5.6 and see if the reverb return follows the send panning. If it doesn’t, the plug-in is not operating in true stereo mode but, rather, blending the left and right out- puts to maintain a balanced stereo output (as it does in the mono in/stereo out oper- ating mode). In Screenshot 5.6, where a mono send (buss 1) is feeding a reverb set to the “mono/stereo” mode, note that the stereo return is not identical left to right—at the moment captured, the right channel is S CREENSHOT 5.6 slightly louder. Th is is because, in simu- M ono aux send to stereo reverb lating the kind of stereo spread created, THE ART OF DIGITAL AUDIO RECORDING by room ambiences arriving at diff erent times to each ear, there are going to be continuous but minor variations in the left and right channels. In Screenshot 5.7, where a stereo send (busses 1 and 2) is feeding a stereo reverb (inputs set to busses 1 and 2), note that the send is panned 75 percent to the left , and as a result, the reverb return is considerably louder on the left side. If an- other audio track were sent to this reverb with the panning set far to the right, the 164 reverb for that audio would appear pri- marily on the right-hand channel of the reverb return.

Send and return and CPU

usage One fi nal note on sends and returns: Be- S CREENSHOT 5.7 sides all of the benefi ts of using this model Stereo aux send to stereo that have been covered above, the ability reverb to use one eff ect on many tracks adds the benefi t of saving on computing power. Whether your plug-ins are being pow- ered by the host computer CPU or through an external card or interface, plug- ins—especially reverb plug-ins—can gobble up computer processing power, so sharing eff ects among tracks can greatly aid in maximizing the effi cient use of your system. WHAT NOT TO DO Reverbs and delays on inserts. Typically it is not a good idea to use individual channel inserts for reverbs or medium to long delays. This is partly because it is often desirable to share delays or reverbs between two or more tracks, which requires the send and return model, but even if the effect (reverb or delay) is being used for only one audio track, the send and return model provides easier use and more fl exibility, as described above. Still, it is a good idea to know how to use a reverb on a channel insert and be able to make adjustments with the wet/dry control in case you have run out of sends., Th ree Fundamentals 5 .3 Auto-switching (Auto-input) A uto-switching (same as auto-input) refers to changes in the monitoring dur- ing the course of normal punch-in recording. Punching-in refers to rerecording parts of a previously recorded performance. M onitoring refers to what you are listening to during playback and recording. Certain signal path procedures, such as auto-switching or using sends and returns, are very common yet dif- fi cult to fully understand. Again, it is most helpful to begin with the logic of what you are trying to do before you look at the more technical aspect of signal path.

I nput-only mode

Inalot of music production, punching-in is used extensively aft er the initial re- 165 cording. You may punch-in and replace a portion of a track or you may punch- in somewhere in the middle and record all the way to the end of the track. In any event, the process of punching-in is dependent on the musician’s abil- ity to hear (monitor) in an appropriate manner. Th is means that the musician needs to be able to hear what was previously recorded on the track up until the time the punch-in is made and recording begins. So the monitoring must be switched while the music is playing—switched from playback (what was previ- ously recorded) to input (what is being recorded). Th e evolution of this kind of switching ability on analog tape recorders marked a major change in capa- bilities. Th e original monitoring default situation, prior to the advent of auto- switching, was what is referred to today as “input-only mode.” In this mode, a track that is armed (in record ready) is always monitoring input regardless of whether it is playing back or recording. Playback is not available until the track is taken out of record ready. In order to complete an eff ective punch-in, it is necessary for the recorder to be able to be switched from playback to record while running. It would not be possible to do an eff ective punch-in while in input-only mode because the performer may not be able to tell where he or she was in the arrange- ment of the music. It was not an easy technical devel- opment for an analog tape recorder, but eventually the electronics were developed and the ability to punch-in was created. As shown later, this was not easy for the computer to accomplish ei- DIAGRAM 5.6 ther, but fi rst the details on Input-only versus auto- switching punch-ins., THE ART OF DIGITAL AUDIO RECORDING

P unch-ins and auto-switching

Let’s say there’s a singer in the iso booth and you’re about to fi x (rerecord) a few vocal lines. As you’re preparing to do this, you want to be reviewing the plan with the singer. Th is means that you need to be in communication with the singer, and that means that you need to be monitoring his or her input. By placing the track into record ready the recorder automatically switches the monitoring status of that track from playback to input (while stopped). Once you’re ready for the punch-in, you will start playback. If the recorder is not in auto-switching (or auto-input) mode, the singer’s track will continue to moni- tor input while the recorder is running (input only). Th at means that the singer can hear him or herself but not the pre-recorded vocal and he or she won’t be 166 able to tell where the entrance for the punch-in is. In auto-switching mode, when playback is started, the singer’s track is automatically switched from input to playback, even though it is in record ready. Th is allows the singer to hear the already recorded vocal. When it comes time for the punch-in, the engineer ac- tivates recording and the track starts to record and also starts monitoring input. Go out of record (punch-out), and playback is monitored again (as long as the recorder is running). Stop running the recorder, and the singer’s track reverts to input (and two-way communication via talkback is available again). Th e gist of the matter is this: in auto-switching, the recorder is automatically switching between input and playback according to the demands of a typical punch-in recording. Th e following diagram indicates the diff erences in the monitoring status of input-only mode and auto-switching. Note that the input-only model is sometimes referred to as “audition” mode because it allows the user to audition whatever is to be recorded, without actually recording. Th at’s because you can play the recorder and be listening to the musician sing or play as long as the new track is in record ready and input- only mode. You cannot hear anything that may have already been recorded on that track, so it is not a useful mode for punching-in, but this might (see below) be convenient for such activities as setting levels or warming up. WHAT NOT TO DO Don’t use input-only or audition mode. This may seem rather odd advice, but in the age of computer-based, nondestructive recording there is little reason to use input-only or audition mode. As explained in the following, the computer environment lends itself to “always being in record” whenever a musician is playing to a track, and this generally eliminates the usefulness of input-only mode., Th ree Fundamentals L et’s say you are setting up a vocalist and you need to have the person sing to the track so you can set the record level and the singer can check the headphone mix. The likelihood is that you’ll be adjusting the level as he or she sings, which probably prevents you from using anything that is recorded, and they’re just warming up anyway. Often singers will say something like, “I just want to try it once—don’t record this.” So if the recording is going to be unusable anyway, and/or the singer doesn’t want the track recorded, shouldn’t you use audition mode? I say no! One of the beauties of DAW recording is the undo function. I suggest that you remain in auto-switching mode and simply go into record from the start in order to be hearing input at all times. You are recording, but the effect is the same as being in audi- tion or input-only mode from a monitoring point of view (you’re always hearing input when you’re recording). When the level testing or trial run is 167 over, you can easily hit “undo record” and that recording is gone, just as if you were in audition mode. But on some occasions, after a supposed audi- tion or test run, I have had singers or other musicians ask, “Did you record that?” You just never know when you might get something good—and with the DAW there’s no risk of losing something already recorded, as there was with analog tape. Sometimes people play or sing particularly well when they don’t think they’re being recorded—and first takes can have a magic that is unreproducible! S ometimes a musician will play or sing something I particularly like dur- ing a trial or warm-up pass. I might say, “I really liked what you played dur- ing the bridge [or wherever] in that warm-up.” In the past, the musician might reply, “I have no idea what I played in that part. Did you record it?” If I had been in audition mode, I would not have been able to play what they had done. Now, if I have recorded it, I can go back and play it for the musician so the person has a reference. With nondestructive recording (and with the price of hard drives so low that storage really isn’t an issue), there is no reason not to a lways be re- cording, even if you delete it later or have responded “Okay” when a musician asks not to be recorded. (If a musician has asked not be recorded, and if he or she doesn’t ask if you happened to record that bit, and if you don’t tell the singer that you did record it, the right thing to do is to eliminate that recording before moving on). Saving practice runs or any number of alternate takes is easy using virtual tracks, covered in section 4.2.

H ow auto-switching works in a DAW

F inally, a note about the way auto-switching is accomplished in a DAW. As I mentioned, this was a technical challenge that had to be overcome in the world of analog tape recorders, and it turned out to be a technical challenge for the DAWs as well. Th e problem for the DAW was that it is not easy for a com-, THE ART OF DIGITAL AUDIO RECORDING puter to start recording. Streaming 24-bit audio onto a hard drive at 44,100 samples (or more) per second is pretty demanding. As a result, it takes at least a few milliseconds for the computer to begin a recording. You may notice this when you go into record from stop—there is a slight delay before the recorder actually starts up and starts recording (more or less of a delay depending on how fast your computer is, how many tracks are in record, how many playback tracks and plug-ins are in use that are making demands on the CPU, and how effi ciently the soft ware is that you’re using). Regardless, any discernable delay is unacceptable in a punch-in situation. Th e recorder needs to respond to the record command immediately. Th e solution in the DAW is both ingenious and benefi cial in unexpected ways. I n order to provide immediate punch-in capabilities, a DAW actually starts 168 recording on any track in record ready as soon as playback is started. Th is is why you may notice a slight delay on startup if one or more tracks are in record ready, even though you haven’t instructed the recorder to start recording yet. Th e DAW is recording on those record-ready tracks from startup, but it is “pre- tending” not to be recording! Th at is to say, it is monitoring playback on those record-ready tracks just as it should be in auto-switching mode prior to being placed into record, even though it is recording on that track at the same time. Unlike an analog tape recorder, the DAW can record and play back on the same track at the same time because it uses random access storage—it isn’t limited by a physical tape track. So, the DAW is recording, pretending not to be recording, and as soon as you tell it to record, it switches to input and places the new audio in the timeline, appearing to act just as it would have on an analog tape recorder track. Th e same is true when you punch-out; the DAW continues to record, but the monitoring switches to playback. Th is allows for instantaneous punching because the DAW isn’t actually punching-in, it isn’t really going into record; it already was in record and it is simply switching the monitoring from playback to input. Th e supplemental benefi t is that all the stuff before and aft er the actual punch was recorded as well. Actually, the ability to uncover or trim back material from before or aft er the punch can be both a blessing and a curse. It’s a blessing because sometimes you may have been late with a punch and you can retrieve the bit that you missed. Or the musician might say, “I think I played a great lick right before the punch; can we hear that?” and in fact, you can hear it and keep it if you want to by uncovering it on the track’s timeline. Th e curse is the way in which this might encourage sloppy punch-in and punch-out habits. While it’s true that if you punch late you haven’t actually missed the point at which you were supposed to punch-in (it’s been recorded and is easily retrievable), the musician couldn’t hear what he or she played or sang at the point he or she was supposed to enter because playback was still being monitored until the punch was made. Th is can be distracting for the musician. When (I must admit) on occasion I have made, Th ree Fundamentals a late punch, the musician will oft en ask, “Did you get the beginning of that?” Th ey couldn’t hear it, so they didn’t know if was actually recorded. Many musi- cians now know enough about DAW operation to recognize that the beginning had been recorded—but it’s still distracting. Careful, accurate punching is still an important part of good studio practice.,

C hapter 6 M ixing Th e Most Creative and the

170 Most Challenging Stage I call mixing “the most creative and the most challenging stage” because there are endless variables to mixing and much less in terms of the concrete guide- lines I’ve been presenting in regard to making good recordings (mic techniques, etc.). Mixing requires imagination and vision in order for you to achieve your sonic goal for the fi nal mix—this is very creative work. But there’s a lot of detail work that needs to be done to serve the larger vision, and there are a lot of tech- nical elements that aff ect your ability to get from your recorded material to your goal for the fi nal mix. Th ese are the challenges. Sometimes I hear mixes of music and my immediate response is, “What were they thinking?” Some mixes sound so wrong to me that I am at loss to understand how the recordist arrived at what it is that I’m hearing. On the other hand, I sometimes focus on the mix of a piece of music that I’ve heard many times and realize how truly odd the mix is and how diff erent it is from what I would have likely done had I been the mixer—yet I have accepted and enjoyed the music (and the mix) without noticing its details. Both cases remind me how subjective mixing is. For the most part, we can assume that the listener does not consciously notice the details of the mix (how loud the vocal is or how aff ected the guitar sound), but we can also assume that these details aff ect the impact of the music on the listener—possibly even to the point of making the diff erence between the listener’s liking or disliking the recording. Th e following is intended to detail the way the DAW tools are used in the mixing process, examine the various elements that should be considered while mixing, and raise the creative issues that each recordist will answer in his or, Mixing her own way. It is organized along the more practical guidelines—what you need to do in order to mix, how you build your mix, and how you fi nish your mix—but the more subjective and creative challenges arise within each part of the process. 6.1 Mixing Requirements W hat do you need in order to eff ectively mix a project? Th ere’s no simple an- swer, but fi rst you must ask both what is meant by e ff ectively and what is meant by the p roject at hand. Being an eff ective mixer requires a certain amount of experience, a critical ear, and usually a healthy willingness to collaborate. Mix- ing is a skill as well as a creative endeavor, and there’s no substitute for time spent mixing to develop that skill. Mixing also requires a good listening envi- ronment and an appropriate set of tools to manipulate sound. What constitutes these technical requirements may vary considerably among recordists working 171 in diff erent styles of music. Having the luxury of a home system or good ac- cess to a commercial facility, along with projects to work on, will allow you to go through the trial-and-error process necessary to develop eff ective mixing skills—guided by the good advice from this book, of course. A s to the project, the nature of the recording and the music you are mixing will greatly infl uence your ability to mix eff ectively. Musical genres have many conventions in terms of how mixes sound; and even if your goal is to defy those conventions, you will likely have limited success mixing styles of music that you are not very familiar with. Th e number of sonic elements in the musical piece is also important to the mixing skill set. Th ere can be masterful mixes of solo piano recordings, but that is quite a diff erent task from mixing a hip-hop track with tons of loops, percussion, samples, instruments, rappers, vocalists, and background singers. Diff erent projects suggest diff erent sets of tools and require diff erent kinds of experience with mixing in order for you to achieve outstanding results. U ltimately, your greatest asset in mixing is the same as your greatest asset in all other elements of the recording process—your ear! Th e more experienced and developed your ear, the better your chances for eff ective mixing of any kind of project. If I were hiring a mixer, I would opt for an ear that I trust far above any considerations of quality of gear being used (though both a great ear and great gear is really what you want).

W hat is mixing and remixing?

Let’s establish exactly what is meant by the term m ixing . As the word suggests, mixing is the combining of audio elements. While mixing in some form has been an essential part of recording from the beginning, it was initially accom- plished by the placement of musicians and microphones as the music was being, THE ART OF DIGITAL AUDIO RECORDING recorded. If the singer wasn’t loud enough in the mix, he or she was moved closer to the mic.. It was with the advent of multitrack recording that the con- temporary process of mixing began. Because many distinct elements are recorded on separate tracks in the typi- cal DAW environment, you must ultimately “mix” these to create a fi nal version of the music. Typically, mixing involves setting the level and panning position; and considering the tonality, dynamics, ambience, and other eff ects of each sepa- rately recorded element. A new stereo fi le that incorporates all of these elements is created and used for burning to CD, posting to the Internet, and so on. You might sometimes be creating a 5.1 surround mix, or even 7.1 surround, or some other confi guration—but stereo is still the predominant delivery format. R emixing used to simply mean doing the mix again; and because of the power of the DAW, recordists fi nd themselves redoing mixes more frequently than ever before. But the word remix has come to have its own, separate mean- 172 ing. Remixes are reimaginations of a piece of music, oft en using completely new elements and eliminating other elements that were used in the initial mixed version. Remixes for specifi c functions—such as club play—are common, but remixes simply as creative exercises have also found a signifi cant role in popular music. Beyond remixing are mashups and other newfound ways of recombin- ing music elements. All of these are extensions of the basic mixing process, and mixing is what I cover here.

Th e mixing environment: Th e room and playback system

I have already discussed room acoustics and monitoring systems at the begin- ning of chapter 2, and that information pertains to the mix environment as well. In fact, control-room and speaker considerations that are important to record- ing become even more critical in the mixing process. I’ve made recordings in some pretty funky listening environments, and sometimes I simply rely on ex- perience: “It doesn’t sound very good in here—and I don’t trust these speakers or this control room—but it sounds good in the recording room and I know the mics are working properly and positioned correctly so I’m going to assume that the recording sounds good.” Th ese kinds of situations have worked out for me with recordings, but they won’t work out when it comes to mixing. A sonic environment and playback system that you can trust is critical to mixing. Near-fi eld monitors reduce the eff ects of room acoustics, but they do not eliminate them. Your room and your speakers must be reasonably neutral. Th is means that frequency buildup and refl ections should be kept to a mini- mum through good room acoustic management, and your speakers need to be studio monitors that have at least reasonably fl at response across the spectrum. All speakers have diff erent qualities, and no speakers are truly fl at, so fi nding the right mixing speaker is usually a process. Research at various discussion group sites, such as, can be useful and give you a lot of ideas, Mixing about available studio monitors. You may have access to a recording-equipment supply store that has monitors set up that you can audition, though those envi- ronments may be quite diff erent from your setup, so the situation isn’t ideal. Of course, budget will probably be a major factor, as well. Once you’ve settled on a good candidate through research, and, if possible, some auditioning or studio experience with a particular speaker model, try to buy them from a dealer that will allow returns, so that when you get them to your studio/home studio you have an option if they just don’t seem right in your environment. U ltimately, a good-sounding room and accurate speakers need to be com- bined with experience for you to create reliably good mixes. Getting used to your room and your speakers requires some time and some trial and error. Learning to listen as a mixer must be supported by confi dence in what you’re listening to, so don’t shortchange your environment or your playback system. Th ere is more on making your mixes translate to all listening environments at the end of this chapter. 173

How mixing relates to composing, arranging,

and performing B ecause mixing involves the ultimate way that a musical recording is going to sound, it shares many of the functions of composition (or songwriting), music arranging, and musical performance. In some fundamental ways, it is impos- sible to separate the mixing process from the writing, arranging, and perform- ing processes; they all interact to form our ultimate experience of the musical recording. As a result, it isn’t possible to completely distinguish the eff ects of the mixing process from these other musical activities. A beautifully composed, ar- ranged, and performed piece of music will be much easier to mix than one with awkward composing, poor arranging, or inconsistent performances. O ne example is mixing a song in which there are two diff erent guitar parts and a piano part, all played in the same register as the vocal melody. No matter how you mix these elements, they are going to be competing for the same fre- quencies. Level and panning strategies—key to mixing—can create some sense of separation between these parts, but nothing a mixer can do will completely solve the overloaded frequency range caused by the arrangement. Th e situation is similar with a performance that feels uncomfortable rhythmically or out of tune. Performance problems such as these will always make the mix sound un- fi nished. And a composition in which the melody jumps awkwardly from one theme to another can never sound settled, under any mixing strategy. Th e above situations are true except to the extent that the mixer actually alters the composition, arrangement, or performance. As discussed in the chap- ter on editing, recordists have powerful tools for altering all the elements of a recording, and more pronounced alterations have become common in contem- porary recording work. Arrangements, performances, and even compositions, THE ART OF DIGITAL AUDIO RECORDING are routinely altered as part of the recording/editing/mixing process. We can alter the rhythm and pitch of performances, we can mute or move elements, and we can reorder pieces to change arrangements and compositions. Two questions arise: Who has the authority to undertake such transforma- tions? and When are they to be done? Th ere is no simple answer to either ques- tion. Th e authority may be centralized in one person—artist, performer, pro- ducer, recordist, or a combination of these—but it is more likely spread among all of them, without clear dividing lines. Good collaborative relationships allow ideas that change compositions, arrangements, or performances and can be suggested at any time during the process. Th ey can be tried and then accepted or rejected by a consensus, though one person will need to have the fi nal say if there is disagreement. And, while there’s oft en an immediate consensus about a change—that is, all agree, “Th at sounds better!”—there can be healthy, and even frequent, disagreement without harming the working relationship if all are 174 working with the spirit of creative experimentation. Th e second question—“When are they to be done?”—is generally an- swered as “At any point in the entire process.” Th is means that editing, fi xing, moving, and so on might get done right at the same time as things get recorded, or in dedicated editing/fi xing sessions, or during mixing. Which brings me back to the question, What is mixing? I recently received an e-mail asking if I was interested in a mixing project. Th e inquiry said that they have budgeted a certain amount to mix fi ve songs. Th e budget works for me, if—and this is a big if—by “mixing” they are not expecting any editing or fi xing as well. If I’m working on an hourly basis, or on my own, then the task of mixing may well get blurred with those of editing and fi xing. Even though composing, arrang- ing, and performing matters may have a strong relationship to mixing, they are separate from the fundamental task of mixing.

Mixing tools

Mixing tools, beyond the room and the playback system discussed previously, encompass a broad world of systems and processors. Th e equipment starts, of course, with your computer and your particular DAW, though every major DAW system is well equipped to handle the basics of mixing. Before I get too far with mixing tools, however, I have to consider one of the major ongoing debates in regard to mixing: should you mix entirely within your DAW, using only digi- tal processing available within the computer (mixing in the box), or should you supplement the DAW with analog equipment (mixing out of the box)? Mixing in or out of the box? Th e notion of mixing “in the box” is simple: everything you do as a part of your mix occurs within your DAW (the computer is the box). Mixing “out of the box” can take myriad forms, from using just one or two analog processors to, Mixing supplementing a mix that’s done primarily in the box, to mixing with an analog console and all analog processing gear (oft en with external digital processing gear, as well). In this book, I limit the discussion to mixing within the box. I’m not arguing that this is the best way to be mixing, but this approach has some distinct advantages in regard to budget and work fl ow, and it has become in- creasingly common at all levels of production, including big-budget projects. B eyond budget, the advantages to mixing in the box include ease of setup and outstanding automation and recall systems (discussed later in this chapter). Th e primary disadvantage is that you eliminate your access to analog process- ing gear, which some people prefer. Some people also believe that analog sum- ming (combining of tracks) is superior to the digital summing within a DAW. While there continues to be considerable debate about the relative merits of analog and digital processing and summing, everyone agrees that the digital options have been tremendously improved in the last several years and there are more digital options than ever before. Th ere’s no simple answer, but the fact 175 is that a great many projects, including some high-profi le projects, are being mixed in the box—including several of my own Grammy-nominated projects. Processing gear (plug-ins) Along with setting levels and pan positions, it is audio processing that occu- pies most of the recordist’s attention in the mixing process. Th e tools of DSP (digital signal processing) include the EQs, dynamics, and ambience proces- sors discussed in the second half of chapter 2. Th ese tools play a critical role in mixing, as you will see in the following section, when I discuss building a mix. Each DAW comes with plug-in versions of most of these tools, but there are an enormous number of third-party developers that supply additional tools for every DAW. Some supply capabilities that are not included with the DAW and some supply higher quality versions of the same basic tools. Obtaining these plug-ins can be a near endless process of acquisition (and expense!). What do you need to mix eff ectively? As you might expect, there is no simple answer to that question. Th eoretically, you don’t need anything more than the tools that come with your DAW. More important than any plug-in is the ear and creative vision that drive the mixing process. Th at said, not having some high-quality processors of nearly every kind can be a distinct disadvantage in trying to create satisfying mixes. I remember very well the fi rst time I got access to an SSL mixer (one of the highest quality analog consoles). As I was working, I started thinking, Th is is why my drums have never sounded the way I want them to—I didn’t have the necessary tools! It’s true that certain qualities to sound are just not available un- less you have the right tools—with either the right capabilities, or the right level of quality, or both. So, again, what do you need to mix eff ectively? As much gear as you have the ear and the experience to use eff ectively—and can reasonably aff ord! It’s not, THE ART OF DIGITAL AUDIO RECORDING always easy to know what that means—and sometimes gaining the ear and the experience fi rst requires having access to the tools in order to learn—but oft en your system will grow and develop naturally with your experience. And the income from your work will provide the opportunity for growth; I still use part of the income from big projects to expand my processing arsenal. Th ere is more about specifi c tools in the following section on building a mix. 6.2 Building a Mix Building a mix is an apt metaphor for the mixing process because mixing is a form of construction. Really, it’s a reconstruction, taking all of the recordings that have already been constructed for the particular piece of music to be mixed and reconstructing them into their fi nal form. Th e following addresses both the strategies and the processes involved in building a mix. 176 While mixing provides endless opportunities for creativity, there needs to be a balance between art and artifi ce. Th e art of mixing encompasses all mixing strategies, both artful and artifi cial. According to the dictionary, artifi ce is “an artful strategy,” but it is also sometimes understood to be a trick. A rtifi cial in mixing may refer to sounds and eff ects that aren’t natural, that wouldn’t occur in natural acoustic environments. Th e art of mixing must employ artifi ce, but it does so somewhere on the continuum between artful strategies that employ only natural acoustical eff ects and those that defy natural acoustics and include any number of audio “tricks” that fall well outside anything possible in nature. I worked on one mixing project where the artist defi nitely wanted to limit my choices to “sounds found in nature”—a perfectly fi ne strategy for mixing a lot of music. On the other hand, some mixing requires a lot of “artifi cial” eff ects and unnatural sonic environments, and these can still sound very musical. Some sense of where your project is going to fall along this scale between art and ar- tifi ce is a valuable starting point for building your mix.

Approaches to listening and listening levels

H ow we listen is an important part of eff ectively building a mix. I have had mu- sicians tell me that they have trouble listening to the balance between frequen- cies (from the lows to the highs) because their ear keeps focusing on the mu- sical content. Some engineers miss musical relationships, like the interaction of counterpoint, because they’re used to concentrating on sound rather than musical ideas. A good mixer needs to be able to listen sonically and musically. Sometimes we need to focus our ear on the way things sound, ignoring musical relationships, and sometimes we need to consider the musical functions before we decide about sound and placement issues. Oft en we need to balance the sonic and the musical contents at the same time. I have dedicated a whole section of the fi nal chapter of this book to lis- tening levels during recording session—it’s an important topic that deserves, Mixing signifi cant attention. Much of what is covered in that section is applicable to the mixing process, but in addition to that material I want to emphasize two points in regard to listening levels while mixing. Th e fi rst is that listening at a variety of levels, from soft to loud, is a valuable part of referencing your mix. Second, ear fatigue is the enemy of mixing—it’s the enemy of all audio work, of course, but especially mixing because of the subtle nature of the critical relationships being manipulated. Referencing your mixes loud is valuable as an occasional part of the process, but most of your mixing should be done at moderate levels. Peak volume readings of about 85 dB SPL represents a good standard for much of your listening while mixing and will allow you to work long hours without ear fatigue (a decibel reader, available from Radio Shack and other electronic supply stores, is a good investment). I n regard to listening at various levels, you need to take into account the Fletcher-Munson curve (and its later refi nements that I discuss in section 2.5) that describes the way the ear’s ability to hear diff erent frequencies changes at 177 diff erent listening levels. Th is explains why it is just as important to not listen too quietly or too loudly when mixing. Loud listening will cause ear fatigue, but quiet listening will cause the ear to misjudge the relationships in the fre- quency spectrum because you don’t hear high or low frequencies as well during low-level listening. But for this same reason, low-level listening can cue you to volume relationships that may be missed during moderate and high-level listening. Th e elements that you want in the front of your mix (vocals or solos, for example) should really pop out during low-level listening; if they don’t, they might not be loud enough in the mix or they might require further EQ work. Subtle background sounds, such as reverbs, are sometimes easier to judge with pretty loud listening. Aft er working at a moderate volume for a while, give yourself a short period of loud listening to reveal some relationships that were not so obvious before, such as an excess of delay or reverb. Use your listening level to monitor various elements of your mix: moderate-level mixing for the general balance of all mix elements, low-level mixing for the level relationships between primary elements, and (relatively) high-level listening to check the re- lationship of quiet elements within your mix.

Preparing your fi le: Tracks, grouping, and routing

W hen you are ready to mix and all (or most) of the recording and editing is done, it is worthwhile spending a bit of time preparing your fi le for mixing. Part of organizing your fi les means creating a logical layout for your tracks. Oft en, during the recording and editing process, tracks get created or moved around to serve whatever is being done at the moment. A guitar track might get put next to the kick-drum track to check timing and a vocal track might get moved next to the piano track to make critical monitoring changes during a take. When mixing, it’s nice to have the tracks laid out in some logical manner. For a typi-, THE ART OF DIGITAL AUDIO RECORDING cal band recording, I organize my tracks as follows, moving from left to right on the mixer: drums, percussion, bass, guitars, keyboards, vocals, background vocals. Of course, your recording may have more, less, or other elements, but you simply make a progression that makes sense to you. P art of organizing your tracks may involve getting rid of tracks that you’re not using. Many DAWs allow you to “hide” tracks so that they’re not visible in the mixer or edit views, but still available if you change your mind later and want to include them in your mix. You should also be able to disable or deacti- vate those tracks so that they are not using any computer resources while they’re on hold. Once you have an organized track list that contains only tracks you’re planning to use in your mix, you’re ready to consider some essential grouping and routing options. Channel groups 178 It’s likely that during the recording process you created some channel groups and possibly subgroups to make working easier. We encountered the notion of grouping in the chapter on editing. A group is simply a means of linking chan- nels together so that you can control all of the tracks as a unit. Editing, chang- ing the volume, or copying and pasting multiple parts are much simpler and more effi cient when done as a group. For example, if you have multiple drum tracks and you haven’t already made a drum group, you will certainly want one for mixing. In general, groups are very valuable in the mixing process, and you will want to go through your tracks and make groups for all the basic relation- ships: a drum group, a percussion group, a background vocal group, and so on, depending on the elements in your recording. You can disable any group while you make changes to one or more of the individual elements separately and then re-enable the group for overall group changes. You may have groups within groups, smaller groups that are also a part of a larger group. A typical example would be the tom-tom tracks group, or the overhead tracks group within the larger drum group. DAWs have some means of showing groups within groups: in Pro Tools, the larger group is categorized by letter (a group, b group, etc.) and when a smaller group appear within a larger group, and the larger group is activated, member tracks from the smaller group are identifi ed with a capital letter and member tracks that aren’t in any other groups are identifi ed with a lowercase letter. In the screenshot on the following page, the drum group is the a group. Because the toms and the OH (overheads) are also grouped separately, they are shown with a capital A while the tracks not in another group, such as the kick and snare, are shown with a lowercase a ., Mixing SCREENSHOT 6.1 Multiple groups within Pro Tools S ubmixes, subgroups Th e terminology is not consistent when it comes to making subgroups or sub- mixes, but the practice is very common. By routing multiple tracks to an aux track (typically a stereo aux to maintain the stereo position of the individual tracks), you can use the aux track to apply processing and automation to a group of multiple tracks. In the example on the following page, six background vocal tracks have been routed, using buss 5–6 to a stereo aux. Th is submix or subgroup channel is being used to apply EQ and compression to all six tracks at once, and to send them all to a reverb (using buss 7–8), as well. You can also automate the level of the tracks together. Th is can save on processing power, as well as making your work go quicker., THE ART OF DIGITAL AUDIO RECORDING SCREENSHOT 6.2 M ultiple tracks routed to a stereo aux with processing M aster fader Th ere can be only one true master fader in a session, but the terminology can be confusing because sometimes what are technically sub-master faders may be identifi ed as master faders. All tracks feed the master fader, and generally the stereo outputs of the master fader are the pair that feed the playback system (amplifi er and speakers). Th e master fader can be used for stereo buss processing. If you place a plug-in on the master fader, that DSP will be applied to your entire mix. Th is can be useful for overall buss compression, EQ, or other eff ects. (Note: the stereo feed from your DAW is sometimes referred as the “2 buss” or just the “buss.”), Mixing Th ere is a problem with fade-outs when using dynamics processing (com- pressors, limiters, expanders, etc.) on your master fader. Because the processors are fed by the master output, the processing is aff ected when creating an overall fade (such as the fade-out at the end of a song). As all the tracks fade, the send to the dynamics processor will drop below the processing threshold. Although the track is fading, the music’s intensity is not meant to be aff ected, so you don’t want the dynamics processor to stop doing its work. Th e way to avoid this is to set up a master auxiliary track—you may want to label this “SUB,” as it is a master submix. If you feed all your tracks to the SUB using a stereo buss, and then feed the SUB to the master fader, you can place your buss processors on the SUB; and then, when you create a fade on the master fader, the overall mix will continue to be processed (via the SUB plug-ins) as the track fades. SCREENSHOT 6.3 A ll channels routed to a sub-master and then to the master fader S CREENSHOT 6.4 A master fader fade-out after the sub-master processing

Mixing: Basic operations

A s with any construction project, there are many possible routes to get from the beginning to the fi nal form; but because eff ective mixing generally involves a whole series of steps and resteps, the exact sequence of events is not necessar- ily critical. Mixing involves drilling down to great detail while at the same time it requires a consistent focus on the overall sound being created. Th e “micro”, THE ART OF DIGITAL AUDIO RECORDING is managing every part of each track’s mixing parameters, including the level, panning, EQ, dynamics processing, eff ects, room ambiences, reverbs, short de- lays, and long delays, that may combine to create the sound of each element. At the same time, you must not lose focus on the “macro,” which involves consid- ering each individual sound in the context of every other sound that is part of the mix. I n this section, I consider each of these mix parameters as part of building a mix. Both micro and macro points of view are included in the discussion, as well as refl ections on the working process. All of these elements have already been discussed as part of our general understanding of the recording process, but here the focus is on the mix, where greater detail and a more creative point of view are required . Th e goal of creative mixing is to fi nd the right sound and the right place for each element to best serve the creative vision. Many factors com- bine to give each element its proper sound and place. L evel and gain structure (balance) Th e number one task of mixing is to establish the relative levels of all the ele- ments in your mix—which are louder and which are quieter. However, as you begin to mix, you also need to be aware of your overall gain structure. Once all the elements are in play, you will want your overall gain—your two-buss level as refl ected on the meters of your master fader—to be at a comfortable level. Too much gain will overload the system and cause distortion, and too little gain decreases resolution and control. Y ou will want to start by playing all your tracks together, setting a quick balance among elements, to see what your overall gain structure looks like, and to imagine a creative strategy for how you will eventually position all the ele- ments. You can adjust all the tracks together to set your overall gain, allowing a fair amount of headroom, as levels are likely to increase with the addition of EQ. At the loudest part of your rough mix, all of your tracks together shouldn’t peak over -6 dB on your master fader. A n important part of creative mixing is imagining the relative levels be- tween elements in terms of f oreground and background. Unless you have very few elements in your mix, it isn’t possible for everything to be in the foreground. How you treat elements in terms of processing will be aff ected by their position relative to foreground and background. (You may remember that I discussed recording techniques in these same terms—how you choose to record elements may also be infl uenced by their ultimate position as foreground or background in the mix.) As you begin to mix, the fi rst element you consider exists in a kind of vac- uum, as you have no other elements to balance it against. Having established an overall gain structure means you can start with the fi rst element at the level it is already set, and that becomes the baseline as you add elements. In a traditional band recording, the fi rst element mixers consider is oft en the drums—and the, Mixing fi rst element from the drums is oft en the kick drum—but some mixers prefer to start with the bass. Diff erent mixers take diff erent approaches, but because you will be returning many times to each element in a mix, it isn’t critical which ele- ment you choose to start with. I return to the question of how you might order the introduction of elements into your mix, and ultimately how you might settle on relative levels, aft er considering the other major parts of the mixing process. P anning C reative use of panning is one of the most frequently underutilized tools in the mixer’s toolbox. It’s useful to remember that the word p anning c omes from p an- orama, which refers to an unobstructed and wide view; and creating a wide and elegant aural panorama is one goal of all creative mixing. Th e complete panning spectrum runs from hard left to hard right, and the creative mixer will make the most of this entire fi eld. I covered the basics of panning in chapter 2, so here I focus on panning 183 strategies for mixing. Th e fi rst strategy is to have a strategy—that is, you want an overall plan for panning elements before you start addressing individual tracks. Certain panning approaches may remain constant. Drums may be panned ac- cording to their physical setup, with the kick and snare tracks centered, the hi- hat track to one side, the tom-tom tracks spread from one side to the other de- pending on the number of toms, and the overhead mics split in hard left /right stereo. Drum panning can adhere to either the drummer’s perspective or the audience’s perspective and either is acceptable as long as it is consistent. (Don’t pan the hi-hat based on the drummer’s perspective and the tom-toms based on the audience’s perspective.) I was a drummer for many years, so I usually pan the drums using the drummer’s perspective because that’s what sounds most natural to me, but if I’m mixing a live recording, I’ll use the audience’s perspec- tive because that’s the way the live audience was hearing the drums. Bass and lead vocals are usually center-panned along with the kick and snare (though it’s perfectly fi ne to stray from this convention if you fi nd a compelling reason to do so). Beyond these generally accepted practices, panning is wide open to cre- ative approaches. Getting the macro of panning established for your mix means considering each element in the mix and placing it in the panning spectrum. You might start with four basic positions (seven total positions)—center, soft left or right, medium left or right, and hard left or right—and place every ele- ment in one of these positions. Your decision will be based on the number of elements, their relationship, and your vision of how they will best fi t together across the stereo fi eld. For example, a tambourine track may belong in any one of these seven places, but the part it plays (simple or complex), its relationship to the position of the hi-hat, its interaction with other rhythmic elements such as the snare drum or a rhythm guitar, its relationship to other high-frequency elements such as a shaker, its history in the style of music, and so on might all, THE ART OF DIGITAL AUDIO RECORDING aff ect your decision. Four tracks of background vocals may be panned in a mul- titude of ways, including spreading them evenly left to right, spreading them across either the left or right panning spectrum, and lumping them together at one spot in the panning spectrum. Your decision may be infl uenced by the relationship of the four parts (which are high and which are low), by the rela- tionship of the parts to the lead vocal, by the existence of other elements in the track that may have similar function such as a horn section, and so on. Th ere are an enormous number of considerations that you might take into account in any panning strategy. Th ere is no substitute for experimentation and creative thinking while making panning decisions, but here are some further guidelines. 1. D on’t be afraid to abandon an initial panning strategy and start again from scratch. 184 2. Aft er you’ve applied your basic strategy for panning all the ele- ments, continue to experiment with slight changes in positioning to fi nd the best possible position for each element. 3. Use the entire panning spectrum. If there are very many elements in your mix, it is almost always the case that one element should be panned hard left and one element hard right. Don’t leave the far ends of the panning spectrum unexplored. WHAT NOT TO DO Panning stereo tracks Just because something was recorded (or sampled) in stereo, that doesn’t mean that you have to use its full stereo capability in your mix. When you create a stereo track, it defaults to placing the two panning controls set to hard left and hard right. Sometimes you will want to leave them set this way, but often you will want to adjust the stereo balance within a stereo recording. For example, even though the piano is recorded in stereo (using two microphones), there may be a lot of elements in your mix and the piano will be heard better if it occupies a smaller piece of the stereo image and doesn’t compete across the entire stereo spectrum. You may want to set the one panning control soft right and the other medium right—keeping the piano on the right side but allowing it to be spread a bit across the spectrum on the right. Or instead, you might want to set both panning controls to hard left and let the piano have its own place at the far left end of the spectrum. The two tracks are still providing complementary information to fi ll out the piano sound, even if they are panned to the same place, making them sound like a mono recording. Too many elements spread out in wide stereo will often make a mix sound indistinct and congested., Mixing 4 . Remember that altering panning changes volume. Th ere is a power curve to panning controls, which means that sounds in- crease in volume as they move farther left or right (the diff erence between center position and far left or right is between 3 and 6 dB, depending on the system). Consult your DAW manual, but your ear is best source for setting volume regardless of specs. Auto-panning is another powerful panning tool that can be eff ective (or distracting) and has become much more versatile in the DAW world than it was in the analog world. Auto-panning refers to “automatic” movement in pan position as the music plays. I will explore auto-panning in the following section on automation. Equalization (frequency range) As previously discussed in chapter 2, EQ represents the most powerful and 185 important of all signal-processing gear. EQ is an essential part of the mixing process. However, I am reminded of a discussion I had with a colleague shortly aft er having my fi rst experiences mixing on an SSL console. He said, “Th at SSL EQ is powerful and can be a great tool, but it can also destroy a mix.” Indeed, EQ can be your best friend or your worst enemy. Used wisely, it can transform mixes into works with greater clarity and impact; and used poorly, it can make mixes sink in a morass of shrillness and/or mud. Th ere are two essential considerations to keep in mind as you EQ elements for your mixes. Th e fi rst is what kind of frequency shaping with EQ is going to make this element s ound best, enhancing the sound of the recording. Th e sec- ond is what kind of frequency shaping with EQ is going to make this sound fi t best with all the other elements in my mix. Typically, these two considerations will have some things in common and others in confl ict. Your job as a mixer is to make the best compromise between “sounds best” and “fi ts best.” Sometimes these two things are really completely complementary, but that is usually only the case in mixes involving very few elements. On a solo piano recording, you can ignore “fi ts best” and only consider “sounds best,” but on a mix involving 15 diff erent instruments, there will need to be a lot of “fi ts best” considerations that override “sounds best.” A typical example of the “sounds best” versus “fi ts best” EQ-ing confl ict would be in regard to an acoustic guitar recording. Acoustic guitar is a full-fre- quency instrument that oft en has very rich overtones throughout the frequency range. A well-recorded acoustic guitar may sound best with no EQ at all, or with a slight amount of high-midrange frequency, high-frequency, and/or low- frequency boost to accentuate the overtones and make the instrument speak, sparkle, and resonate most fully. Th e fullness of an acoustic guitar is wonder- ful for solo guitar or in small ensembles, but it is oft en problematic when the instrument needs to fi t in with a larger group or in a rock-type setting. Th e, THE ART OF DIGITAL AUDIO RECORDING rich low end of the guitar tends to get muddied up with the bass and other low-frequency sounds. In a mix with drums, bass, electric guitar, vocals, and possibly many other elements, that full-frequency acoustic recording takes up way too much space. In a dense mix, it is likely that you will want to severely cut frequencies from the acoustic guitar, especially in the lows and low-mids, and you may want to accentuate the higher frequencies beyond your normal “sounds best” sensibility in order to get the acoustic to cut through in the mix— the guitar must fi t in and making its presence known without competing with too many frequencies from other elements (panning plays an important role in this equation as well, as discussed above). One might well ask, “How do I know what ‘sounds best’ and what ‘fi ts best’?” Here, there is no easy answer; in fact, there is no one answer or best answer. Certainly, there are some general criteria that most (but not all) re- cordists would agree on, but these are themselves somewhat vague. “Sounds 186 best” is rich in pleasing and musical overtones. “Sounds best” is well balanced through all the frequency ranges that are appropriate to that particular instru- ment. “Sounds best” is warm and present. “Fits best” is focused on the frequen- cies where there is the most space for this particular element. “Fits best” sounds like it belongs in this environment. “Fits best” sits in a mix with a clear identity and place. While many might agree on these descriptions, exactly what kind of EQ-ing might be employed to achieve them could diff er pretty radically from one recordist to another. I have been surprised by the proliferation of “presets” for EQ-ing various instruments that are found as a part of many EQ plug-ins now. You may even get (or you may be asked to buy) EQ presets from well-known recordists for certain EQ plug-ins. I fi nd this odd because each particular recording of any given instrument, and each particular use of that instrument within a particu- lar recording, is best served by an individual approach to EQ-ing that element. Th at said, it is true that approaches to EQ-ing certain instruments may be rela- tively consistent within a specifi c genre of music, so perhaps these presets are useful in pointing people in the right direction. Perhaps. But they also might have a negative eff ect in making people think that there is a “right way” to EQ a snare drum or an acoustic guitar, or that all they need to do is apply the preset EQ for each element and their mix will be EQ’d in the optimal way. My advice is, sure, go ahead and explore the presets, but u se your ear and don’t be afraid to make changes to the preset or of even taking a completely diff erent approach. I have given some advice on using EQ in chapter 2, but you will not fi nd any packaged formulas for EQ-ing here. You must explore on your own. Dynamics processing Dynamics processing plays a major role in mixing. Most of the time, the em- phasis is on compression and limiting, as opposed to expanding and gating. Th is is evident in the section on the basics of dynamics processing in chapter 2,, Mixing where most of the discussion focuses on compression and limiting. As with EQ, dynamics processing can be your friend or your enemy. Eff ective use of dynam- ics requires the technical mastery of the tools and the development of listening skills in the service of your creativity. C ompression has two distinct functions in mixing. One is the subtle con- trol of volume dynamics that evens out performances and helps them retain their presence throughout a mix. Th is fi rst function of compression is generally pretty transparent; the goal is for the dynamics control to leave the sound as unaff ected as possible in any way other than shrinking the dynamic range and thereby leveling out the performance. Th e second is the use of compression to create a variety of obvious eff ects. Th e most noteworthy eff ect from certain kinds of compression is the addition of impact through a concentration of the audio energy. Th is is most frequently heard on drum tracks in many genres of contemporary music. U sing compression for the subtle control of dynamics can be an enormous 187 aid in getting elements to sit comfortably in mixes. Featured elements such as lead vocals and bass are particularly susceptible to problems from too great a dynamic range. Th e basic argument for using compressors is laid out in section 2.6, where I introduced the functions and operations of dynamics processors; elements that have less dynamic range can be heard more consistently when competing with a lot of other sounds. As a general rule, the greater the number of elements in a recording, the more help can come from compressing them. In many contemporary recordings most elements are compressed, and there is fre- quently overall compression applied to elements in subgroups (such as drums), as well. Th ere may also be additional compression on the overall mix. One well- known producer has said, “Compression is the sound of rock and roll.” My overall creative vision for the sound of the mix, along with the density of the mix and the relative position of each element, dictates how aggressively I use compression in any given mix. In relatively spacious recordings without a lot of elements, I rarely go above a 3:1 ratio and 4 or 5 dB of compression on the loudest sounds. On a dense mix, I might use ratios as high as 6:1 or higher and hit 7 or 8 dB of compression at the maximum levels. In dense mixes, I might use a bit of limiting as well as compression on some elements to tame the peak levels. I use a bit of buss compression (overall compression on the entire mix) on most mixes as well. Gentle buss compression acts as a kind of glue that helps blend all the tracks together, although too much glue can make mush of the tracks. Aggressive buss compression can be used as an eff ect—to add impact to mixes. Th e diff erence between gentle and aggressive buss compression has to do with ratio and threshold—higher ratios and lower thresholds ramp up the aggression—but processors also have characteristics that are the result of many other elements in their design. Many compressor plug-ins have settings that simulate a wide range of compressor types from their analog antecedents. Mix-, THE ART OF DIGITAL AUDIO RECORDING ing is the place to explore all kinds of compressor types and functions, from the most gentle and transparent to those with the most “personality.” Th ere is one type of control over dynamic range that is not really a part of the mixing process and that is brick-wall limiting—it belongs to the mastering process covered in the next chapter. But because this process has such a pro- found eff ect on the sound of mixes, and because it will be applied moderately to heavily most of the time in mastering, you need to integrate it into your mixing practice as well. I cover the basic idea of how to integrate this into your mixing process in the fi nal section of this chapter on delivering mixes and I cover it more thoroughly in the following chapter on mastering.

Mixing: Creating ambience and dimension

Certain mixing processors can add ambience and dimension to your recording. 188 Th ese processors are the delays and reverbs covered under FX (“eff ects”) in sec- tion 2.7. Most FX processing of this type is done as part of the mixing process, although you might use delays as an integral part of a sound when recording (chorusing on a guitar, for example), and you might use some reverb for moni- toring your vocal (meaning you add reverb to the vocal for listening purposes, but you do not actually record that reverb as a part of the vocal recording). Waiting until the mix to add these kinds of eff ects allows you to create a unifi ed a mbient environment for your fi nal audio presentation. Th e combina- tion of delays and reverbs creates a kind of delay pool that, though made up of individual eff ects on each element, also combines to create a sound stage that you want to consider as a whole. As you build your sonic landscape (or sound stage), both musical and technical considerations come into play. You may wish to construct a naturalistic environment—one that sounds true to a real-world setting, such as a nightclub or a concert hall. Contemporary popular music tends more toward unnatural environments, in that many diff erent kinds of ambience are used within one mix, even though it wouldn’t be physically possible for all the elements to be in those environments at the same time. Some mixes combine many diff erent eff ects but limit them to the kinds of reverbs and delays that are found in natural environments such as rooms, theaters, and concert halls. Other mixes incorporate unnatural delay eff ects, such as “gated reverbs” with abrupt cut-off s, and ping-ponging delays that bounce back and forth between left and right speakers. Whatever your approach to the environ- ment that you are creating, it is the construction of a sound stage that is one of the most creative parts of the entire process of mixing. U sing delays in mixing I covered the basics of short, medium, and long delays in section 2.7. Here, I discuss some of the fundamental ways that they are used to enhance elements during the mix. Short delays can be used to thicken, to add interest, and to, Mixing expand the stereo fi eld of a sound. Medium delays can open up a sound while reinforcing the rhythm. Long delays can also reinforce rhythm, they can call attention to a sound, and they can be used for special eff ects. Short delays are oft en used to thicken sounds. Th e classic “chorus” eff ect thickens elements in the way that vocal choruses are used to create thickness with many voices. Sometimes the short delays that create the chorusing eff ect are used without the modulation that is characteristic of the chorusing eff ect. Th is is generally referred to as a doubling eff ect, and it may be created with straight delays or it may be created with micro-pitch shift ing to further thicken the sound without modulation (usually shift ed in pitch either up and/or down 6 to 9 cents). Th ese kinds of eff ects can be used on virtually any kind of audio, though they are generally not eff ective on short drum and percussion sounds. Th ey can sound good on rhythm and lead instruments, as well as vocals. Of course, thickening is a two-edge sword—it can enhance and add interest to al- most any element, but too much of it in a mix makes the mix too thick and 189 blurs the sound. You must pick the elements that will benefi t the most from short delay eff ects, and this will vary from mix to mix and in relation to diff er- ent genres. For example, punk rock may not call for too much thickening with short delays (though fl anging on the vocal can be very eff ective and appropriate to the genre), whereas electronica might benefi t from quite a lot of short delay eff ects, creating a wall of thick, lush ambience. S hort delays can also be used to spread elements across the stereo fi eld. If a sound is split evenly between right and left , it will sound center. It is mono if there is no diff erence between the sound in the left and the right channels. If you introduce a short delay on either side, the sound will suddenly acquire a stereo spread (anywhere from 3 to 30 ms delay would be typical). Although that application can be useful, it is somewhat artifi cial sounding. More subtle variations on using short delays involve less radical panning options, such as having the original signal soft right and the slightly delayed signal mid-right, just spreading the sound slightly across the right side. Two or more short delays, with slightly diff erent delay times and panning positions, can yield endless pos- sibilities for thickening and spreading sound across the stereo spectrum. O ne classic short-delay application involves two delayed and pitch-shift ed signals—perhaps one delayed 15 and one delayed 25 ms, with one pitch-shift ed up 7 cents and one pitch-shift ed down 7 cents, and then split hard left and hard right. Th e original, unaff ected signal is centered. Th e two delayed and eff ected signals can be pretty quiet and still provide thickening and stereo spread to the original signal. You can also collapse the two signals a bit (bring them in from hard left and right to closer to center) if you want a less audible eff ect. Th is type of eff ect is used fairly commonly on vocals, lead guitar, and other upfront sounds. M edium delays provide a sense of space by simulating medium to large environments. Medium delays in the 125 to 175 ms range are oft en referred to, THE ART OF DIGITAL AUDIO RECORDING as slapback. A very audible version of this eff ect can be heard on many Elvis Presley recordings, and this eff ect has become somewhat identifi ed with his vocal sound. Th e advantage to slapback delays is that they provide a sense of space without the complex and potentially cloudy eff ect of reverb—though they may also be used in conjunction with reverb, as a further enhancement rather than as a replacement. Although they can be used to broaden the stereo spread, they are generally panned to the same location as the direct signal to provide a more subtle eff ect. You can use feedback (multiple repeats) on a slapback delay, but generally one slap provides a cleaner sound. Th e level of slapback delay can vary from very rather obvious (the Elvis eff ect) to rather subtle, where you can’t really hear the delay but you notice a change in the depth of the sound if the eff ect is removed. Y ou can set the delay time for a slapback eff ect by ear, but you might want to set it based on the tempo of the piece of music you are mixing. Using “in 190 time” delays—delay times that are based on the musical subdivisions of the tempo—generally support musical propulsion, while delays that are contrary to the beat can diminish the rhythmic energy. Most delay plug-ins have options for setting delay times based on beat divisions (quarter notes, eighth notes, etc.). Of course, the plug-in must “know” what the tempo is in order to do that, so you need to have the fi le referenced to the correct tempo. Even if your music wasn’t recorded to a set tempo, you can usually determine the approximate tempo using various tap tempo tools that allow you to tap on one key of your computer keyboard in time with the music and get a read-out of the tempo (check your DAW for this “tap tempo” function). L ong delays can really open up a sound and can also be used for all kinds of special eff ects. It is especially important to use musical timing when setting long delays, as they can really confuse the rhythm if they are not in time. When used as an eff ect to suggest a very large space, it is typical to use some feedback, simulating the characteristics of sound bouncing back and forth in a large space (15 to 30% would be a typical range for feedback). As with slapback eff ects, nor- mally the long delay would be panned to the same position as the direct signal. Th e volume of long delays can range from a subtle eff ect that is only audible at high volumes to an obvious repeating eff ect that is easily heard. With such a strong eff ect that is creating new and distinct rhythmic patterns, it is generally used rather sparsely. Long delays are commonly used for special eff ects such as obvious and audible repeats, stutters, and cascading sounds. Such eff ects can be playful and fun, and they can add elements that become integral parts of an arrangement. To some extent, however, the ease of editing in a DAW has replaced the need to use long delays to create some of these eff ects. We can copy and paste, using a musical grid, and create repeating eff ects that can be more easily controlled than those made with a delay. In any event, repeats occupy an important role in the creation and mixing of popular music., Mixing U sing reverbs in mixing R everbs are the principal tool used to create (or recreate) ambience when mix- ing. As more thoroughly explained in section 2.7, reverbs simulate or actually reproduce the eff ects of real-world environments. When used creatively, re- verbs provide tools for mixes of great depth and interest. When used without suffi cient planning and careful listening, reverbs can be a primary source of problems that produce a lack of clarity in mixes. F irst, keep in mind that reverbs cover the entire range of room acoustics, from closets to bathrooms, nightclubs, concert halls, and outdoor arenas. Most reverb plug-ins organize their sounds by type of space (rooms, clubs, theatres, concert halls, etc.) and by type of reverb simulation (plates, springs, chambers etc.). Th ere are two primary qualities to reverbs: length and timbre. Length is expressed in seconds and can run from .1 to 7 seconds or longer, but most reverbs are in the .3- to 3-second range. Timbre of reverbs ranges from warm (concert halls) to bright (plate simulations) with rooms, chambers, theaters, 191 and the like varying depending on the nature of each individual space. WHAT NOT TO DO Tweaking reverb presets With the huge range of reverbs available, it is usually possible to fi nd the ambience that you want without having to do much in the way of tweaking the sound. If the preset you choose doesn’t sound like what you’re looking for, don’t spend time trying to tweak it into shape. Go for another preset that gets you closer to what you want to use as a starting point. My number one rule for selecting reverbs is: Don’t use a longer reverb when a shorter reverb will do. Often, the goal of creating an ambience for an element is satisfi ed with a relatively short reverb, such as a room with a length from .3 to 1 second. These short reverbs create depth and interest without washing a lot of sound over an extended period of time. If you have a variety of room reverbs available, from small to large and from warm (wood rooms) to well-balanced (standard rooms), to bright (tile rooms), you can use these to create much of your overall ambiance pool. S hort reverbs (such as rooms) are oft en the best choice for any element that has very much rhythmic interest. Th e more rhythmically active a part is, the more it will get clouded by longer reverbs. Long-sustaining elements can be treated with longer reverbs to give them a lot of depth without smearing the sonic landscape as much, and sometimes a single lead instrument or voice can be enhanced by a longer reverb. When using multiple reverbs, the combined eff ect must be considered—another argument for using shorter reverbs., THE ART OF DIGITAL AUDIO RECORDING R everbs do not have the same distinct kind of timing quality as delays have because they decay slowly, but tempo should be a consideration in reverb selection. You can use tempo timing as you do with delays to set certain kinds of reverbs. Gated reverbs (and some rooms) have a pretty steep decay, and you can set their length to a quarter note or a half note to good eff ect. Because of the more typical slow decay, timing most reverbs is best done with the ear, listening to how the trailing off of the decay fi ts into the overall rhythm of the piece. As you adjust the length of the reverb you will begin to hear what length seems to allow the decay of the reverb to become a part of the rhythm. Th e longer the reverb, the more diffi cult it is to have it interact with the music rhythmically, but the less important it is because the slow decay of long reverbs will tend to blend the reverb in without disrupting the rhythm. I generally don’t share reverbs, which means that each element in the mix gets its own distinct reverb. I may use the same reverb plug-in multiple times, 192 but with diff erent settings that recall diff erent spaces and reverb types. Nonethe- less, I use reverbs on the send and return model, rather than on the direct chan- nel insert, because it allows for easier fi ne-tuning and for panning variations. In a typical mix I use diff erent reverb settings (rooms, halls, plates, etc.) on the drums, the percussion, the piano, the organ, the rhythm guitar, the lead guitar, the horn section, the background vocals, and the lead vocal. I oft en use more than one reverb on some elements, such as drums and lead vocals, and blend them. Of course, not every mix has all these elements, and some mixes have other ele- ments, but the principle is that each element may benefi t from its own ambience. Th e individual sound can be tweaked separately, and you will have the opportu- nity to create a more distinctive sound and place in the mix for each element. It might seem tempting to send many things, or everything, to the same reverb so they “sound like they’re all in the same room together,” but in doing so you may create a muddy ambience pool, with elements competing for the detailed refl ections that reverbs are capable of. In some instances, this may be the right approach—perhaps on a live recording—but in general we are able to create more distinctive and interesting recordings by combining many reverbs, each one suited and balanced for the each specifi c element. You may have to limit your number of reverb plug-ins because of limits to your computer’s DSP power, in which case you need to be creative to get the most out of your limited resources; but most contemporary computers have enough CPU power to drive as many plug-ins as you need, even for the most complex mixes. Panning reverb returns P anning reverb returns is an important part of creative reverb use. Th ere are three input/output confi gurations for implementing reverbs: mono in/mono out, mono in/stereo out, and stereo in/stereo out. Exploring these confi gura- tions, along with the more detailed possibilities with reverb panning, is an im- portant part of the mix process., Mixing M ono in/mono out reverbs are handy when you want to place the reverb return directly behind the direct signal in the panning scheme (e.g., guitar panned 37% left and mono reverb return panned 37% left ). You can also use these mono reverb returns to push the ambience farther to the edges of the pan- ning spectrum (e.g., guitar panned 75% left and mono reverb return panned 100% left ). Try to avoid too many instances of the most common confi guration— mono in/stereo out, with the stereo outputs (returns) split hard left and right. Th is spreads the reverb return across the whole panning spectrum, and more than a couple reverbs in this confi guration can blur a mix rather quickly. Rather than having the returns fully panned, you can use this confi guration to spread the reverb a bit over the spectrum (e.g., guitar panned 60% left and the two re- verb returns panned 40% left and 80% left ). You might spread the return even farther but still avoid using the entire spectrum (e.g., guitar panned 35% left and the two reverb returns panned 70% left and 20% right). Because the stereo 193 outputs oft en have considerable phase diff erences in order to create a spacious sound, it can create problems if they are panned too closely together or to the same position. For this reason, it’s best to use a mono in/mono out confi gura- tion or just one channel of the stereo return when a single point reverb return is desired. Reverb confi gurations that have stereo inputs use varying strategies for feeding those inputs to the reverberation algorithms and generating stereo re- turns. Many reverbs sum the inputs to mono at some stage in the processing, so that the return remains equal in both channels no matter what panning strategy is used to feed signal into the input. True stereo reverbs maintain the stereo position of the signal’s input in the reverb’s output. Th at means that if you feed a signal to the left input of the reverb only, then the reverb for that signal will be returned only on the left channel. T rue stereo reverbs can be very useful in mixes with multiple channels of one type of element. For example, if you have six tracks of background singers, you can feed them all to a stereo submix by using a stereo send to a stereo aux track, and then feed the stereo submix to a true stereo reverb. Th is will put the same reverb on all of the singers’ voices, while maintaining the panning posi- tion of each voice within the reverb return, helping to create a distinct position for each voice while blending them in the same reverb. Advanced techniques with delays and reverbs Th ere are many more advanced techniques for using delays and reverbs than I have space to cover here, but I will mention a few and encourage you to use your ear and imagination to fi nd more. To begin, you can combine any number of the techniques described above to create more complex ambiences. It would not be uncommon for a vocal to have some kind of short delay/doubling eff ect, a slap delay, a room reverb, and a hall reverb—all used rather subtly but com-, THE ART OF DIGITAL AUDIO RECORDING SCREENSHOT 6.5 Backing vocals routed to a stereo aux and then to a stereo reverb bined to create a complex ambience. A solo saxophone might have a long delay and a medium reverb. A lead guitar might have a slap delay and long reverb, or it might have a doubling eff ect, a long delay, and short reverb. Th ere are endless possibilities for combining eff ects. When combining delays and reverbs, you can apply your eff ects either in parallel or serial. In parallel means that each eff ect is independent of the other, and this is the most common confi guration when combining delays and re- verbs. Serial eff ects feed from one to the next. A typical serial usage might be a signal that is sent to a long delay and then the delay is sent to a reverb, so that the delayed signal is soft ened and spread by the reverb. As explained in section 2.7, delay and reverb eff ects are typically used in a send and return confi guration, with the send being post-fader so that the level, Mixing of the eff ect follows the level of the direct signal. By sending to an eff ect pre-fader, you can create unusual ambient results. Pre-fader sends to reverbs allow you to use the sound of the reverb only, without any of the direct signal, and that can create some eerie and unusual eff ects (screenshot 6.7). Th e output fader of the audio channel can be set to zero, but the signal still feeds the reverb because the send is set to pre-fader. By sending a direct signal to a long delay, and then sending from the delay pre- fader to a reverb (with the delay channel output set to zero), you can create a reverb that follows the direct signal aft er a long 195 delay. Th is same eff ect could be created using a long pre-delay setting on a reverb plug-in, but reverbs do not necessarily pro- vide pre-delay lengths that would be equiv- alent to a normal quarter note or longer. Y ou can duplicate a track in your DAW and then radically EQ, heavily limit, or otherwise process the duplicated track and send that processed sound to a reverb SCREENSHOT 6.6 or delay pre-fader. In this way, you create S erial routing from a delay unusual eff ects without having to use the to a reverb more radically altered source track as part of your mix. Th ese more extreme-sound- ing eff ects may be combined with the more normally processed direct sounds for subtle but unusual results. As you can tell from the above examples, the creative way of accessing and combining eff ects is limited only by your imagination. Exploring routing possibilities is a big part of accessing these more advanced processing tech- niques. Th e fl exibility of DAW routing, and the variety and easy access to so many DSP eff ects, provides tremendous opportunities for new approaches to creating sonic landscapes.

Mixing: Procedures

How do you proceed through the mixing process from beginning to end? Th ere is no standard answer to this question. Diff erent recordists apply diff erent pro- cedures; even the same recordist will use diff erent procedures in diff erent in- stances. Nonetheless, I off er some general advice on how you might eff ectively, THE ART OF DIGITAL AUDIO RECORDING move through the mixing process. I would expect you to adapt this to your own working process. Setting levels: Building and sustaining interest S etting levels for each individual element is the primary activity of mixing. Th e goal with levels is to build and sustain interest over the entire musical timeline. Eff ective setting of levels can be approached with a variety of techniques, but one general practice that I have found particularly useful is to build the primary elements of your mix fi rst, and then 196 add the parts that interact with those primary elements. In a typical rock mix, that would mean setting levels for drums, bass, and one or two rhythm instruments (guitars and/or keyboards) and then the lead vocal. You don’t have to work in this order, though record- ists oft en do. It’s much easier to get a proper relationship between the lead vocal and the rhythm section without other elements con- fusing the balance. By getting these primary instruments into reasonable balance, you S CREENSHOT 6.7 have a framework within which to add other C reating “reverb only” elements. Th e proper levels for lead guitar or effects by using pre-fader other solo instruments, background vocals, sends horns, and so on all need to be set in relation- ship to the lead vocal and the rhythm section. Of course, what is meant by “proper” is certainly subjective—but what is proper for you and the music you are mixing is aided by this procedure, regardless of how that may translate into the specifi c levels you set. A nother important tip for setting levels is to vary your playback volume. Th e ear processes sounds diff erent at diff erent volumes, as discussed in sec- tion 2.5, and therefore your mixes sound diff erent to you (and everyone else), depending on the volume that you are listening at. To properly balance levels, you need to evaluate your mixes while listening, from very quiet to quite loud. As previously noted, most work can be done eff ectively at moderate listening levels, but a quiet listen can be especially helpful in setting the most important level relationships. Th is is because your ear fi lters out much of the higher and lower frequencies at low volumes, revealing the fundamental level relationships among prominent elements. Sometimes you can miss a simple problem at nor-, Mixing mal listening level—for example, the snare drum is too quiet or too loud—and a quick check at a quiet level makes this obvious. On the other end of the spec- trum, you can miss details in low-level sounds when listening at moderate lev- els and a louder listen can reveal these details, such as murkiness in the reverb pool or clipping on an eff ects send, as well as extraneous noises or sounds (like quiet pops from edits that were poorly cross-faded). Headphone listening is also valuable for revealing low-level problems. But, a note of caution here: don’t revise basic mix decisions such as level relation- ships based on headphone listening. Despite the popularity of iPods and the prevalence of earbud listening, your best chance at getting mixes that will trans- late across all playback systems is through moderate listening levels on studio monitors that you are very familiar with. Th ere are too many variable listening possibilities—from the huge variety of home stereo listening environments and speaker setups, to cars, to computer playback systems, to TVs, to blasters, to headphones, to earbuds—to possibly check mixes in all environments. Use al- 197 ternative playback levels and systems to gain more information, but in the fi nal analysis you must trust your studio monitors at moderate listening levels. Th e three-dimensional mix O ne primary goal in mixing is to achieve the best possible three-dimensional mix. It is up to you to defi ne what is best, but one way to do that is to consider your mix as a three-dimensional object (variations on this approach have been used by others in describing mixing methodologies). Th e three dimensions are height, width, and depth. In mixing, the notion of height has two possible meanings. Th e obvious one is level. You can imagine the relative volume levels of each element as relative height relationships—the louder the element, the higher it is—and, as already noted, the fi rst job of mixing is setting the level for each element. Height, however, can also be considered in terms of frequency range. You can think of the frequency range on a vertical scale—ranging from lows to highs—with the higher frequencies viewed as higher in height. A proper height relationship might be considered to be a balance in the frequency ranges from low to high. Listening for balance throughout the frequency range is an important part of the mixing/listening process. Although you can use a spec- trum analyzer to check frequency balance, I recommend this only for gathering a very limited amount of information. An analyzer might reveal problems in areas that your speakers don’t reproduce well (very low or very high frequen- cies), but they might also lead you to make unwise decisions by showing fre- quency bulges or defi cits that are a natural part of the program material you are working on or the style of mixing that you wish to create. For most decisions regarding frequency balance, your ear is a much better guide than a spectrum analyzer. Width in mixing is defi ned by the panning spectrum from left to right. As I have pointed out, panning represents one of the most powerful tools in, THE ART OF DIGITAL AUDIO RECORDING creating eff ective mixes. It helps to think of panning as width, and as a three- dimensional mix as a goal, because it encourages you to use your entire spec- trum from left to right. Small variations in panning can dramatically alter the sense of space within a mix. D epth is the subtlest and most potentially artful and creative part of creat- ing a three-dimensional mix. As with height, depth may be thought of in two diff erent ways. Depth can be created just by volume relationships between ele- ments. Th e development of foreground and background elements through vol- ume relationships, as discussed earlier in this chapter, is one way to create the sense of depth in your mixes. Th e other is the delay pool made from all the delays and reverbs that you are using. As discussed, these delay elements can also have a signifi cant eff ect on panning and the sense of width in your mixes. M ixes as three-dimensional entities is really just another way of thinking about all of the practices already covered in this chapter. However, it provides a 198 concise way to view and evaluate your mixes, and it gives you a visual metaphor for imagining your mix. While this visual metaphor can be helpful—and we live in a culture that is heavily oriented toward seeing over the hearing—I cannot stress enough that, in the end, you must use your ear. All that really matters, to quote Ray Charles again, is: “What does it sound like?” DIAGRAM 6.1 Three-dimensional mixing model Revise, revise, revise H ere is a fi nal bit of general advice on building mixes. Everything I have dis- cussed in this chapter is subject to constant revision as you mix. You have to do some things fi rst and other things later to build a mix, and I’ve made sug- gestions on strategies for doing this, but you also have the option of returning to anything you’ve done previously. Th us, EQ and compressor settings, pan-, Mixing ning positions, reverb choices and amounts, not to mention just basic level placement, should all be subject to review and revision as a mix progresses. For this reason, it is valuable to save mixes under diff erent names once you think the mix is getting close to completion. I use a numbering hierar- chy; for example, if the song title were “Blackbird,” I’d start with a fi le named B lackbird Mix 1 . When that mix seemed close, or if I wanted to try a diff erent tact and was planning on altering a bunch of elements, I would save my mix as B lackbird Mix 2 before proceeding. If I stop work for the day, when I open the mix the next day to continue working, I would name it B lackbird Mix 3. If I decide to make an edit—let’s say I wanted to try cutting out the third verse—I would save that mix as Blackbird Mix 3 Edit . Once I am convinced I want to keep the edit, I would go back to a simple numbering hierarchy, but I would always be able to easily fi nd the last mix I did before I made the edit. Keeping track of mixes by using a naming structure is especially useful when working with other people, so that you can identify mixes as the pro- 199 cess continues. I discuss collaboration on mixes more thoroughly in the last section of this chapter. 6.3 Automation and Recall A utomation and recall capabilities have been greatly expanded within the DAW environment. A utomation refers to the ability to alter settings in real time as a mix plays. Recall refers to the ability to remember and restore all the settings in a mix. Th e ease with which a computer can handle data management has re- sulted in the ability to automate virtually every parameter in a mix. Th e nature of computer fi les means that if you have done all your mixing in the box (as discussed above, under “Mixing Tools”), you can have complete, accurate recall of your mixes in the few moments it takes to open the fi le. Th e extent of automation capability can be either a blessing (greatly in- creased creative options) or a curse (you can get lost in the endless number of possibilities). Th e ease, speed, and accuracy of the automation functions are only a blessing. As I discuss automation in mixing, I focus on the practical side of things, but I also touch on some of the creative capabilities that are open to the recordist as a result of automation in a DAW.

Online versus off -line automation

M any of the capabilities of DAW automation will become clear as I explore the diff erences between online and off -line automation. Online autom ation refers to changes made in real time. Th at means that faders or rotary knobs or other controllers are moved as the music plays and the automation system remembers whatever moves are made. Th is operates on the recording model; movements are “recorded” as they are made, and then played back on subsequent replays. DAWs usually use the term w rite for the act of recording, writing automation, THE ART OF DIGITAL AUDIO RECORDING data as controllers are moved and then reading them upon playback. Th e pro- cess oft en resembles recording in that the automation function needs to be armed and the “write ready” mode oft en consists of a fl ashing red light, just like the “record ready” mode for audio recording. Online automation follows the automation model established by the high-end analog recording consoles with integrated computers . Off -line automation refers to changes made independent of playback, usu- ally utilizing a graphic interface. Off -line automation functions similarly to the editing process and generally uses many of the audio editing tools in slightly altered fashion. Although the automation is controlled off -line, there can be im- mediate playback auditioning of the changes made. Some analog consoles have limited off -line functions, but the DAW has vastly expanded the capabilities of this approach to automation. B efore delving into the specifi cs of these two systems, I explore the pros and 200 cons of each. Online automation has the advantage of real-time input that allows the recordist to be responding to aural information, and it has a tactical compo- nent that means you can use the fi ne motor control in your fi nger for automation moves. Online automation has the disadvantages of being dependent on physical response time, which can be diffi cult when trying to do things such as raise the volume of one word in a continuous vocal line. To take advantage of the fi nger’s motor control, online automation also requires a hardware interface for your DAW. Moving controllers with the mouse does not provide nearly enough fi ne control for most of the kinds of changes made during the automation process. Off -line automation has the advantage of exceeding fi ne control over both the position and amount of controller changes—for example, raising the vol- ume of one word in a vocal line by exactly 1.2 dB is very easy with off -line auto- mation. Off -line automation also has the advantage of certain kinds of automa- tion moves, such as time-based auto-panning, that are impossible using online automation. (I explore these in more detail in the section “Details of Off -line Automation,” below.) Off -line automation has the disadvantage of not having a physical component (fi nger movement) and of being a completely diff erent process for those used to working online. I spent many years using the automation systems on SSL consoles, which had taken analog/digital online automation systems to new heights of function- ality and user friendliness. Nonetheless, I now do all of my automation off -line in Pro Tools. Th e ability to have precise control of parameters has proved too big an advantage, even over the familiarity of the online model. Some recordists fi nd that they prefer to control certain functions online—fades, for example— but most functions are faster and more accurately done off -line (and many are impossible online). Many recordists do not have a hardware interface for their DAW, and the constraints of mouse movement mean that they will naturally use off -line automation; but many of those with access to physical controllers are still tending toward off -line automation for most functions., Mixing D etails of online automation Th e basic “write/read” functionality of online automation is enhanced in many ways, though the details vary among DAWs. In most systems, you begin with a write pass, during which you create some of the basic automation moves that you want to hear. Once you’ve made one basic write pass with online au- tomation, you probably will work in one of various updating modes. A typical update mode might be called “touch.” In touch mode, the previous automa- tion is read until you move (touch) a fader or other controller, and then new automation begins to be written. Th ere may be two types of touch mode—in Pro Tools, touch mode retains all automation written aft er you release the controller you touched to begin rewriting, and the latch mode erases all the automation past the point of the touch update. Th e choice of which of these to use depends on whether you are updating a section in the middle of some established automation (touch) or working across a timeline from beginning to end (latch). 201 Another common online automation mode is “trim,” which updates al- ready written automation. If you had a bunch of automation moves on the lead vocal of a song’s chorus, for example, but decided the whole thing needed to be a little louder, you would use the trim mode to increase the volume (trim up) the entire section. Th e trim function would change the overall volume while retaining the previous automation moves. D etails and further functionality of online automation will vary in diff er- ent DAWs and with diff erent hardware controllers. If you have access to physi- cal controllers, I recommend that you familiarize yourself with their use, but that you also explore off -line automation for increased automation accuracy and functionality. Details of off -line automation Off -line automation, using a graphic interface, allows for very fi ne control of au- tomation data and the opportunity for some unique automation eff ects. Off -line graphic automation uses a horizontal line to represent a scale of values: the higher the line on the graph, the greater the value of the parameter setting. For volume, the horizontal line represents the fader setting—all the way up is the maximum fader level (+12 dB on many systems) and all the way down is 0 dB (equivalent to off ). Th e following screenshot shows some volume automation created by rais- ing and lowering certain parts of a vocal take. Th e line represents volume, with greater volume (output fader position) indicated when the line is higher and less volume when lower. In the background, you can still see the waveform of the vocal, allowing you to pinpoint the places that you wish to raise or lower volume. Although the actual movement of the volume by raising or lowering the line on the graph is done off -line (the music or program material is not playing), you can immediately audition the results by having the curser placed just in front of the passage being automated and playing back the results., THE ART OF DIGITAL AUDIO RECORDING SCREENSHOT 6.8 Volume automation on a vocal trackAsmentioned previously, the big advantage to this kind of off -line auto- mation control is the ability to easily select the exact portion of audio that you wish to control and then to make very precise changes in parameters. Most sys- tems allow control to 1/10 of a dB (.1 dB increments), and this allows for very fi ne tuning. Aft er using this technique for a while, you begin to become familiar with the likely results from certain degrees of parameter changes. I have a good idea of whata1dB or 2 dB (or 1.5 dB!) change in volume is going to sound like, 202 so I can oft en make exactly the right automation move for what I want to hear on the fi rst try. In any event, I can easily revise a move by whatever increment I want in order to achieve the result I want. Some systems show both the new absolute level as you move a portion of the vertical line and the change in level. In the following screenshot, you can see the readout is showing the original level (-2 dB) and then in parenthesis is the new level (-.8 dB) and the change in level (1.2 dB). Th e change in level is preceded by a triangle, which is the Greek symbol for change (delta). SCREENSHOT 6.9 Off-line automation readout Level changes in auxiliary sends can also be created off -line, allowing for easy implementation of special eff ects, such as a repeat echo on one word within a vocal line. By accessing the eff ects send level in the graphic automation mode, you can take a send that is set to 0 dB (so no eff ect is heard) to whatever level you wish in order to create the special eff ect. Because the graphic representation of the program material is seen in the background, it is easy to isolate the eff ect send on something like one word. Breakpoints indicate the spots where the graphic line moves in position. In Screenshots 6.8 to 6.10, all the movement between the breakpoints is linear. SCREENSHOT 6.10 A utomating a send so that one word goes to an effect, Mixing Online automation will create nonlinear data, which is refl ected in the graphic readout by multiple breakpoints. Many DAWs provide tools that allow you to draw nonlinear or free-hand automation data off -line as well. To prevent over- taxing the computer’s CPU, you might be able to thin the nonlinear automation data, as shown in Screenshot 6.11. SCREENSHOT 6.11 Nonlinear automation data as written, below as thinned Th ese same tools might be confi gured in various other graphic arrange- ments, such as triangles or rectangles. Th e graphic shapes are typically used in one of the editing grid modes. Grids set in musical time—for example, a quarter-note or an eighth-note grid—allow for some great special eff ects done in musical time. Th e following screenshot shows two diff erent panning eff ects, the fi rst using a triangular shape to create smooth movements between hard right and hard left , and the second using a rectangular shape to jump from right to left and back again. Th e general eff ect is oft en referred to as auto-panning, as it is the automatic and regular changes in panning position. SCREENSHOT 6.12 Variations in “auto- panning” type effects using off-line panning automation Th e following eff ect uses the same triangle-based automation editing tool on off -line volume rather than for panning. Th is creates a tremolo eff ect in mu- sical time (tremolo is created through cyclical changes in volume). S CREENSHOT 6.13 A tremolo effect using off- line volume automation Advanced automation techniques A utomation is created using the tools I have discussed, but it can become com- plex when many elements are combined and manipulated in great detail. You can create elaborate graphic automation that alters changes on every word in a, THE ART OF DIGITAL AUDIO RECORDING vocal, and you can automate volume, panning, sends, and plug-in parameters on every track. Automating plug-in parameters off ers a near endless number of possible real-time changes through automation, but it also threatens to cre- ate complexity with little audible advantage. Th e depth of possibilities though automation provides wonderful creative opportunities, but they need to be balanced against maintaining a coherent vision of the overall sound being cre- ated. Sometimes mixes can be overworked to the point that the bigger picture is lost in the details, so the mix doesn’t hold together. Sometimes simple mixes sound the best. One convenient technique involves trimming volume on elements in your mix without using the automation functions. I described the trim function above, in discussing online automation, and you can trim sections of automa- tion off -line, using the graphic interface, as well. However, when you wish to trim the volume up or down on an entire track, it is oft en quicker and more 204 convenient to use the output function of one of your plug-ins. Some DAWs provide a separate trim plug-in for just this purpose. By raising or lowering the output on a plug-in, you eff ectively trim up or down that track, retaining all of the volume automation already written for that track. Compressor outputs are oft en good candidates, but it could be a dedicated trim plug-in or one of many other options, depending on what plug-ins you already are using on the track you wish to trim. In the analog world, we used to call this “fooling the automa- tion” because it allowed global volume changes to a track without the time- consuming job of trimming an entire track in real time (as was necessary on most automation systems within analog consoles). It’s easier and quicker to trim off -line now, using the graphic interface, but “fooling the automation” with plug-in outputs is still a convenient way to make adjustments. Although it should be clear from the above discussions, and it will cer- tainly be clear once you start working with automation, any automating that is done in one mode will be refl ected in the other mode. Th at is, online automa- tion moves show up in the off -line graphic automation mode, and off -line au- tomation moves create the same real-time eff ects, such as fader movement, that online automation creates. Advanced automation practices may involve use of both modes of operation to create the automation data you want. For example, you might write a piece of automation online for the creative engagement of working as the music plays, and then make small changes to details in off -line mode where you are able to fi ne-tune all the parameters. As you gain familiar- ity with your automation system, you can explore the best ways to achieve your creative vision. Recall Recall r efers to the ability to recall all the parameters of a mix. Th is includes automation, panning, plug-ins and their settings, and anything else you have done to create your fi nal mix. Th is used to be a very diffi cult, if not impossible,, Mixing process when using analog equipment. Eventually, elaborate computer-assisted analog consoles were developed that could remember the position of every fader and knob on the console and could display those graphically. Nonetheless, an operator had to reset each parameter on the console by hand. In addition, someone (usually an assistant engineer) had to log all of the hardware outboard gear that was used, what the signal path was, and what the setting was for each parameter on each piece of gear—and all of these had to be reset by hand. Th is was a long and tedious project, and as you might imagine with so many settings involved, not always successful. W hile the debate continues over in-the-box mixing (mixing entirely within the DAW) versus use of some gear outside of the DAW, in regard to recall, in-the-box mixing provides the ultimate in convenience and reliability. In the time it takes to open a session fi le (less that one minute), you can re- call complete and perfectly accurate mixes. Many of us have come to rely on this capability, especially as remote mixing has become more common. Remote 205 mixing—sometimes called unattended mixing— refers to working with clients in other locations by sending mixes over the Internet and taking feedback and making revisions aft er the client has had an opportunity to review the mix (see the Appendix for information on some of the formats commonly used for shar- ing mixes). DAW recall has opened up the possibilities for these kinds of mix- ing strategies that rely on easy, accurate recall at the click of a mouse! 6 .4 Mix Collaboration, Communication, and Delivery In the end, mixing is almost always a collaborative process. What used to be a bunch of people with their hands on faders, trying to make mix moves in real-time because there was no automation, has become mixes of enormous complexity recalled and replayed eff ortlessly. And what used to be groups of recordists and artists working late into the night, trying to get a mix done be- fore the next recording group came in and broke down the console in order to start a new session, has become a series of mixes and responses oft en sent via the Internet from remote locations and sometimes going on for weeks. In between are any number of combinations of collaboration and communication used to complete a mix. It’s not possible to cover them all, but I discuss some collaborative possibilities to consider and some ways of talking about mixes as they progress. D elivery of mixes has also come a long way from the ¼-inch 15 IPS tape master. While delivery formats have always been in fl ux, contemporary digi- tal fi le formats off er a large number of possibilities. Fortunately, there is much less of a compatibility problem than when a particular piece of hardware was required for each possible delivery format, as DAWs can usually handle most digital audio fi les. A larger question remains about the best way to deliver your, THE ART OF DIGITAL AUDIO RECORDING mix for mastering, and I begin that discussion here, before delving into it more deeply in the following chapter on mastering.

M ixing collaboration and communication

Y ou can’t separate working together on mixing with communicating about mixes, as the ability to talk about mixes is required in order to collaborate on them. Mixing collaboration now comes in many forms, both technical and in- terpersonal, and happens both in close contact and remotely. Working with others remotely means using some medium for communication (phone, text- ing, e-mail, etc.). Th is can have some advantages—having to put mix notes into writing can make revisions easier and clearer for the recordist, though some- times the written word can be as obscure as the spoken one (“Please make this mix more purple”). 206 Having the language for communicating about mixing is largely a matter of having built a vocabulary for talking about mix and sound issues. Some things are easy and straightforward—“I think the vocal needs to be louder”—though this leaves the question of how much louder still an open matter. “I think the vocal needs to be a lot louder” or “a little louder” helps clarify things, but the exact degree of change that is going to satisfy the request is still a matter of trial and error. Working in collaboration is another reason I like off -line automation. I can adjust the vocal up 2 dB, and if my collaborator says that’s too much, I can say I’ll split the diff erence (up 1 dB) and we can work from there, knowing ex- actly what changes have been made and adjusting in defi nable degrees. Mix issues other than questions of volume start to create a greater need for a shared vocabulary. Questions regarding frequencies, as controlled by EQ, have inspired a huge vocabulary of descriptive words, some more easily under- stood than others. Words that rely on the scale from low to high frequencies are more easily understood and interpreted. Th ese include bass, middle, treble, or b ottom and top. Other words that are used pretty frequently are suggestive but less precise and thus open to more interpretation—words such as b oom, rum- ble, thump, fatter, warmer, honk, thinner, whack , presence, crunch, brighter, edge, brilliance, sibilance, and air . Th ese might be pretty easy to interpret, especially if they are used oft en among frequent collaborators, but they may also mean very diff erent things to diff erent people. Other words, such as the sometimes inevitable color references or highly subjective terms such as “magical,” really give the recordist almost nothing to go on. Th e most precise language for EQ is specifi c frequency references, and with the proliferation of engineering skills among musicians and other con- tributors to the mixing process, these are becoming more frequently used. Sug- gestions such as “I think it needs a little boost around 8 K” or “Perhaps we could thin this sound a bit around 300” (meaning dipping at 300 Hz) are becoming increasingly common in mix collaborations. Th e recordist may still need to ad-, Mixing just somewhat from the suggested frequency—it’s impossible to know exactly what the eff ects of any given frequency adjustment are going to be without lis- tening—but this language is certainly the most precise and the easiest to re- spond to. Communications about ambience and eff ects can be more obscure. A re- quest for a sound that is “bigger” probably refers to a desire for increased ambi- ence—but not necessarily. Again, suggestions that a more “mysterious” or “un- usual” mix is desired leave the recordist without a good idea of how to proceed. With the widespread use of recording gear, however, specifi c suggestions and references are more common. A guitarist may well suggest, “How about some long delay on the lead guitar?” Th e guitarist may even be more specifi c: “Can we try a quarter note delay on the guitar?” Th e more exact nature of the delay (overall level, amount of feedback, etc.) may be left to the recordist or may con- tinue to be part of collaboration as an ongoing discussion of details. Some terms can suggest changes in mix ambience pretty clearly. Certainly 207 “wetter” and “dryer” are accepted terms describing relative amounts of reverb and/or delay, though how to implement a request for a wetter vocal or a wetter mix still leaves a lot of options open to the recordist. Similarly, terms such as “closer” or “farther” generally can be interpreted as references to types or de- grees of ambience, though again the way to accomplish such changes can vary widely. I t is very helpful for a recordist to have a variety of terms available to try to help the collaborators clarify what it is they want out of a mix. Sometimes, when a person is struggling with what he or she wants out of the sound of their vocal, for example, the person can be greatly aided by being asked if it should sound more “present” or “closer” or perhaps “bigger” or “richer.” Th is can give the col- laborator a term that you might then have a chance of interpreting technically, as opposed to something like, “Could you change the way the vocal sounds?” Of course, you can, but how? Don’t rely on your collaborators to clearly express their interests; develop the vocabulary to help them (and you) create mixes that you all love. Finally, when working remotely, make sure you are listening to and col- laborating on the same mix! I have had confusion with artists over elements in a mix, only to discover that we were not referencing the same mix. Th is is why I number and/or date the CDs I give to artists. I can then refer that information back to a specifi c mix fi le so that changes are made from the correct starting point.

Delivering mixes

Th e best way to deliver mixes depends on answers to a couple of key of ques- tions: To whom are you delivering them? and For what purpose? Th e mix for- mat must be appropriate for the person who is receiving the mix. Oft en, you, THE ART OF DIGITAL AUDIO RECORDING will need to deliver mixes in a variety of formats to diff erent participants in the process. In a commercial project, you may need to deliver one mix to the artist, one mix to the record company, one mix to the Webmaster, and one mix to the mastering house. To a large extent, questions surrounding the mastering be- come an important part of how you deliver your mixes. Th is is true whether or not your project is going to undergo a formal mastering process. If your project is not going to be mastered beyond your fi nal mix, then you will need to incor- porate at least some of the standard processing for mastering as a part of your mix. If it is going to be mastered, you will probably want to deliver a separate mix format to everyone involved in the project other than the mastering house and two diff erent formats to the mastering house. I cover most of these topics in the following chapter on mastering and also in the Appendix. As a part of this chapter on mixing, however, I want to alert you to the fact that you will need to have a good understanding of the 208 mastering process in order to fi nish your mixes properly, whether or not they are going on for fi nal mastering. As I mentioned in the above section on uses of compression in mixing, there is a type of compression that has become an essential part of the mastering process, and that is brick-wall limiting. Th is is discussed more thoroughly in the following chapter on mastering, but for now know that brick-wall limiting has a profound eff ect on mixes. For this reason, you will probably want to use it prior to fi nishing your mixes, so you have a better idea of what they are going to sound like aft er mastering. You will also probably want to use it on all mixes (including rough mixes) that you give to the people you are working with, so that what they hear will sound more like what the fi nal recording is going to sound like. In most cases, the only time I create a mix without brick-wall limiting is when I make the fi le that is going to be used for mastering, so that the limiting can be applied as a part of mastering instead. Even then, I also supply the mastering house with a brick-wall version, so they can hear what the artist has been hearing and something close to the way I imagine my mix will sound aft er the mastering process restores the limiting that I have removed for their working fi le. T aking multiple mixesWeused to record multiple versions of a mix, simply as a matter of course. Because it was so diffi cult or impossible to recreate a mix once the studio was reconfi gured for another session, we would try to anticipate changes that we might want to consider. Th e most common variations on mixes were ones with diff erent lead vocal levels. We’d take a mix and then a “vocal up” mix in case we wanted a louder vocal. We might also take a “vocal down” mix, or two mixes with diff erent “vocal up” levels, or a “drums up” mix, and so on. Th e problem, of course, was that there were an endless number of possible options, and the time and materials it took to run alternative mixes started to defeat the purpose., Mixing If you are mixing in the box, then the only reason to take multiple mixes is to have diff erent possibilities to review. Otherwise, it is usually most effi cient to make one mix for review and then simply open the fi le and make revisions as desired. Even if you are supplementing your mix with some outboard gear, if most of the work is done in your DAW, it might be fairly easy to log the settings on a few external pieces, so as to allow for pretty simple recall. Many recordists and artists have come to depend on ease of recall as a means of providing op- portunities to live with mixes for a while, or to work remotely, with easy revi- sions being an essential part of mixing collaborations.,

C hapter 7 M astering O ne Last Session

I am calling this chapter on mastering “One Last Session” because mastering is typically the last part of the process that starts with recording or assembling audio and ends with providing audio destined for the end user: CD, Internet download, Internet streaming, game audio, DVD, and so on. Mastering also is typically done in one session, unlike the recording, editing, and mixing pro- cesses that frequently take place over multiple sessions. However, depending on the size of the project, and the degree of scrutiny of all the details, master- ing can require several sessions or at least several rounds of revisions. While the following chapter is hardly intended to be exhaustive, it provides a basis for understanding and undertaking the mastering process. As you progress through the various stages of creating audio programs, it is important to be familiar with each previous stage before undertaking the next one. For this reason, a good mastering engineer has a strong basis in recording, editing, and mixing. 7.1 What, Why, How, and Where What is mastering? Why do we need to do it? How might you accomplish it? And what is needed in terms of a facility in order to eff ectively master a project? Th ere are no simple answers to these questions, but fi rst you must ask both what “eff ectively” is and what the “project” is at hand. To master eff ectively requires a certain amount of gear/soft ware, experience, a critical ear, and usually a healthy willingness to collaborate. Mastering is a skill, and there’s no substitute for time spent mastering to develop that skill. Having the luxury of a home system, and, Mastering having projects to work on, will allow you to go through the trial-and-error process necessary to develop mastering skills—guided by the good advice from this book, of course. Th e nature of the music in the project, as well as the recording and mixing already done, will greatly aff ect your ability to master eff ectively. Musical genres have many conventions in terms of how fi nal masters generally sound, and even if your goal is to defy those conventions, you will have limited success master- ing styles of music that you are not very familiar with. Th e number of elements in the fi nal audio program is also important to the mastering skill set. Th ere can be beautiful masters made from solo piano recordings, but that is quite a dif- ferent task from mastering a compilation of large ensemble recordings from a variety of sources. Diff erent projects will suggest diff erent sets of tools, and dif- ferent kinds of approaches to mastering. Ultimately, however, your greatest asset with mastering is the same as your greatest asset with all other elements in the recording process—it’s your ear! Th e more experienced and developed your ear, the better your chances for eff ective mastering of any kind of project. Finally, if you notice a marked similarity between this introductory paragraph and the 211 one for the previous chapter on mixing, it is because the overviews for these two parts of the recording skill set are remarkably similar.

What is mastering?

L et’s establish exactly what is meant by the term m astering. As the term suggests, mastering is the creation of a fi nal “master” version of your audio program. Th is fi nal version is what is usually delivered to the manufacturer to replicate as CDs or for other forms of duplication or dissemination, such as audio placed on the Internet for streaming or downloading, or loaded onto a DVD as audio to ac- company video, or placed into a game as audio to accompany game play. Typi- cally, a mastering session involves working with the fi nal mix or mixes that, in combination, form the complete audio program for any particular project.

Why master?

Th e goal of mastering is to create the best fi nal version possible, and to put that version into the correct format for its fi nal destination or destinations. Th e “best” version possible is, of course, a subjective process that requires creative decisions and may vary considerably depending on who is doing the mastering. It is also very much dependent on what happened before, in the recording and mixing of the audio, as these will have been completed before mastering begins. Th e fundamental task of mastering is to make all the audio elements work to- gether in their fi nal delivery confi guration. I cover the creative process in the following sections on the “how to” of mastering. C reating the correct format is the other part of mastering, and this is purely technical. Diff erent audio applications require diff erent fi le formats, and, THE ART OF DIGITAL AUDIO RECORDING their creation may require particular media (CD-Rs, DVD-Rs, hard drives, lac- quers, etc.). I discuss delivery formats at the end of this chapter.

How do you master? Th e basics

In order to make the master that has been creatively and technically optimized for fi nal delivery, there are a variety of typical tasks. Generally, the most essen- tial job in mastering is to set fi nal levels for all of the elements. Beyond this, it is the job of mastering to balance the sonic characteristics of all the elements so that they fi t well together. Finally, it is a part of mastering to put all the elements together exactly as they are meant to be in their fi nal delivery—sequencing and creating the spaces between all the songs on a CD, for example. I cover these level, sonic characteristics, and sequencing considerations separately. Th ere may be other tasks in mastering as well, many of which I cover in the section on advanced mastering techniques. 212 L evel Th ere are two basic aspects to setting levels in mastering—absolute level and relative level—and they interact, so they need to be considered as one process. Absolute level refers to the volume of the particular audio element (such as each individual song on a CD) and r elative level refers to how loud the ele- ment sounds relative to the other elements in the project. I use the model of mastering a CD made up of a variety of songs in the following discussion, but you could be mastering any number of diff erent kinds of audio for diff erent projects. Just substitute “audio element” for “song” in the following if you’re working on something other than a traditional CD. Because of the dynamic range of audio, trying to balance the volume from song to song is a subjective process. Th e key to level balancing in mastering is to focus on the loudest part of each song. Th e goal is to give the listener a consistent experience when listening to the entire CD. If the loudest part of each song is balanced with the other songs, then the listener will never feel like a particular song on the CD has suddenly gotten louder. When the loudest part of each song on the CD is relatively balanced, the quieter sections will vary depending on the dynamics of the song, but this is the nature of musical dynamics and it doesn’t usually present a problem. Brickwall limiting Absolute volume is the volume relative to digital zero. Digital audio has a binary code for volume (along with everything else), and the maximum volume is digi- tal zero. Contemporary mastering tools include a brickwall limiter that allows the recordist to push the program material up against the digital volume ceiling. Brickwall limiting increases the absolute volume of your program material, and therefore aff ects the relative volume between songs., Mastering To understand brickwall limiting, you might begin with the processing known as peak normalization . To normalize a piece of audio means to fi nd the loudest sample (peak) and to raise the volume of the entire audio piece to a given maximum volume. While this may be digital zero, it is usually recommended that you stop just short of digital zero to prevent misreadings by the CD player that may result in distortion. Th e typical normalization (and brickwall) ceiling is -.2 dB (2/10ths of a dB below digital zero). Normalizing raises the volume of every sample equally, placing the loudest sample at whatever limit you set (e.g., -.2 dB). Because the overall volume of each sample is raised the same amount, it doesn’t change the dynamics of the audio piece. Brickwall limiting takes the idea of normalization and extends it into lim- iting. Instead of just placing the one loudest sample at -.2 dB and everything else in the same relative volume position, brickwall limiting allows you to push all the samples above a user-set threshold up to the ceiling. It does this by setting a “brick wall” at the ceiling point (-.2 dB, for example). As the threshold lowers, more and more samples are lift ed up to the brickwall volume limit. Th e lower the threshold, the greater the number of samples that are lift ed to the ceiling of 213 the brick wall. In the following screenshot, the threshold is set to create about 3 dB of brickwall limiting at the moment that the screenshot is captured. SCREENSHOT 7.1 A brickwall limiter, THE ART OF DIGITAL AUDIO RECORDING Th eoretically, a song that has considerable brickwall limiting isn’t any louder than a song that has been normalized to the same limit, in the sense that there aren’t any samples in the brickwall-limited version that are louder than the one loudest sample in the normalized version. However, the brickwall- limited version may sound considerably louder because so many more of the samples are reaching the volume ceiling. O ver time, the extent of brickwall limiting has increased in a sort of es- calating “volume war” to make one CD sound louder than another. Th ere has also been considerable debate about the eff ects of brickwall limiting and the decrease in dynamic range that is created by the process. Some go so far as to argue that brickwall limiting has eff ectively killed popular music by fl attening musical dynamics to such an extent as to make music unpleasant to listen to. It is true that consumers tend to listen to CDs many fewer times than they did in the past, but the extent to which this is the result of brickwall limiting (versus the natural results of a less patient society), we can’t really know. 214 Balancing levels B alancing the levels in mastering a popular music CD is usually accomplished by adjusting the threshold on the brickwall limiter. Th is means that relative levels are controlled by the degree of limiting used—by the extent to which the song is pressed against the absolute level. Th is is necessary because, once you have started to use the brickwall limiting process, the only way to make a song louder is to increase the brickwall limit. If you try to simply raise the volume aft er hitting the brickwall ceiling, you will get digital distortion (audio pushed beyond digital zero). O nce the brickwall-limiting process is begun, you will generally raise or lower a song’s overall volume with the limiter’s threshold control. While you can lower overall volumes rather than lowering the extent of limiting with the threshold control, this will oft en produce undesirable results. If the ceiling of song 1 is set to -.2 dB, with many samples hitting that limit, and song 2 sounds louder than song 1 with the same threshold setting on the brickwall limiter, you may be tempted to reduce the ceiling of song 2. If you set the ceiling (output) to -1.2 dB, for example, song 2 will oft en sound unnaturally quieter than song 1. Th e eff ect of raising the threshold of song 2, to decrease the extent of brick- wall limiting, will usually produce a more desirable result in level balancing. Exceptions to this are likely to result from program material that is sonically very diff erent. If you have a full-band record with one or two songs that are just acoustic guitar and voice, for example, you may fi nd that you do need to lower the overall level of the acoustic songs to prevent their sounding louder than the band tracks. Th e best approach to achieve level balancing is repeated listening, focus- ing on the loudest sections in each song. DAWs allow you to easily jump from one song to the next, and mastering sessions sometimes sound like a jumble of, Mastering snippets as you jump from song to song, listening to short segments of each. It oft en takes many rounds of listening and making very small adjustments before you begin to feel as if the overall level of each song is well balanced against the others. Sonic characteristics By sonic characteristics, I refer to the qualities of the sound that might be ad- justed with your typical DSP tools. Th ese would be EQ, compression, and ambi- ence. EQ adjustments in mastering are common, though usually rather subtle, whereas additional compression or expansion (outside of the brickwall limiting discussed above) is less common, and adding ambience, such as added reverb or delay, is quite rare. Nonetheless, all these tools and many others—including fades or editing—are sometimes part of the mastering process. Wanting to adjust the sonic characteristics of songs may involve two dis- tinct goals. Th e fi rst is to make each song sound as good as possible. Th e second is to give the songs a sonic consistency from the beginning of the CD to the end. Th e fi rst goal should be approached cautiously, with the assumption that 215 the mixer has made the song sound the way everyone involved in the produc- tion wanted. I try not to second-guess the production process that has come for mastering. I might add a very small amount of EQ, or even compression or reverb, based purely on the sound of the individual song, but I need to feel strongly that the song would be improved. Otherwise I accept the mixes and I focus on the second goal. Sonic consistency generally means that the frequency balance from song to song sounds uniform. If one song has a stronger low end or a brighter top than the following song, one will suff er by comparison. In these cases, EQ ad- justments are not made so an individual song sounds “better” but, rather, to bal- ance the frequencies from one song to the next. Of course, the decision whether to dip the low end or the high end of one, or boost the lows or highs in the other, is part of the creative process. I oft en listen to all the tracks on the CD and try to fi nd one that I feel has the best overall frequency balance, and I use that as my model. I will then EQ other songs to match the frequency balance of my model song as best as possible. Again, as song keys and arrangements vary, this can be a highly subjective endeavor, but it may serve as a good working model. I fi nd that it is helpful to work with the songs in the same sequence as they will appear on the CD. Ultimately, all the songs should work together in any order, but sometimes knowing the fi nal sequence can help guide both level and frequency adjustments. S equencing and spreadsInatypical CD mastering session, creating the fi nal sequence of songs and the time between songs (spreads) is usually the last step. You may actually burn your CD-R master straight from your DAW, in which case sequencing and spreads, THE ART OF DIGITAL AUDIO RECORDING will be handled in the same program as all the other mastering functions; but oft en I prepare all the fi les in the DAW, create the master fi les for whatever for- mat is required, and then do the fi nal sequencing and spreads (assembling) in a separate program. Deciding on the sequence of the songs for a CD is an art in itself. Flow, energy, and commercial concerns are part of the decision-making process. For- tunately, the ease of burning CDs at home means that the creative team can try out a variety of sequences either before mastering or as part of the mastering process. Changing the sequence is simple and CD-Rs are very inexpensive, so if you aren’t sure of your sequence, it’s worth trying out numerous possibilities. S preads between songs can be deceptive because they depend a lot on the volume of playback. If some of the songs fade, or even have a short tail of reverb or other ambience at the end, the apparent time before the next song depends on how much of that fade or tail is heard, and that depends on how loud the music is. Quiet listening will make the spread seem longer and loud listening will make them seem shorter. Moderate listening levels are the best compro- 216 mise for setting spreads. O ne technique for setting spreads is to play the end of each song and have one person—whoever is most responsible for setting the spreads—indicate the moment he or she thinks that the next song should enter. Th e person might tap on a table, say “Now,” or whatever. You can assemble the master, song by song, in this way. Or you might just place a default time (usually 2 seconds) between each song. Th en, once the master is assembled, listen to each spread and adjust according to taste. Some burning programs have the ability to play a user-defi ned amount of time at the end of each song and the beginning of the next, essentially playing each spread for you in sequence. It’s best to listen to all of the spreads, making notes about each one as you go; conversation during the listening process means you will likely miss hearing the next spread play. You can then adjust the spreads that felt as if they needed more or less time and listen again until you’re satisfi ed. Creating and delivering your master Th ere are a lot of technical requirements for making a proper CD-R master, but fortunately most of them, such as fi le format and P&Q codes, are taken care of automatically by your CD-burning program. Some burning programs still off er a choice between TAO (track-at-once) and DAO (disc-at-once) burning proto- col. CD-R masters must be burned using DAO protocol, but that is the default for most programs. Th ere are options for what information is added to the audio program, and in the digital age this has become increasingly important. CDs can code the song title and artist name to every song by using a CD-burning program capable of adding CD text. You should make sure that your master has that information encoded, so that it will appear on CD players and computer pro-, Mastering grams that read those data and radio stations that transmit them. CDs can also have an ISRC (International Standard Recording Code) identifi er included for each song. Th is code that provides ownership information, so that tracks can be digitally traced for royalty collection, administration, and antipiracy problems. You have to register to get these codes—they used to be free, but now there is a charge to get your codes. Search ISRC code registration on the Internet for more information. Once your CD-burning program has everything sequenced as you want it—the protocol set to DAO, with the proper spreads and all the text and coding information correctly entered—you are ready to burn your master to a CD-R and send it off for manufacture. Th ere seem to be endless debates as to what CD-R medium is best (which manufacturer, which color, etc.) and what speed masters should be burned at. Th ere has been a lot of testing done, and the up- shot is that it really depends on your burner. No one medium is better and no one burning speed is best. In fact, tests show that sometimes faster burn speeds result in CD-Rs with few error readings. Most of the time, pretty much any CD-R, burned at any speed, will create a master with error rates well below the 217 danger level. If you fi nd a particular brand of CD-R that seems to work well with your burner, and you are getting good results at a particular burn speed, then you might just want to stick with that. You might want to try diff erent brands of CD-R and diff erent burn speeds, and do some listening tests to settle on a way to create your masters with your system. O nce the master is burned you will need to listen to it before sending it for manufacture. It does no harm to a CD-R to play it, as long as it’s handled care- fully, by the edges. Occasionally there are problems with a blank CD-R or with a burn, and there can be audible pops or distortion on a particular burn. You want to listen carefully to the master to make sure it doesn’t have any of these unforeseen problems before sending it to the manufacturer. M ost professional burning programs also create a printout that shows the track list, the time between tracks, index times, cross-fades, and so on. Manu- facturing plants like to have a copy of this printout to confi rm what they are seeing when they analyze your master, but most plants will accept your master without the printout. Be sure to burn a duplicate master for yourself so you can compare it to the manufactured CDs that are sent to you later by the plant. You should not be able to hear anything more than the very slightest diff erence be- tween your burned CD master and the completed CDs from the plant. I n regard to digital delivery formats other than the CD-R, see the Appen- dix. How do you master? Advanced techniques Mastering may encompass a variety of tasks beyond the basics covered above, though most mastering sessions don’t get much more complicated than what I’ve described. Certain things, such as editing, are usually considered part of, THE ART OF DIGITAL AUDIO RECORDING S CREENSHOT 7.2 A printout from a master created in the program Jam the recording or mixing process, but they sometimes end up getting done at mastering sessions. I have received fi les in which the ending fade of songs was saved for mastering, but that is the exception rather than the rule. In some instances, sections of songs may be treated diff erently in master- ing. Most common would be something like an acoustic guitar introduction that sounds a bit too quiet or too loud prior to the entrance of the full band. Th e brickwall limiting or other mastering processing may have changed the rela- tionship between the two elements enough to require some adjustment. In that case, the intro might be raised or lowered in volume. Manipulating individual sections of songs is certainly possible, and I know of mastering sessions where a lot of volume and EQ changes were made to songs on a section-by-section basis. An example would be, say, a little high-frequency boost on the choruses and a little level boost on the bridge. But this starts getting very close to remix- ing, and if there are a lot of section-by-section changes, or if you’re not getting what your really want by trying to work this way, you will need to go back to the mixing stage and have the changes made there. I have found that on proj- ects that I both mixed and mastered, I have occasionally gone back to mixes to, Mastering make changes right in the middle of the mastering session; this is a luxury that is made quite easy if all the mixing and mastering is done in the DAW, so that recalling and changing mixes can be very quick. A recent trend in mastering is called separation mastering. Th is involves delivering s tems of the fi nal mixes that can be processed individually in master- ing. Stems—a term that comes from a common practice in audio delivered for fi lms—refers to submixes of certain elements that can be recombined to create the fi nal mix. In a recent project, I delivered four stereo stems for mastering: drum set minus snare drum, snare drum, all other instruments, and all vocals. Th e advantage may be twofold: you can control the amount of snare drum more easily, and that is the element that oft en gets the most suppressed (lowered in level) by brickwall limiting; and you might maintain slightly greater breadth in your stereo image and a bit more clarity because the elements are not as inter- mingled by the stereo buss processing (typically compression but sometimes additional EQ, analog saturation simulation, etc.). Th is second advantage may instead be a disadvantage to separation mastering. Stereo processing integrates elements in a mix in a way that might be desirable and that will be diminished 219 by separation mastering. Stereo buss processing, such as compression and ana- log saturation simulation, may add punch and warmth to the fi nal mix. Th is ste- reo buss processing could be added in the mastering stage when the stems are combined, but then mixing becomes increasingly removed from the fi nal sound of the recording, making it more diffi cult to mix eff ectively. Th e diff erences are fairly subtle, but I have found that, more oft en than not, I prefer masters made from the stereo mix to those created by the separation mastering technique. Separation mastering also adds time and expense to a project. In regard to creating the fi nal sequence and spreads, there may be the de- sire to do some more elaborate transitions between songs. Th is could include cross-fades where one song begins as the previous song is fading. You may also need to create separate CD track identifi cation number at places where there is no audible break in the music (as in a live music recording). You may also want to include an “invisible track” that occurs at the end of a CD without a track identifi cation number. Th e ability to create these kinds of advanced sequencing techniques will depend on the ability of your particular CD-burning program and will be something you would want to investigate in acquiring a program to use for mastering purposes.

Where do you master? Mastering environments and tools

C an you do your own mastering? If so, what do you need to do it? Th ere are no simple answers to these questions. What is required for good mastering is as follows: 1. very good playback system in a room that you trust. Th e system must be fl at (within reason) and the room consistent through the, THE ART OF DIGITAL AUDIO RECORDING frequency range. Th e system should extend comfortably through the entire frequency range—which may mean the addition of a properly calibrated subwoofer. Being able to evaluate the low end, especially the lowest frequencies that might not show up anywhere but in a nightclub or other environments that use sub- woofers, is an important part of mastering because material may have been recorded and mixed in environments that don’t have that capability. 2. Ataminimum, high-quality brickwall limiting and EQ pro- cessing. Th ese are critical tools. Other processing gear is valu- able, including compression, multiband compression, reverb, and analog saturation simulation soft ware. Th ere are mastering “suites” of plug-ins made by soft ware makers that contain at least the basic tools needed in forms optimized for mastering. Because mastering is done to single fi les of completed program material (mono, stereo, or surround), there is not a concern about delays 220 (latency) that may be caused by excessive plug-in processing (this can be a problem in mix situations). Some mastering soft ware uses phase-aligning algorithms that make for very high qual- ity processing, but the phase-aligning process means that these plug-ins introduce signifi cant delay. Because the whole program material is being processed at the same time, these delays don’t aff ect alignment with any other audio and these processors can yield outstanding results. 3. A CD-burning program that burns using DAO protocol (usually standard).Preferably you want a program that has cross-fade, CD numbering, indexing, and ISRC-coding capabilities. Th ere are many ways to burn CDs, including some very simple programs built into some computer operating systems. Th ey probably are all capable of creating CD-R masters that would work fi ne for manufacturing, but you do need to verify that the disc is being burned using the DAO protocol. More elaborate burning pro- grams off er the capabilities that may be essential in many master- ing situations (such as text and IRSC coding). 4. Th e ability to hear audio programming material in ways that allow you to make accurate and creative judgments about the relative levels and sonic qualities of the material you are master- ing. Th is is the result of the ear-training process that requires experience and attention to the essential issues of mastering. What it is that constitutes an appropriate listening environment for mas- tering, or the proper soft ware or hardware tools, or the ear and creative capa- bilities to utilize the gear that you have, is open to interpretation. Some pretty, Mastering basic combination of the above is enough to get you started, and as with all things audio, experience is the best teacher. Once the master is sent off and ap- proved by the manufacturing plant (or Webmaster or DVD authoring person, etc.), you have fulfi lled your responsibilities that may have started when the fi rst sound for the recording was made (or imported or downloaded, etc.). You may be asked to verify that the manufactured product is worthy of approval, but un- less the problem is with the master you made, it isn’t your responsibility to fi x any problems. Th e wonderful world of audio is a constantly shift ing landscape of creative approaches, working procedures, formats, protocols, listening devices, and de- livery methods. Some of the information in this book will be dated almost im- mediately upon its release, but most of it will refl ect the audio creation and delivery terrain for a long time to come. Audio reproduction as a fundamental form of media expression is here to stay.,

Chapter 8 Three Best Practices E asy Ways to Raise the Level

of Your Sessions Th ere are three aspects to running recording sessions that are oft en inade- quately covered or completely overlooked in recording textbooks, yet these are vital to creative and productive work in the studio. Session fl ow, talkback opera- tion, and playback volume all contribute in some very obvious—and some not so obvious—ways to getting the most out of a recording session. Th is chapter explores these three elements from both the technical and the creative points of view. 8 .1 Session Flow Th e whole idea of “best practices” in running comfortable, creative recording sessions can be contained within the notion s ession fl ow. How is the session progressing? How is the balance between a relaxed atmosphere and focused work being handled? Are the musicians being given the opportunity to perform at their best? Are the goals being achieved? Is the work getting done effi ciently but with enough room for creativity? Th ere are four primary elements of inter- est in regard to session fl ow: the verbal, the technical, the musical, and the eco- nomic. Th e verbal r efers to what is said during a session—what is appropriate conversation, what is constructive feedback, and what might be best left unsaid. Thetechnicalinthis context means understanding how the technical demands of making a good recording may be balanced against the creative demands of making good music. Th e musical requirements of good session fl ow require an understanding of the fundamentals of music in ways that promote the most effi cient and creative recording sessions. And fi nally, one cannot escape the eco-, Th ree Best Practices nomic considerations that almost always form a backdrop to the day’s activities in the studio, even in the home studio.

Verbal fl ow

C onversation during a recording session is vitally important to session fl ow. To start, let’s distinguish between face-to-face conversation and conversation via the talkback system. Th e comments that follow may also apply to talkback conversations, but I reserve specifi c observations and advice about talkback for later in this chapter. Also, in many home studios, there is no talkback system because there is no separation between the control room and the recording room. F irst, the obvious: criticism needs to be constructive. Very general com- ments like, “You can do that better” are rarely helpful or appropriate. Even specifi c observations like, “You’re rushing the beat” can provoke a defensive reaction, whereas something like, “It feels a little rushed to me” or better still, “Does it feel a little rushed to you?”—which invites the musician’s input—helps to maintain a more relaxed atmosphere while addressing issues that may be important to getting the best possible musical performances. Rather than “Your 223 part is too busy,” you might invite input by asking, “Do you think your part would work better if it wasn’t quite so busy?” Comments or suggestions such as, “Can you try being more expressive with the lyrics?” get into emotional territory, as opposed to more objective mu- sical feedback. It’s much less complicated when you are making an observation such as, “You missed that chord change in the chorus” than when you want to get more or diff erent emotional content from a performer. I recommend going slowly with interactions regarding emotional content. Generally, it is advisable to develop a working relationship and get a feel for a performer’s creative pro- cess before getting into these sensitive areas. Once you’ve established a certain level of trust and respect, it may be possible and helpful to push a performer to- ward a deeper emotional commitment to the performance. Th is trust can build over the course of a single session, or it may require a few opportunities to work together before you can enter into delicate considerations of expression in per- formance. Again, it may be best to put these types of suggestions into the form of a question: “Do you think you can bring even more emotion to those verses?” M ore of the obvious: keep extraneous conversation to a minimum. Th ere is a time to tell a story or make a comment that is unrelated to the work at hand, but those times are pretty rare in the studio. Typical studio etiquette involves a brief preliminary chat before the session actually starts—about personal things or the weather or whatever—but once the work begins, it is important to stay focused on the music. Th is applies to the musicians as well as the recordist. Th e most frequent complaint I hear aft er recording sessions is that one of the participants talked too much. Th is doesn’t necessarily mean the person said too, THE ART OF DIGITAL AUDIO RECORDING many words, but it does mean the person interrupted the session fl ow too oft en with unnecessary and extraneous conversation. Be mindful of what you say during recording sessions. Positive feedback is a tremendous boon to performers. Some people describe record producers as cheerleaders, and honest enthusiasm can spur performance while creating a positive environment for creativity. If you work professionally, you may fi nd yourself recording music or musicians that you don’t feel much of a connec- tion with. You need to fi nd what is positive for you about the music and the performances, so that your enthusiasm can be honest. Lies told in the studio will ultimately be recognized, and dishonest enthusiasm is no better than si- lence, butifyou can’t fi nd something positive to say about virtually any music, then you probably shouldn’t be a recordist. A gain, criticism is essential—oft en it is the core of your job, if you are assuming production responsibilities—but it must be constructive, it must be specifi c, and it must be balanced with honest enthusiasm. 224 WHAT NOT TO DO Don’t allow strangers or guests at recording sessions unless you are sure that everyone involved wants them to be there. Playing and recording music is a very intimate process. It is important to be sure that only those whom everyone involved really wants to be at the session are in the room. Even if it seems like the performer is completely comfortable and relaxed, he or she may be unhappy about the presence of a particular person but not willing to speak up about it (especially in that person’s presence). Sometimes it falls to the recordist to ask a person to leave. In any case, carefully monitor who is in the room during recording sessions. If someone new walks into a session in progress, always make some contact with that person and try to ascertain that the person is welcomed by everyone involved. For many musicians there is considerable anxiety around performing in the recording studio. While encouragement is basic, there are specifi c tech- niques that are helpful in putting a performer at ease. Th e kind of self-con- sciousness that goes along with anxiety of recording may produce uncomfort- able or nervous performances. One of the most surprisingly helpful comments for a struggling, self-conscious performer is, “I can hear you thinking. STOP THINKING!” Th e humorous element soft ens the somewhat awkward request for less self-consciousness. Very oft en performers will recognize that they are “thinking” too much and that they just need to relax and play. It’s no accident that play is the term used for making music., Th ree Best Practices A long these same lines is the request for a musician, regarding the con- struction of his or her particular part, to “make it more boneheaded!” Th is is a comment born of the tendency for musicians to overplay—especially in the studio. Overplaying is usually a symptom of anxiety and self-consciousness. Again, humor eases the request for an altered approach to performance—a re- quest that may be interpreted as criticism. For those of us who work regularly with studio performance, the diff erence between a self-conscious performance and a comfortable one is usually apparent, though “usually” is an important qualifi er and sometimes reading performances is diffi cult. And, of course, there is much more involved here than a simple distinction between relaxed and overthought; there are considerations regarding musical execution and other subjective elements in judging performance. Nonetheless, a lack of self- consciousness goes a long way toward an outstanding musical performance, and the right feedback from the recordist can do a lot to keep the session fl ow positive and productive. WHAT NOT TO DO 225 Don’t be too eager to be sensitive to the performing musician. While much of healthy session fl ow revolves around being sensitive to a musician’s needs (“Can I get you some water?” or “Are you hearing yourself okay?”), it sometimes sends the wrong message if you are overly attentive. The classic example is asking, “Would you like a break?” too frequently. The intention may be to make sure that the musician is fresh and at his or her best, but the subtext is likely to be, “You’re not doing very well, maybe if you took a break you’d do better.”

T echnical fl ow

T echnical issues are important (most of this book is dedicated to them!), but it is helpful to keep in mind the true order of importance in regard to session goals. When it comes to the technical part of making a recording, I have one guiding rule: the music always comes fi rst! Th e primary goal is to encourage great performances, and aft er that comes the goal of making a great-sounding recording. If one must be sacrifi ced to the other, certainly it is the technical details that should be sacrifi ced for the sake of fostering the best possible per- formance. Th is dynamic between performance and recording technology fre- quently comes into play in subtle ways during the course of a session. H ow much tweaking of sound before recording begins is one of the issues that most frequently need to be balanced against getting the optimal perfor- mance. Here, the proper approach can be complicated and the recordist needs, THE ART OF DIGITAL AUDIO RECORDING to be sensitive to the musicians and the situation. Let’s say you’ve put a mic in front of a guitar amp and you’re getting ready to record some rhythm guitar. You ask the guitarist to play a bit of his or her part, and you listen to the sound that you’re capturing. You think it sounds a little thin, so you go and move the mic a few inches farther from the speaker. You come back into the control room and ask the musician to play the part again and you listen. You think it sounds better, but what if you pulled the mic another inch away? Well, maybe it would sound better still, but is the guitar player getting anxious to get started? Is he or she remaining focused on the music or becoming hyper-sensitized to the sound being produced? Th is is a judgment call on the part of the recordist. Is a slight improvement in the sound worth stretching the patience of the musician? And at some point you have to ask yourself whether the sound would actually be better, especially given how subjective the judgment of sound is. In terms of priorities, the search for the “perfect” sound should be placed well below the state of mind of the performer. Th at said, sometimes the reverse may be true. For some musicians (and guitar players are notorious for this), the pursuit of the “perfect” sound is a 226 major part of their pleasure in the studio and it is intimately tied into how they perform. If considerable time and energy is spent exploring the fi ner details of capturing his or her “sound”—swapping mics to fi nd the “best” one for the job at hand, using multiple mics, fi ne-tuning the mic placement, and so on, then the musician might feel inspired to perform better. Th e right balance between tweaking and getting on with the playing requires a subjective judgment, but the primary factor is the state of mind of the musician, not the actual diff erence made by small tweaks. Of course, certain technical matters require attention—a signifi cant buzz, a crackly cable, etc.—and there are times when these have to be resolved even WHAT NOT TO DO An anecdote regarding technical issues versus performance Recently I was standing on stage during setup for a performance by a well-known jazz musician. The person responsible for the live sound was setting up microphones on the drums as the drummer was warming up. I saw the sound person stop the drummer and ask him if he could move his ride cymbal up a bit so that a microphone would fi t comfortably beneath it. This is an example of a very bad job of balancing technical demands with a musician’s comfort. It is not appropriate to ask a musician to adjust his or her setup for the sake of technical convenience. The recordist’s (or live sound engineer’s) job is to create the most comfortable playing environment possible for the musician, and technical concerns should be addressed accordingly., Th ree Best Practices if it is inconvenient for the performer. However, fi nding the balance—knowing when to tweak and when to get started—is a very important part of the record- ist’s job. I t is diffi cult enough to play music; the technical elements should interfere as little as possible.

M usical fl ow

It will be diffi cult to record music eff ectively if you do not have some basic mu- sical knowledge. Most recordists have some musical background; you need to have an understanding of some music fundamentals, or you will not be able to do a good job with session fl ow. Knowing the fundamentals of musical rhythm, such as counting, bars and beats, and so on; essential songwriting terminol- ogy, such as v erse, chorus, bridge , and the like; and basic music theory, such as simple scales and chords, is essential to communication in recording sessions. When the performer says he or she wants to punch-in on bar 4 of the verse, you should be able to do that without further instruction. If the musician says “I want to take it from the modulation,” you should know where that is. If the band says they want to listen back from the turnaround before the guitar solo on a 12-bar blues, you should know where to start the playback. You don’t need 227 to know how to play an instrument, but you do need to know music basics so that you can communicate with the performers. Th ere are a variety of books to assist in this process (see especially E ssentials of Music for Audio Professionals, by Frank Dorritie). Besides being able to navigate to appropriate parts of the song based on the musical language, you need to bring some understanding of musical process to the task of making recordings. A key musical element in promoting good ses- sion fl ow is managing run-up time when doing punch-ins. Th is means know- ing the best place to start playback when someone is getting ready to replace one section of a recording (punch-in). If you start playback too far back, the musician may lose his or her focus by the time the punch comes and may play the wrong part, or a singer may lose his or her note, which means s/he doesn’t come in singing the correct pitch. If you start too close to the punch-in point, the musician or singer doesn’t have enough time to prepare, to fi nd the groove or the pitch reference needed for a good entrance. When I fi rst heard a very experienced background singer comment on how much easier a session had been because I was carefully managing the playback start time, I was surprised. I have since come to realize how much diff erence this makes in the comfort and performance of musicians, and thus in smooth session fl ow. I n practical terms, what is the correct amount of run-up time to a punch- in? Th is varies depending on the tempo and the preference of the musician, but a good guideline for a typical song would be a little more than one complete vocal line or a little more than two complete bars. Th is allows the musician or singer enough time to get oriented, without losing focus on what he or she, THE ART OF DIGITAL AUDIO RECORDING intends to do. If a singer can hear a complete vocal line ahead of his or her en- trance (which means you must start a few beats before the vocal line entrance for the singer to get oriented), that’s usually suffi cient time. If a musician can fi nd the beat and then count two bars to his or her entrance, that’s also usually enough time. If the tempo is fast or the music complex, sometimes a little more time is necessary—perhaps even two vocal lines or a bit over four bars. Some- time fairly early in the process you might want to ask the musician if you are using a good starting point for punching-in on a part. Some musicians prefer longer or shorter run-up times. Th e main lesson here is that this is an important concern when it comes to maximizing creativity, and along with a practical un- derstanding of music fundamentals, it is vital to good session fl ow.

Economic fl ow

Finally, you can’t ignore economics as an essential part of the recording process. Budgets and deadlines may be critical factors, especially when dealing with OPM (other people’s money) or with record-company release schedules. On the other hand, recording yourself at your home studio may make economic 228 factors virtually meaningless. I say “virtually” because even recording yourself at home has certain economic consequences. If you never get your recording done, you’ll never have the opportunity to see if it has any economic potential! Whatever the level of economic pressure, this does get refl ected in creative deci- sion making. All kinds of decisions, from what instruments to use on a song to how much time you take to complete a lead vocal may be aff ected by budget and timeline. Th e response to the same vocal take may vary from “Let’s do it a few more times and then we’ll put together the comp” (as described in section 4.2) to “Th at was good; we just need to fi x one part in the fi rst verse and it will be done.” It’s great to keep economics from dominating the creative process in the studio, but it’s not always possible. Budgets need to be clear, and it is important that economics are kept in mind from the very fi rst day of the project. Th e best way to keep economic pressures from seriously hampering recording sessions is through plenty of advance planning. From the very beginning, you should guard against allowing a project to fall behind budget without considering the consequences. 8.2 Talkback As discussed in the preceding section, good communication is key to good ses- sion fl ow, and at the heart of communication in many studio situations is the talkback system. Th e talkback system provides a way for those in the control room to communicate with those in the recording room. A talkback system in- volves a microphone that feeds the headphones and/or recording-room speak- ers when communication is desired., Th ree Best Practices WHAT NOT TO DO Don’t even use a talkback system if you don’t have to! That is to say, if you can work in the same room with the musicians, so that you can communicate directly, without any talkback system, this is the most comfortable way to work. A lthough n ot using a talkback system provides some distinct advantages, it is oft en not practical or not possible. Th at’s because bypassing a talkback sys- tem comes with the following potential problems: 1. Having a live mic in the same room with the playback speakers, which causes leakage onto the recording and/or feedback 2 . H aving insuffi cient space in the control room to accommodate the musicians and their instruments 3 . H aving fan noise or other external noises in the control room that compromise a live recording 229 A simple circumstance whereby you can bypass the talkback system is when the recording doesn’t involve any live microphones, such as when record- ing someone playing a synthesizer or recording a guitar or bass guitar using only a direct input (DI). In this case, headphones are not needed and it makes much more sense for everyone to be in the same room, thereby making com- munication easy. Another example is one in which the mic can be separated from the musician, such as when recording electric guitar with the musician in the control room and the amplifi er and mic in a recording room. Sometimes guitarists prefer to be in the same room with their amp (or need to be if they’re controlling feedback), but generally the ability to have direct communication, without talkback, makes having the guitarist in the control room the most de- sirable setup. But oft en talkback is necessary. And because communication is at the heart of good session fl ow, and good session fl ow is at the heart of a successful session, p roper talkback operation and etiquette are essential! I address the tech- nical issues regarding setting up a hardware and/or soft ware talkback system fi rst, and then take on some of the oft en-overlooked issues regarding talkback operation.

Setting up a talkback system

Th ere are many kinds of talkback systems, and they come (or can be built) with a variety of features and options. Generally speaking, if you are using a hard- ware recording console as part of your setup, you probably have a built-in talk-, THE ART OF DIGITAL AUDIO RECORDING back system. Th is means that there’s a small microphone built into the console that is activated by a talkback button. Th e button opens the mics routing into the main and/or monitor output, so that anyone who is listening through the mixing console can hear someone who is talking into the microphone. Some consoles have elaborate routing options for the talkback mic, allowing control over which users or which systems (headphones or speakers, for example) re- ceive the feed from the talkback mic. I’m not going to go into all the various console confi gurations here; you’ll have to consult your user’s manual for that. I do address the basic kinds of talkback systems that are available, including a look at some of the concerns of special interest if you are not using an outboard mixing console but only your computer and audio interface. Th ere are two kinds of talkback button operations: momentary and latch- ing. Sometimes you can select between the two. A momentary operation means that the button must be held down for the talkback microphone to be active. Latching operation means that pressing and releasing the button opens the talk- back mic and leaves it open until the button is pressed and released again. Th e incoming audio (from a singer’s mic, for example) must be managed 230 in some way because of the possibility of a feedback loop when using a talkback system. Th e feedback loop may be caused by the following: (1) the talkback mic in the control room is switched on by pressing the talkback button and the engineer’s voice is carried into the studio and is broadcast through the singer’s headphones; (2) the engineer’s voice leaks out of the headphones and is picked up by the singer’s microphone (which is typically only a few inches from the singer’s headphones), then that voice is broadcast through the control-room speakers (set to monitor the signal from the singer’s microphone); and (3) the sound of the voice coming through the speakers feeds back into the talkback microphone, creating a loop that runs continuously, building quickly into feed- back. It may sound pretty unlikely, but it is a pretty direct path from point A to point B to point C, and is a very common cause of feedback. Here’s the abbrevi- ated version of the signal path for the potential feedback loop: T alkback mic → Singer’s phones → Singer’s mic → Speakers → Talkback mic (feedback) I n order to prevent feedback, you need to manage what happens to the incoming audio when the talkback microphone is opened. Th ere are three pos- sibilities: muting the incoming audio, dimming the volume of the incoming audio, and leaving the incoming audio unaff ected. Each is considered in the following sections. Momentary systems that mute the incoming audio Th is is the classic talkback system, and the one that is most common on hard- ware mixing consoles. When the talkback button is depressed, the talkback mic is activated and the incoming audio is muted. Th is prevents any possibility of, Th ree Best Practices feedback by cutting the feed from the microphone in the studio to the speak- ers in the control room. It also means that when the talkback mic button is depressed, the musician in the studio cannot be heard. Th e challenge for the talkback operator is to switch the talkback mic on and off at the appropriate times—on to talk, off to listen. Th ere is more on dealing with this operation in the following section that discusses talkback operation and etiquette. Momentary systems or latching systems that dim the incoming audio Th e notion of dimming (decreasing the volume), rather than killing the incom- ing audio, is a relatively new development in talkback systems. Th e obvious advantage is that two-way conversations can occur because the incoming audio is not completely muted, as it is in the traditional system. Th e reason for dim- ming the audio is that it will (one hopes) prevent the feedback loop from de- veloping. If the audio coming through the speakers is soft enough, it will lose enough energy so as to be unable to make the complete loop back through the talkback mic, the headphones, and the singer’s mic. Th is generally works pretty well. As long as the sound is suffi ciently dimmed (some systems provide vari- able dimming) and the headphones don’t get too close to the microphone, an 231 open, two-way conversation may be possible. In this circumstance, a latching talkback button is convenient because it means that the operator doesn’t have to keep the button depressed during the two-way conversation. Th e operator must, however, remember to unlatch the talkback once ready to record, to avoid unwanted sounds (talking from the control room, leakage from the speakers, etc.) to be fed into the headphones during recording. Latching systems that do not aff ect the incoming audio A latching system that neither mutes nor dims incoming audio is likely to not really be a system at all, but the result of a typical talkback arrangement created when there is no hardware mixing console or the mixer doesn’t have a built-in talkback mic. In this case, you are simply connecting a microphone in the con- trol room and sending its signal out to whoever is on the headphone monitor system. Operation is controlled by muting or unmuting the channel that the mic is plugged into. Th e problem, of course, is the possibility of creating a feed- back loop. Th is is less of a problem if everyone is on headphones (typical of the one-room home studio setup, where the speakers are muted during recording to prevent leakage into the mic in the control room). A feedback loop is still possible in this circumstance, but less likely because it requires leakage from the operator’s headphones back into the talkback mic, rather than from the moni- tor speakers (which are muted). T alkback systems of this type, using soft ware only, is less than ideal. Not only is there no muting or dimming of incoming audio, increasing the likeli- hood of feedback, but also there is no physical button to push, so on/off op-, THE ART OF DIGITAL AUDIO RECORDING eration is controlled by a mouse click on the channel mute box in the soft - ware mixer. Th is can be awkward and slow, whereas talkback button operation should be easy and quick. Nonetheless, this can be workable if you are careful with the playback volume over the speakers (or if you’re all using headphones) and you are quick with the mouse. Th e nature of the on/off mute control is that it is the same as a latched button operation, and because there is no mut- ing, two-way continuous conversations are possible. Th ere is a soft ware plug- in available from that provides dimming capabilities for some computer systems and there are dedicated hardware talkback systems available from I would expect expanded options on this front in the days ahead. WHAT NOT TO DO Don’t turn on the talkback mic when there is a loop that will cause feedback! Unfortunately, this is easier said than done and all of us who work in this 232 fi eld have, at one time or another, inadvertently activated the talkback and been greeted with feedback that is highly annoying (not to mention potentially damaging) to the musicians wearing headphones at the time. It is appropriate to use extreme caution when fi rst operating the talkback during a session or after making changes in microphones and signal path. Try to check the talkback level before the musicians have put on their headphones (either by putting on the phones yourself and having someone else talk into the talkback or using an assistant as the guinea pig).

Using the talkback button

Th e heart of the talkback system is the button or switch used to open the signal path from the microphone that permits talkback. What do you need to know about using the talkback button? Believe it or not, s essions can sink or swim totally based on how eff ectively that little talkback button is used! Poor operation of the button can bog sessions down, cause miscommunication, and raise the frustration level so that little or no good work may get done. Th ere are two primary elements to good talkback operation. Th e fi rst is knowing when to turn the talkback mic on and off , and the second is the ability to operate the talkback button for conversations among multiple people in the studio and the control room. Th e fi rst may seem pretty straightforward— on when you want to communicate and o ff when you don’t—but it isn’t always quite that simple. You don’t know what people are going to say, and the people in the control room don’t always know whether or not you have the talkback, Th ree Best Practices button depressed. Th at means that someone might say something that he or she doesn’t want the musician in the recording room to hear, and the remark is accidentally heard. An off -hand remark such as “He never plays anything right all the way through, so we’ll just have to edit the pieces together” may be an ac- curate analysis of the situation, but may not be something the producer wants the musician to hear. As you might imagine, a mistake of this kind can blow a session (or an entire relationship). How do you know what someone else in the control room might be about to say? Th ere’s no way to know, but you can try to avoid disasters by making it clear to everyone when the talkback is on and when it isn’t. W hen you press the talkback button, you may want to say something right away, making it clear that you’re in communication with the recording room, and as soon as you think the communication is over, you should let go of the button or switch the latch to offsothat the mic is dead. Knowing when to turn the talkback on or off can be diffi cult—and it re- quires some experience to actually be good at it—but even more of a challenge is managing the talkback when there are multiple participants to the conversa- tion. In some circumstances, such as latched talkback operation, a multiperson conversation may occur without any special maneuvering by the operator, but with the most common kind of hardware system (a momentary system that kills 233 WHAT NOT TO DO Never let a musician endure silence after a recorded performance. The best way I can explain the above is to tell a story from early in my career. I was recording a vocalist (and a close friend), and she had just completed a lead vocal take in the studio. At the end of the take, I got involved in a brief discussion with another musician in the control room rather that responding to the singer’s performance. When I fi nally got on the talkback, the singer chewed me out: “Never let me stand here waiting for you after I’ve just poured my heart out,” she said. (Or something to that effect, maybe not using quite such polite language.) This made me realize what a serious mistake I had made. Performing music is very personal and often deeply emotional. If you are sharing in the experience as part of the process, you need to let the performer know that you are paying attention. Ever since that time, after any performance in the studio, I immediately get on the talkback and say something—even if it’s just, “That was good; give me a moment while I talk it over with x .” Never allow a musician to wonder whether you were even listening, or whether perhaps the performance had been so bad that you were at a loss for words. Even if that’s true, you must fi nd some words to reassure the musician that at least you are with the person and going to help him or her in the process of making a good recording., THE ART OF DIGITAL AUDIO RECORDING the incoming audio), this can be quite demanding. You have to anticipate the conversation as best you can, trying to switch so that each speaker can be heard (button down when the person in the control room is talking, button up when the speaker is in the recording room). One solution is to have multiple talkback buttons. Some studios have a talkback button on a long cable that stretches to anywhere in the control room and can be passed among people when needed. Th is requires some special wiring, and it doesn’t always work out if the speaker isn’t experienced in operating a typical talkback—people tend to forget that they can’t hear the other speaker until they let the button up. In any event, because communication is such an important part of the recording process, thoughtful operation of the talkback system is critical to good session fl ow. 8 .3 Playback Volume Controlling the volume of the playback is one of the most critical (and ne- glected) elements in running a productive recording session. Th e person con- trolling the playback volume is aff ecting the creative process in signifi cant ways, but oft en even the operator (probably you!) is unaware of the eff ect the playback 234 volume is having. Th e engineer is responsible for the playback volume that everyone hears in the control room (though that might just be you, if you’re working by yourself). Regulating playback volume is critical to session fl ow, to accurate listening for decision making, and to session fatigue. Finding the appropriate playback level requires a sensitivity that can dramatically aff ect both session fl ow and musical outcomes. Even if you’re not in control of the playback volume, you should still keep these things in mind and request diff erent listening levels when appropriate. Listening levels may need to be adjusted fairly frequently, depending on need. I cover the following six elements in considering playback volume during sessions. 1. Ear fatigue is an important consideration over the course of a session. 2. Quieter levels make pitch and rhythm accuracy easier to detect. 3. Louder levels make very high and very low frequencies easier to hear. 4. Loud levels are important for certain kinds of performances. 5. Controlling volume is an important part of the talkback/conver- sation matrix. 6. Everything sounds better when it’s louder!

Ear fatigue

While mental fatigue (lack of concentration) is the biggest challenge over the course of a long session, ear fatigue ranks a close second—and ear fatigue con- tributes to mental fatigue, as well. Your ears can take only so much sound over, Th ree Best Practices the course of a day. Persistent loud-volume listening will shut down your ear’s ability to hear, and eventually everything will start to sound muffl ed. But before things have gotten to that point, your ears will start to lose some of their ability to hear detail. I’m not going to go into issues about actual ear damage, which can be caused by very loud studio monitoring over extended periods of time, but even moderately loud levels sustained over the course of a day can cause ear fatigue, which really prevents you from being an eff ective listener. Y ou can have a SPL (sound pressure level) reader in the studio and be monitoring it for levels, but truthfully, I think we all know what loud is. It is more fun to listen louder, and I address this in the last part of this section, but eff ective listening requires low-level listening most of the time. Try to train yourself (and those you’re working with) to listen at pretty low levels. Knowing when to turn the volume up (again, covered in sections coming up) is also im- portant to workfl ow, but generally the problem is too much loud level listening. Th e key to low-level listening and prevention of ear fatigue is to start the day listening as low as you comfortably can. Your ears are very fresh, and you can listen at a pretty low level and still hear all the detail that you need, in most cases. Over the course of the day, there is going to be a natural tendency for playback volume to creep up, so by starting low you have the best chance of 235 preventing too much high-volume playback.

Quieter levels for detecting pitch and rhythm accuracy

I t is a little known fact, but quieter levels can greatly increase the productivity of your recording sessions. Volume tends to mask performance details. Problems with both pitch and rhythm tend to be much easier to hear when playback is low in volume. In fact, very low playback—lower than the comfortable listening level for most work—might best allow you to hear inconsistencies in pitch or rhythm. As volume increases, the ear hears more detail in frequencies (see section that follows), and this can distract from hearing small discrepancies in pitch or rhythm WHAT NOT TO DO Don’t listen very quietly if you have an overly nitpicking artist! Very low-level listening for checking performance details can backfi re if you are working with someone who is overly critical of his or others’ performances. If I’m working with someone whom I think is spending too much time trying to correct pitch or rhythm elements, I avoid the low-level listening technique because it can encourage obsessive correction. Although typically I turn the playback down, I have at times turned the playback up a bit when certain artists say, “I think part of that line is out of tune; can we listen again?”, THE ART OF DIGITAL AUDIO RECORDING accuracy. Very loud listening levels distract in many ways (and dulls hearing as well), so subtle detail in performance is easily missed during loud playback.

Louder levels for hearing high and low frequencies

Sometimes louder listening levels are necessary. As noted in section 2.5 regard- ing EQ, equal-loudness contours describe the ways that our ears begin to lose the ability to hear higher and lower frequencies as overall volume decreases. We detailed how this explained use of the smile-curve EQ application and the presence of loudness options on some playback systems. It also explains why we sometimes need to monitor fairly loudly. If you want to hear detail in very high or very low-frequency sounds, you need to monitor at a higher level so that your ear captures the details in the those frequencies. Usually this can be done for a relatively short period of time, as you fi ne-tune EQ or do critical level bal- ancing. Th e ear perceives frequencies at diff erent relative volumes depending on the overall listening level (this is described by the equal-loudness contours). To maintain perspective on your recording, you will want to listen at diff erent levels as part of your working process.

Loud levels for certain musicians

Th ere are circumstances when the playback level in the control room needs to satisfy the demands of a performing musician. For example, if you’re recording electric guitar with the amp isolated but the guitarist in the control room with you, then the musician is relying on the playback level for his or her perfor- mance. Th is is a diff erent situation from when the playback level serves only the recording process. In these situations, the musician should be allowed to control the level (not literally; you’re still operating the knob, but you’re asking for feedback on the level until you get it where the musician wants it). Th is does mean that sometimes the level will be somewhat louder than what you prefer, and this can cause ear fatigue (or worse), but it may be necessary to the process. If you know you’re going to be recording something that requires loud playback (rock guitar is a common example), try to schedule that for later in the day so you don’t have to do a bunch of sensitive work aft er having spent a few hours monitoring loud electric guitar. While fairly loud monitoring may be appropriate in some circumstances, it is n ot a ppropriate to allow the monitoring volume to get to the uncomfort- able level. If the musician keeps asking for it louder, beyond your comfort level, you can ask the musician to wear headphones so that you don’t have to monitor that loud. If that doesn’t work, you have the right to say that the musician will have to fi nd someone else to run the session if he or she wants it that loud—ul- timately, you must protect your ears. It rarely comes to that kind of confl ict, and musicians who ask for really loud playback will likely come to their senses if you suggest that you won’t be able to work at that volume., Th ree Best Practices WHAT NOT TO DO Don’t buy in to the argument that certain music has to be listened to loud all the time. Note that at the beginning of this section regarding loud monitoring levels I say “for certain musicians,” not for certain kinds of music. Some may argue that if you’re working with heavy metal, or dance-club music, or rap or punk, or whatever, that you need to monitor louder in order to capture the spirit of the music. This just isn’t true. More effective work gets done on every kind of music when reasonable monitoring levels are maintained. Sometimes loud listening is necessary, sometimes it’s fun (and that’s good, too), but it’s never appropriate all of the time .

C ontrolling volume as part of the talkback /conversation

matrix C learly, the volume of playback in the control room aff ects the ability for people 237 in the room to have conversations. Th is is something you need to be sensitive to as the operator of the playback level. It is oft en helpful to be listening to play- back and talking at the same time, and this requires a tightly controlled play- back volume—loud enough to hear the details but quiet enough to be talked over. Th is may be a diff erent level depending on who’s talking. It also may af- fect the playback duration—that is to say, monitoring conversations also means deciding when playback should be stopped because the conversation has over- taken the listening. When done with sensitive attention to the situation, this change is oft en transparent to the other people in the room. Th ey don’t even notice that the playback has been turned down or stopped, because they’re having a conver- sation. But it allows that conversation to happen and for the creative process to move forward, as opposed to a situation in which people keep raising their voices to be heard until they have to ask for a lower volume or for you to stop the playback so they can start their conversation over again. Th is wastes time and causes frustration—both negative outcomes in a recording session.

Everything sounds better when it’s louder!

Th is is generally true (up to a point), and it’s part of the constant struggle to be really creative while making recordings. We saw this when it was applied to mid- to high-frequency EQ (section 2.5). Because that kind of EQ adds appar- ent volume as a part of the frequency boost, there’s oft en an initial response of “Th at sounds better,” which can lead to over EQ-ing. If you want to get more of a kick out of what it is you’re recording, turn it up! But, the problems in, THE ART OF DIGITAL AUDIO RECORDING doing this are many, as just described: ear fatigue, inability to make accurate judgments about pitch and rhythm, and an environment in the control room that makes communication diffi cult. Loud playback has its place, and at the end of the session you might want to do some pretty loud listening, just for fun. But ultimately, playback level is a tool and it must be used to further the session’s goals.,

Addendum 1 H ow to Walk into a Commercial Studio

and Be the Engineer My ascent to the ranks of professional recording engineer was, in brief, as fol- lows. I had been a professional drummer for a few years and had the chance to do a bit of recording as the drummer in various bands I had been in. In 1979, I acquired one of the fi rst Tascam 144 cassette 4-track tape recorders and it changed my life. I started making recordings and learning the intricacies of this amazing little recorder/mixer. Although it isn’t quite true that “Everything you need to know about recording you can learn on a cassette 4-track,” it is amazing how close to the truth that is. Th at little machine had output faders, pan con- trols, EQ, aux sends and returns, and various I/Os on the rear panel, including inserts. With it you could do overdubs, punch-ins, and bounce tracks. It was a 239 miniature version of an entire multitrack recording studio. I graduated from the cassette 4-track in my living room to an Akai 12- track in my garage. Th e 12-track was also an all-in-one recorder/mixer that had the expanded capabilities aff orded by the extra tracks. I began recording band demos in my garage for next to nothing. One day, one of the bands I was work- ing with said they had cobbled together enough money to go into a commercial studio to do a recording, and they wanted to know if I would come with them and be the engineer/producer. I said yes, though I had never been an engineer at any studio other than the ones in my home. I successfully got through the session and my career as a professional recordist was offi cially launched. F or many people, including a fair number of those reading this book I would guess, the idea of going into a commercial studio and being the engineer is just too intimidating to consider. Even though you’ve been running home stu- dios for years, and are really good at making everything work for those “home” recordings, the idea of being the engineer at a studio that has an unfamiliar mixing console or control surface, patch bay, microphones, and outboard gear seems out of reach. I want to encourage you to expand your notion of what you are capable of. Th e secret to being an outside, guest, or “independent” engineer at a com- mercial studio is that you’re not expected to necessarily know the intricacies of the particular console at that studio, and you’re certainly not expected to do the patching in the patch bay. Th is is why you are assigned a second or assistant engineer for your session. Commercial studios have to provide someone who, THE ART OF DIGITAL AUDIO RECORDING knows the ins and outs of their particular setups, including the functions of the console. Th e assistant is also expected to do all the patching. Th e person is there to answer your questions and to make sure that everything is working for you. Asaguest engineer, your job is to know what it is that you want to do, not exactly how to do it at the particular studio you are at. You need to know mostly all of the basics that I cover in this book, but none of the specifi cs of implemen- tation at a particular studio. It’s perfectly acceptable for you to ask about the microphone input or the bussing system of the studio’s console. Of course, you need to know the general principles behind getting from the mic to the console, and the basics of proper gain structure in doing so, but you can certainly ask for the specifi cs or ask the assistant to set up one signal path on the console so you can see how the routing works. It’s perfectly fi ne for you to ask the assistant to suggest a microphone for a particular application. It’s not possible for anyone to be familiar with all the microphones available. Even with familiar mics, it’s a good idea to fi nd out from the assistant which ones are considered particularly good at that studio, espe- cially for critical recording functions such as vocals. Th e main point is this: as a guest engineer at a commercial studio, your job is to know what it is that you want to do from a technical standpoint and 240 the basics of how such a thing is done, without necessarily knowing any of the specifi cs as to how that is accomplished at the particular studio. If you’re ac- complished at making your home studio work, then you already know what it is you need to do and you’re ready to be a guest engineer at a commercial studio. Yes, you might require more assistance than a more experienced engineer, but you will be able to make the session happen and fulfi ll your role. Aft er a few sessions, you will fi nd it rather easy to adapt to a new console or control surface and a new work environment. Th e principles are always the same—the specif- ics of signal-path routing always follow the same basic concepts. What’s more, the whole notion of what a recording session is—from setup through line tests, to recording and reviewing recordings, to working through all the creative and technical issues that result in getting the work planned for the day done—is the same in the general sense, no matter what studio you’re at. And the assistant engineer is there to help you through the details. Th e one exception I would make is in regard to the DAW. Pro Tools re- mains the default standard DAW for commercial recording studios. Just as the 2-inch 24-track tape recorder was the standard previously (and they continue to sit in the corner of the control room or in the machine room of most com- mercial studios), Pro Tools is now the only piece of technology that is reliably found in almost every commercial studio around the world. For this reason, I highly recommend to any of you who aspire to work in this fi eld commercially that you acquire and learn Pro Tools. You will be expected to know how Pro Tools works in order to be an eff ective guest engineer at most studios. Asking questions about the console, control surface, patch bay, or studio confi guration, Addendum 1 would be expected of a guest engineer. Th ere are some Pro Tools questions that would be expected as well; resolving the I/O setup between a fi le that you bring from home or from another studio, and the I/O confi guration at the studio you’re working at, is something that you may well need the assistant to do for you. But in general, just as you will need to know what has to happen in order to have a successful session, you will need to know how to operate Pro Tools suffi ciently to be running the program as part of that session. H aving the confi dence to take your sessions to studios outside your home/ project studio expands your capabilities enormously, giving you the opportu- nity to try out new gear and new recording spaces, learn how other studios operate, and meet people in the recording community as colleagues and peers. Th e fi rst few forays are likely to make you a bit nervous; if you do it with your own band or project you might feel a little less pressured than if you’re working for someone else, but I encourage you to take the plunge. In many cases, you will fi nd that you’re more ready and more capable that you realize—all that time in your bedroom, living room, or garage really is closely akin to the way record- ings are made in all places around the world.,

Addendum 2 Researching and Buying Gear I nternet vs. Brick and Mortar

B uying recording gear (and by this I mean both hardware and soft ware) is oft en an obsessive and confusing endeavor. Th e Internet is a spectacular resource, but it also removes us from the ability to get in there and muck around with things. What follows may be obvious to those of you with a lot of experience already, but I am responding to a lot of questions that I get from students about the best way to make decisions and, ultimately, to purchase new (or used) gear for their studio. What Do You Need? 242 Although this is one of the most common questions I get about gear, it can also one of the most diffi cult to answer. It’s pretty easy to answer this question if it is regarding a particular studio function that you wish to have. Do you need a microphone? If you’re going to record vocals, for example, then of course you do. Th ere are basic items that you need to make your studio a studio. But, in fact, there are a lot of diff erent ways of working and of creating diff erent kinds of music; you may not need a microphone at all if you are doing all instrumental, all electronic music. DoIneed a control surface? Do I need a large-diaphragm condenser mic? Do I need an impulse response reverb plug-in? Th ese questions are more dif- fi cult to answer. You probably don’t absolutely need any of these things in order to get your work done, so it’s a question of quality or convenience, and these questions usually don’t have clear-cut answers. You may want these things, and they may improve the quality of your recordings or the convenience of your work environment, but there is an endless list of things that can improve the quality of your recordings and make your work easier to do. Where do you draw the line? W ell, budget is the great limiter. You need to be able to aff ord new gear, or justify it based on the income profi le of your studio. Clearly, I can’t make these judgments for you, but I can off er a bit of advice on studio upgrade decisions. Th e fi rst consideration is this: every link in the chain—in the signal path—is critical, so buy gear that is appropriate to the weakest link or upgrade that weak- est link. Th at means that if you have an inexpensive mic preamp and less than, Addendum 2 high-quality analog-to-digital conversion into the computer, you shouldn’t buy a $5,000 microphone. Buy a mid-quality mic—in the $500 to $1,000 range— that will hold up until you upgrade the other elements in the signal path and it becomes the weakest link. Perhaps then you’ll be ready for a more expensive mic. If you have some very high quality gear in a signal path with low-quality gear, you are not getting the most benefi t from the good stuff . I oft en tell people that you geta5to 10 percent improvement in quality for double the price. Of course, this is not literally accurate, but it points to the fact that upgrades in quality can oft en be very expensive without bringing vastly noticeable results. Sometimes the results from individual upgrades can be very apparent. For example, a diff erent kind of microphone that is better suited for certain tasks—say, a good-quality condenser mic when you only had dynamic mics previously—can result in a signifi cant change in the quality of your re- cordings. If you upgrade each element of your signal path by 10 percent, the dif- ference can be quite apparent, but also quite expensive. In any event, chose your upgrades carefully to maximize the benefi ts. Th ere is more about the specifi cs of deciding what to buy in the following section on research. Research: Try Before You Buy or Rely on Word-of-Mouth? 243Isit possible to buy gear successfully based completely on word-of-mouth, without ever trying the gear? Yes, although this is not the most desirable way to buy. Is it okay to buy gear that you’ve tried out at the store or used in a session at somebody else’s studio? Yes, but again, this is not the best way to make buy- ing decisions. Ideally, you use a combination of “word-of-mouth” research and some hands-on experience. I put word-of-mouth in quotes here because the In- ternet provides the opportunity for getting a lot of written user feedback—not exactly word-of-mouth, but a close equivalent. Th e problem with Internet research, as well as recommendations from friends and colleagues, is that not everyone has the same response to gear. What sounds sweet and warm to one person may sound relatively harsh and cold to another. By the same token, your hands-on experience with a piece of gear in an unfamiliar environment, like a store or someone else’s studio, may produce a somewhat diff erent response than your reaction to that same gear when you have it in your own studio. A nother problem with Internet research is the sheer bulk of information out there. You can fi nd contradictory opinions about almost anything, and it can be diffi cult to sort out the valuable information from the casual, and sometimes simply wrong, comments. If you research gear consistently over time, you will probably fi nd some sites and/or reviewers whom you trust. Th ere are moder- ated discussion groups, free-form discussion groups, blogs, reviews as a part of commercial Web sites where the gear is being sold, and random reviews. Nega-, THE ART OF DIGITAL AUDIO RECORDING tive reviews can be particularly helpful in balancing what tends to be primarily positive comments—apparently people are more motivated to sing the praises of their new acquisitions than complain about them. Th is is probably motivated in part by a desire to justify a new purchase. In any event, don’t let a few nega- tive reviews scuttle the deal—otherwise, you’ll never get a ny new gear—and don’t let a few over-the-top raves convince you that you have to have something. Read enough comments and reviews until you feel as if you have a fairly bal- anced understanding of how people feel about the gear you’re researching. Pay attention to how they are using the gear and what their studio environment is to see if it matches your needs and interests. In some cases, most notably with plug-ins, you have the option of try- ing before buying. Th is is the best possible situation because you get hands- on experience in the studio environment where you are most comfortable and where you’ll actually end up using the gear. Almost every plug-in company of- fers free trials of all their plug-ins, either on a time-limited basis or with some of the functionality disabled. Th ese represent your best opportunity for making a purchase that you’re going to be happy about. For hardware purchases (but typically not for soft ware), most stores off er a return option, though returning things can be a hassle. Th is brings us to the fi nal topic in regard to buying gear: 244 where to buy. Buying: Store versus Internet versus eBay W here to buy is complicated by several factors, including price, convenience, and return capabilities. Th ere are advantages to buying from your local dealer, most notably ease of return, but it’s also positive to support your local record- ing community and the gear dealers are an important part of that community (though your local store may be a part of a large, national chain). Th ere are a lot of Internet stores that sell gear; some of them also have brick-and-mortar stores. For those not located near physical stores, Internet shopping makes pretty much everything easily available, and many of these dealers have gen- erous return policies, as long as you’re willing to deal with the repacking and return shipping chores. My preference is to shop at my local independent audio gear dealer. I am fortunate to have a very good one in my area. Ideally, the salespeople at your dealer are not paid on commission, and are therefore less motivated to sell you as much gear at the highest price point possible, and are also more likely to take the time to help you fi nd what you really want—and even to save you money where possible—on the understanding that you will become a long-term cus- tomer. Nonetheless, I shop the online stores and eBay to see what prices are like before I buy from my local dealer. I won’t necessarily demand that they match the lowest price out there, but I don’t want to pay a large premium for shopping with my local dealer., Addendum 2 Th ere is a huge amount of audio gear available on eBay, and it is a good place to research prices. It is also a great resource for buying used gear, but that is a specialized market. I do not recommend buying used gear on eBay unless you have a lot of knowledge about the gear you’re buying and are an experi- enced eBay user who feels that you know how to use the system to judge the likely trustworthiness of the seller. To its credit, eBay has made a huge amount of used and vintage gear available to people around the world that would other- wise have had great diffi culty in fi nding it. To be confi dent in buying on eBay you need to read and trust the feedback system. You also need to explore the feedback content, as there are some un- scrupulous sellers who sell a bunch of cheap items to build up positive feedback and then sell one expensive item that is never delivered. Th us, eBay has done more to guard against fraudulent sellers over the years, but scams still happen. You also need to be able to trust the products because, in general, returns on eBay will be more complicated or impossible, so if that’s a concern, you’re much better off with a real or virtual store. Th at said, eBay oft en has new or nearly new items at the best prices. Th at’s because some items being sold are gift s that people received and never used or that were used very few times and then aban- doned, thus selling for well under the price you would fi nd anywhere else. Th ese items may not be returnable, so, again, you have to trust the seller and the prod- 245 uct to buy under these circumstances. Also, sometimes the best price for a new item on eBay is more than the price of the item through a normal retailer. Just because it’s on eBay, that doesn’t mean it’s cheaper than from the alternatives. Buying audio gear is a joy and a disease. New gear can stimulate the cre- ative process, as well as allow for higher quality work, but endless gear research, purchases, and learning curves can become a distraction from making record- ings. Plan carefully, shop wisely, and take some breaks from the endless cycle of upgrading.,

Appendix D igital Audio Formats, Delivery, and Storage

Of all the sections in this book, this may be the most diffi cult one to keep up to date. Digital formats are a constantly shift ing array of fi le types, sampling rates, and bit depths. Audio delivery demands fl uctuate, depending on the ultimate use for the audio, and the same audio may need to be delivered in a variety of formats for a variety of uses. Digital audio storage options are constantly ex- panding, but questions of compatibility and longevity remain as potential prob- lems with storage and archiving. Th e following is certainly not exhaustive, but it provides a primer for both technical and practical considerations at the time of this writing. 246 Digital Audio Formats—Recording Audio recording formats diff er primarily in their bit rate and sample depth. You may think of digital audio as the computer-language equivalent of tak- ing a picture of audio content. Digital audio formats will vary based on the amount of information contained in each picture (bit depth) and the number of pictures taken per second (sampling rate). CD audio is set to a bit depth of 16 and sampling rate of 44.1 kHz. Th is means that each “picture,” or each sample of audio that is converted into digital code from the original analog sound, contains 16 bits of information. In computer language, “16 bits” refers to 16 ones or zeros, each one counting as one bit. Th e number of “pictures” or bytes of information used to create CD audio is 44.1 kHz, which means there are 44,100 lines of 16 ones and zeros used to describe each second of digital audio contained on a CD. Early digital recorders used lower bit depths and sample rates to record audio, but with the advent of the ADAT format, multitrack tape-based systems that were roughly equivalent to the CD standard came to be widely used (16- bit, 48 kHz). Computer-based systems (DAWs) also used something akin to the CD standard, but it was the migration of DAWs to a 24-bit format that was critical to their widespread acceptance as the recording devices of choice. Th ough the fi nal audio program is oft en reduced back to 16-bit for CDs, or even lower resolution for mp3s and other formats that use compression to re- duce fi le size, the 24-bit standard allows for much greater detail than 16-bit in the initial recording. Soft ware engineers have found a variety of techniques to, Appendix take advantage of that detail in the fi nal conversion from 24-bit to lower reso- lution formats. S ample rates above 44.1 kHz are available in many DAWs, and recordists vary in their use. Th e 48 kHz was the digital standard for high-quality audio before the CD standard was accepted, so it remains an option on most DAWs. While 48 kHz off ers the benefi ts of slightly more information per second, it has the disadvantage of requiring complex conversion to get to 44.1 kHz if the fi nal delivery is going to be for CD production. Some engineers chose 48 kHz none- theless, but I prefer to record as 44.1 kHz to avoid the sample rate conversion when the program material is prepared for CD manufacture. S ample rates of 88.2 kHz, 96 kHz, 176.4 kHz, and 192 kHz are available with some systems and are used by some recordists all the time and by oth- ers for specifi c projects. Th e advantage is greater detail, although listening tests seem to indicate a pretty modest improvement—as opposed to the dif- ference between 16-bit and 24-bit audio, which sounds like a dramatic shift in detail to most professional participants in critical listening tests. In general, program material with a lot of very complex harmonics and great dynamic range—such as solo piano, string quartet, and the like—will benefi t more from these higher sampling rates than dense material such as found in most popular music. Th e higher sampling rates also require a lot more processing 247 power for running plug-ins and fi les with complex automation, and they need twice or four times as much disc space to store the audio. For these reasons, I fi nd most recordists on most projects using the 24-bit, 44.1 kHz audio fi le format. Audio fi les also require a certain amount of nonaudio information, gen- erally contained as header information that precedes the actual bits and bytes of the audio that has been converted from analog to digital. Th e nature of this header information, and the format used to deliver it, is what diff erentiates fi le types such as Wave fi les and AIFF fi les. Th ere are many other fi le protocols, such as red-book audio for CDs and orange-book audio for CD-Rs, and there are other DAW recording formats—mostly legacy formats like the Pro Tools Sound Designer II fi les—but Wave fi les and AIFF fi les dominate the DAW re- cording landscape. Wave fi les use the .wav appendix and AIFF fi les use the .aif appendix. Because Wave fi les went through a variety of forms, there has been a move to standardize the Wave fi le format under the name Broadcast Wave Format that uses the .bwf appendix. Th e main advantage to the Broadcast Wave Format is the inclusion of metadata, including a timecode stamp. Th e inclusion of the timecode stamp with the audio allows you to import audio from one DAW to another while maintaining the correct audio region locations. Despite the diff erences, most DAWs can recognize and utilize any of the variations in Wave fi les. I n general, using .wav or .bwf fi les for your recordings is the best idea, as it gives you the most widespread compatibility across systems. However,, THE ART OF DIGITAL AUDIO RECORDING AIFF fi les are required in certain delivery situations, such as for many DVD authoring houses, because some popular DVD authoring programs recognize .aif fi les but not .wav fi les. Many DAWs can handle mixed fi le formats (e.g., some .wav fi les and some .aif fi les), though not Pro Tools, which requires a single-fi le format for each session. In any event, almost all of them can con- vert from one format to another if you need to do this for production or de- livery purposes. Digital Audio Formats—Consumer Th e fundamental information regarding digital audio formats for consumers remains bit depth and sampling rate. As described above, CD players use a 16- bit, 44.1 kHz format—or 44,100 16-bit samples every second—to decode the audio program. Th at’s a lot of ones and zeros, but a second is a long time in musical terms (oft en two or more beats) and sound is complex. Whether or not the CD standard does an adequate job of defi ning audio detail has long been debated. Certainly, soft ware engineers and recordists have found ways to pack more detail into the CD audio format. In any event, consumer formats with more and less detail proliferate, but it is the format with considerably less detail, the mp3, that threatens to overtake (or already has overtaken) CD audio as the 248 new standard audio format. Th e mp3 format uses a variety of sophisticated techniques to try to re- tain as much fi delity as possible while reducing the fi le size considerably from the CD standard—typical mp3 fi les are about 1/10 the size of their CD audio equivalent. Unlike CD audio, mp3s may use a variety of bit rates and sampling rates and can still be read (played) by an mp3 player. Th e standard for mp3 is a bit rate of 128 kbps and a sampling rate of 44.1 kHz, but there are many lower and some higher resolution options available. Th ere are also a variety of encod- ing schemes available. M p3s became very popular because they allowed audio to be transmitted and downloaded relatively quickly over the Internet. As Internet connections and computers have gotten faster, the options for downloadable audio have in- creased and we are seeing more and more options for higher quality audio, including audio in the CD format, available for purchase and download. Commercial audio formats that provide higher resolution fi les than the CD format have been developed, but none has found much traction in the mar- ketplace. Competition between formats such as DVD-Audio (DVD-A) and Super Audio CD (SACD) hasn’t helped higher quality audio fi nd a consumer base. Surround sound (5.1 audio format) has found a large user base for home theater use, but it has yet to attract much interest in audio-only formats. Audio professionals need to be familiar with surround-sound audio-delivery formats (below) if they work on sound for fi lm, video, computer games, or other sur- round-oriented consumer products., Appendix Digital Audio Delivery Th e best method for delivering digital audio depends on its ultimate purpose. Here, I cover delivery for CD mastering, CD manufacturing, Internet applica- tions, fi lm and video applications, and video games. In many instances, it will be necessary for you to talk with the person who will be working with the audio that you are delivering, as diff erent applications require diff erent audio formats even though they may ultimately be put to the same use (e.g., stream- ing audio over the Internet can use a variety of source fi le formats, but the particular Webmaster you are delivering to may require a certain format for their application). Delivery for CD mastering Although diff erent mastering engineers and mastering houses will want dif- ferent fi le formats, depending on the programs they are running, there are two primary considerations for how to deliver your mixed master to the mastering engineer (even if you are the mastering engineer, too). Th e fi rst is to provide the highest quality fi le format possible. Th is generally means maintaining the bit depth and bit rate that you used for your individual fi les before creating the mixed master. If you recorded at 24-bit, 44.1 kHz (as I usually do), you will want to deliver your mixes in that same format, if possible. If you recorded at 48 kHz 249 or at a higher sampling rate, you will want to maintain that sample rate as long as you’ve cleared the format with the person who will be doing the mastering. One of the keys to providing the highest quality fi les is to do as little fi le conver- sion as possible prior to the mastering stage. Th e fi nal CD master will have to be 16-bit, 44.1 kHz, but assuming you started with higher resolution fi les, conver- sion to this format should be postponed until the last stage of fi le processing. Th e second requirement is to provide fi les without any brickwall limiting. Because brickwall limiting has become such a prominent part of fi nal music delivery to the consumer, and because it aff ects the sound so dramatically, I fi nd that I must complete my mixes using a brickwall limiter so that I can hear the likely eff ects of its use. However, in mastering I deliver (or use myself, if I’m doing the mastering) my fi nal mix with the brickwall limiter removed so that it can be added back in as the fi nal processor before creation of the mastered mix. If I’m delivering fi les to a diff erent mastering engineer (not doing it myself), I provide a fi le without brickwall limiting for use in the mastering, but I also pro- vide a version with brickwall limiting so the mastering engineer can hear what I consider to be the actual sound of the fi nal mix. D elivery for CD manufacturing If you are doing the mastering for CD release, the master you deliver will be a CD-R that is an exact version of the way you want the manufactured CD to sound and play. Along with the music, mastered with all the processing and, THE ART OF DIGITAL AUDIO RECORDING sequencing issues handled just as you want them, the CD-R should contain the metadata that the artist or record company want encoded along with the disk. Typically, this includes the title of the CD, the artist’s name, and all of the song titles along with the ISRC codes that I discussed in the chapter on mastering. Some CD-R burning programs allow you to print out a document that contains the critical information regarding the timing and encoding of the burned mas- ter. Manufacturing houses like to see this document to confi rm what they are reading from the CD-R master, but it is not essential and most manufacturers will accept masters without the printout. It is important that you have given your CD-R master a careful listen to make sure that it doesn’t have any fl aws that might have come from a poor burning run or a faulty CD-R. In terms of burning protocols for CD-R masters, there is only one essen- tial and that is that to use the disc-at-once (DAO) burning protocol and not track-at-once (TAO). TAO has become rare, and some burning programs no longer even off er it as an option, but you should check to make sure that you are burning DAO. In terms of what brand of CD-R medium to use and what speed to record at, the opinions vary, but independent lab tests have not shown that recording at slower speeds or using higher priced “premium” CD-Rs pro- duce better results. In fact, in some instances, faster record speeds produce 250 discs with fewer errors. In most cases, almost any CD-R medium and burn speed will produce error rates well below anything near a danger level that would produce any negative results when used for manufacturing. Th e best advice is to fi nd discs and burn speeds that work well for your burner and use those as your standard. Delivery for Internet applications Th e ultimate fi le format that will be used for Internet applications may vary widely, but the delivery fi le is most frequently an mp3 which is then converted or reprocessed as needed by the Webmaster. Protocols for downloading and streaming vary, and the Webmaster may ask for fi les in a variety of formats as well as mp3s, including mp4’s, RealAudio, and/or QuickTime Audio. If you are delivering audio for these kinds of applications, you may need to invest in soft - ware that will convert to a variety of formats, or you can ask the Webmaster if they can handle the conversion for you. I always try to deliver the audio in the CD format as well, so that the client has this on fi le for reference or for use in later applications where higher quality audio can be used. M any of these Internet fi le protocols, including mp3s, contain more en- coded metadata information than a CD-R. A musical category can be desig- nated, which will enable the music to be sorted and potentially recommended in consumer searches. Information about the original CD release, number of tracks, position of this track in the sequence, whether the CD was a compi- lation, and so on, can be included with the fi le, as well as have a link to the, Appendix artwork if this has been posted at a particular Internet address. I expect digi- tal fi le formats to continue to add metadata capabilities to further integrate music tracks into the media datastream that is contained on an individual’s computer. Delivery for fi lm and video Audio for fi lm and video may require synchronization with the visual elements. Obviously, dialogue requires sync, but so do most sound eff ects and music cues. In order to work eff ectively to picture, you will need to import a movie fi le into your DAW. Th e movie fi le should be a “window dub,” which means that the SMPTE timecode location number has been burned into a small window at the bottom of each frame. Establishing and maintaining sync through the use of timecode is beyond the scope of this book, but a few words about fi le formats may get you started with understanding the requirements for this kind of delivery. Audio that accompanies picture may end up in a variety of formats, from VCR tapes to big-screen movie projection, but the most common delivery for- mat right now for picture with sound is DVD. In any event, the fi le format that will be required will vary depending on which editing and/or authoring program is being used. Surround sound (typically 5.1 surround) is increasingly common for fi lm and video, so you may need to supply both stereo and sur- 251 round audio fi les (see below regarding the surround format). You will need to work closely with the other content providers, including the authoring, editing, and packaging people, if you are providing sound that is to accompany visual elements. Delivery of surround-sound fi les S urround comes in various formats, but the dominant format is 5.1 surround, made up of left , right, center, rear left , rear right, and LFE (low-frequency ex- tension) channels. Th e rear channels are oft en referred to as the “surround” channels—they feed the “surround” speakers in back or to the sides of the lis- tener. Th e LFE channel is, in fact, a distinct channel, so there are actually six channels of audio, but because it is not full frequency (carrying only subwoofer information—typically from about 90 Hz and below), it is referred to as the .1 channel of 5.1. Format requirements for delivery of 5.1 audio may diff er, but the standard is 48 kHz, 16-bit AIFF fi les, as this is what is used in the most prominent au- thoring programs. Surround for DVD will be encoded as an AAC fi le for Dolby or some other codec for a diff erent surround format, such as DTS. Usually the audio person supplies the 48 kHz, 16-bit AIFF fi les, and the encoding is taken care of at the DVD authoring stage. If you are required to supply encoded fi les, you will need to get either a program that does the encoding or an add-on for your DAW that allows you to do this encoding within the DAW., THE ART OF DIGITAL AUDIO RECORDING Th e standard order for 5.1 fi les is as follows: C hannel 1: Front left C hannel 2: Front right Channel 3: Center C hannel 4: LFE Channel 5: Rear left C hannel 6: Rear rightItis critical that the fi les be in this order for them to encode properly. Delivery for video games Formats for delivery of audio for video games may vary, but it is likely that you will be asked to deliver a stereo mix, stems (described below), and possibly a 5.1 surround mix. Because video games require so much music to accompany the many hours of game play, each audio element may get used in diff erent ver- sions at diff erent times. In order to do this, stereo stems are made, taken from the fi nal stereo mix. A stem is simply an element taken from the larger mix of the composition; taken all together, the stems recombine to form the original composition and mix. A typical group of stems might be broken down as fol- 252 lows: drums, percussion, bass, guitars, and keyboards. In this case, there would be fi ve stems. More complex compositions may require more stems, such as drums, high percussion, low percussion, bass, rhythm guitars, lead guitars, horn section, piano, keyboards, lead vocal, and harmony vocals—making a total of 11 stems. Once the fi nal mix is done, stems are made by simply muting all other tracks and running a “mix” of each particular stem element. Again, in all of these collaborative projects that combine audio and other elements, you will need to coordinate your work with those working on other parts of the project. Digital Audio Storage Hard drives have become the primary medium for audio storage. Th e key hard- drive confi guration options are computer interface, size of drive, speed of drive, drive buff er size, and drive bridge. Th ere are new developments regarding each one of these drive options so frequently that the following information can be used as a guideline, but you may need to do additional research to determine your best options at any given time. For audio storage, it is best to use the fastest available interface, though of course both your computer and your DAW must support it. Th e most common interfaces are USB-1, USB-2, fi rewire 400, and fi rewire 800. Th e fi rewire 800 connection will be the fastest, and should be used when possible. SATA drives, which are replacing the traditional ATA/IDE drives in many new computers, use a new interface protocol called eSATA (external Serial ATA), which is faster, Appendix still. A USB-1 interface is not fast enough to handle typical recording require- ments; it can be used for storage, but not for recording. Hard-drive storage sizes continue to expand, and to get cheaper and more readily available. It can be problematic for a computer to manage very large drives; the hardware and operating systems don’t always keep up with the latest in available drive capacities. However, drives as big as a terabyte (1,000 giga- bytes) are becoming common, reasonably priced, and can be managed by most recent model computers. Because audio requires quite a bit of storage space, and because you get bigger drives for comparatively less money, the big drives represent good value for audio storage. A terabyte drive might hold as many as 10 complete, typically sized CD projects or more (depending on how much audio was recorded for each project, of course). Th ere are portable hard drives (3.5-inch drives) that are powered from your computer (buss powered, meaning that no AC is required) and may con- nect via USB or fi rewire. Very small USB fl ash drives have become common. Th e fl ash drives currently come in sizes up to 256 GB (gigabytes), with larger models on the way. Th ese little drives are inexpensive and fi t in your pocket— great for transporting data, such as grabbing fi les of a single song to move from the studio to home. You may be able to eff ectively record on a fi rewire portable drive, but it is unlikely that you will be able to record to or play back from a fl ash 253 drive that uses current technology. Drive speed is an important factor in allowing for large quantities of data transfer as is required for large audio sessions. Older drives and some of the portable drives spin at 5,200 or 5,400 rpm, and this can create problems with larger fi les. Drives that spin at 7,200 rpm are much better suited for audio. Th ere are a few drives running at 10,000 rpm, but this is not necessary for even very large audio fi les. Th e newer Solid State Drives (SSD) are faster still, but as of this writing only available with relatively smaller storage capacity. It seems likely that SSD drives, without the moving parts of a traditional hard drive, will fi nd their place, especially for remote and portable recording systems. D rive buff er or cache size is also important, and larger drives require larger caches to function smoothly when handling large amounts of audio. Although drives as large as 1 terabyte will probably provide adequate performance with 16 MB caches, 32 MB is recommended for 1 terabyte and above. Th e chipsets that handle the hard-drive operations also aff ect data-transfer speed and reli- ability, and some have been developed specifi cally for streaming large quantities of audio and video. Th e Oxford 911 chipset for FW400 (Firewire 400) connec- tions, the Oxford 912 for FW800, and the Oxford 934 for SATA drives are fre- quently used by drives that are maximized for handling a lot of data. Multiple hard drives can be set up in RAID enclosures (redundant array of independent disks) that require only one connection to the computer. M any of the specs described above are changing so frequently as to re- quire new research each time you buy a new drive. Th ere are various packagers, THE ART OF DIGITAL AUDIO RECORDING of drives that are optimized for media (audio and video), and it is a good idea to use them as resources for the latest in specs and stick to their products, if possible; not all of them charge substantially more just for being “specialized” media drive. DVD-R (recordable DVD discs) can be used for relatively small fi le storage, and even CD-Rs hold enough data for some backup or transfer functions. Th e plethora of legacy storage media, from Exabyte tape drives, to zip drives, and back to the variously sized fl oppy drives, reminds us that storage formats come and go.

Online Glossary Link

A comprehensive glossary of audio terms requires a lot of entries. Th ere is not the space to undertake such a project here, but fortunately there is a very good audio and recording glossary available on the Internet. Th e online audio store Sweetwater is an excellent source of information about gear, as well as one of many good options for online purchasing, and it has an outstanding glossary provided as a public service. Th e glossary can be accessed here: 254,


Note Italic page numbers indicate photographs, screen shots, or diagrams. Numbers of the recording room, 13–14; minimizing, 1/3 octave EQs, 49 22–23 3-to-1 rule, 26 terminology for, 207 16-bit digital audio format, 246, 247, 248 amplifi ers, headphone amplifi er and mixer 24-bit digital audio format, 246, 247 systems, 91 60-cycle hum, fi ltering out, 54–55 See also guitar amps; mic preamps analog distortion vs. digital distortion, 65 A analog EQs vs. digital EQs, 49–50, 54 absorption (of sound) analog gain, conversion to digital, 65 and frequency response, 11–13 analog gear vs. digital gear, 3, 6, 28–29, 74, isolation and, 11 174–175 absorption materials, 13 analog mixers/consoles, 30, 41, 42–43, 175 acoustic bass, recording/miking, 105–106, 105 automation systems on, 200 acoustic guitar analog routing, 29 EQ-ing, 185–186 analog simulation, 71, 74 recording/miking, 107, 107 analog summing vs. digital summing, 42, 175 acoustics analyzing programs, 145 255 of the control room, 14 arranging, mixing as related to, 173–174 of the recording room, 10–14 art, of mixing, 176 active DIs, 94–95 See also creative endeavor; creative listening adjusting asking about headphone mixes, 89, 93 expansion, 66 attack times on compressors, 60–61 gain, while comping, 136, 137; while mixing, audio. See audio regions/sub-regions; audio 182–183 tracks; and other audio .topics; digital sonic characteristics and consistency, 215 audio; incoming audio (in talkback See also pitch shift ing/correction/adjustment; systems) timing adjustment audio channels, 31 “ad-libbed vocal vamps”, 138 inserts on (see inserts) advanced editing, 138–153 primary input/output, 31–32 adjusting timing and pitch, 142–148 sharing eff ects among, 159 global cuts and additions, 138–142 two processors on one channel, 155–156, 156 miracle edits, 152–153 See also audio tracks; channels; channel strip silence function, 66, 151–152 strips time compression or expansion, 148–151 audio channel strips, 30–31, 31 aesthetics, creative listening, 6–7, 9 I/O settings, 31–32; mono/stereo See also creative endeavor confi gurations, 32–33 Aguilera, Christina, “Genie in a Bottle”, 143 audio dynamics, 55 AIFF fi les, 247–248 audio fi les AKG C414 mic, 20, 104 delivery of (see digital audio delivery) AKG C452 mic, 20, 99, 102 digital formats, 246–248 AKG D112 mic, 20, 96, 97, 105 header information, 247 ambience managing, 84–85 eff ects, 188, 198, 198 (see also delays; preparing for mixing, 177–181 reverb(s)) removing unused, 85, Index audio fi les (continued) Auto-Tune program, 73, 73, 146–147 storage of, 252–254 auto-tuning devices, 73, 73 types, 247–248 aux channels/inputs/tracks, 31, 39, 159–160 audio levels (input levels) master auxiliary track (SUB), 181, 181 aspects (absolute and relative), 212 recording with compression on, 57, 58 balancing (see balancing levels/elements) routing multiple audio tracks to, 179, 180 detection by compressors, 60, 62 aux sends (sends), 34–36, 35 setting (see setting levels) for headphone mixes, 35, 91 audio processing. See signal processing output routing, 35–36 audio production as pre-/post-fader, 34–35, 35, 91 essential information taught in this book, xi setting levels in, 202 guiding principle, 6–7, 9, 74, 198 uses, 35 primary practices, 3–6 See also editing; experience; gear; mastering; B mixing; recording; recordists; signal path; background elements in mixes, 182, 198 signal processing; sound background noise, reducing, 66 audio regions/sub-regions, 119–120, 120 background vocals analyzing programs, 145 copying, 137 editing features, 119, 120–127 panning strategy, 184 fades, 127–130 stereo reverbs for, 193, 194 on a grid, 123 balancing audio tracks with eff ects, 160 locking in place, 124–125 balancing levels/elements moving, 122–123, 125, 137 in mastering, 212, 214–215 normalizing, 213 in mixing, 182–183, 197–198 nudging, 126–127, 143–144 band pass fi lters, 49 placing, 122–125 “bands” on EQs, 44 quantizing, 144–145, 146 bandwidth (Q) (EQ parameter), 44–45, 46 recombining elements, 138, 172 shelving a starting frequency, 46–47, 47 256 removing unused elements, 85 bass buildup, 11 returning to original place, 124 bass drum. See kick drum selecting, 125 bass instruments sending to reverb, 159–160, 161 as a baseline element, 182–183 sliding/shuffl ing, 123, 123, 125, 143 compression of, 187 sounds of clicking or popping in, 127 panning strategy, 183 tempos for, 149 recording/miking, 104–106, 105 transitions in and out of, 127 bassoon, recording/miking, 115 trimming, 125–126, 201, 204 bass trapping, 12, 13 audio store online, 254 “Believe” (Cher), 143 audio tracks (tracks) (dry signals), 132–133 bi-directional/bipolar mics. See fi gure-8 mics balancing with eff ects, 160 bit depths for digital audio formats, 246, 247, breakpoints on, 202–203, 202 248 hiding, 178 boost and dip (EQ parameter), 44 invisible track, 219 brass instruments, recording/miking, 112–114, organizing, 177–181 113, 115–116 panning between eff ects and, 161–164 breaking rules, 7, 158 routing multiple tracks to an aux track, 179, breakpoints (on audio tracks), 202–203, 202 180 brickwall limiters, 60, 65–66, 213 virtual (see virtual tracks) for mastering, 220 audition mode. See input-only mode brickwall limiting, 65 authority for mixing, 174 in mastering, 188, 212–214, 213 automation (of mixing), 5–6, 199–205, 206 in mixing, 188, 249 auto-panning, 185, 203, 203 and the snare drum, 219 auto-switching (of monitoring), 165–169 brighter-sounding mics, 108, 112 in DAWs, 167–169 Broadcast Wave Format, 247 vs. input-only mode, 165, 166 buff er size of drives, 253 and punching-in, 166–167, 168 burning masters, 217, Index “the buss”, 180 chorusing eff ect, 68, 189 buss compression, 187–188 vibrato vs., 74–75 buss routing. See internal routing clacking-type percussion, recording/miking, buying gear 104, 104 from dealers, 17, 244 clarinet, recording/miking, 115 EQ presets, 186 clearing audio, 122 how much to get, 174 clicking sounds in audio, 127 mic preamps, 33 click tracks, 92, 93 mixing speakers, 172–173 clipboard, 120 research options, 172–173, 243–244, 254 close mics, 106–107 stores, 244–245, 254 close miking, 69 trying before buying, 243, 244 closer (term), 207 upgrading the weakest link, 242–243 coincident pair technique, 25, 25 buzzes, 88 collaboration, on mixing, 174, 205, 206 fi ltering out, 48, 54, 73 See also communication in recording bypassing the talkback system, 17–18, 229 sessions comments in recording sessions, 223 C commercial digital audio formats, 248 cache size of drives, 253 commercial studios, engineering as a guest in, cardioid mics (directional mics), 21, 21 239–241 placement of, 25, 26, 27–28 communication in recording sessions, 223–225 proximity eff ect, 109 about EQ, 50, 206–207 See also fi gure-8 mics listening levels and, 237 cascading sounds, 190 about mixing, 206–207 CD-burning programs, 220 See also terminology CD digital audio formats, 246, 247, 248 comping (composite editing), 135–138 CD manufacturing, digital audio delivery for, adjusting pitch, timing, and gain while, 249–250 136–137 CD mastering, digital audio delivery for, 249 detailed work in, 137–138 257 CD projects “It could have happened” approach, 136 fi le management, 84–85 vocal comping, 135–136, 137, 138 storage of, 253 complex patching, 82–83 center-image stability problems, 27 composing, mixing as related to, 173–174 centering mics, 109–110 compression (of dynamic range), 55, 55 cents (pitch increments), 146 frequency-conscious compression, 62–63 channel fader. See fader group compression, 156–157, 157 channels (on mixers/DAWs), 39 in mastering (see brickwall limiting) group controls, 37 in mixing, 187–188 groups, 37–38 of multiple tracks, 179 names for/notes on (track names/notes), of the overall mix, 157, 187–188 38, 39 recording with, 57, 58, 94 types, 31, 38–40 uses, 56–57 See also audio channels; aux channels; See also compressors; limiting; time master fader compression or expansion channel strips (on mixers/DAWs), 30–31, 31, compressors, 55–66 78–79 audio level detection by, 60, 62 fader, 34, 35, 37, 79 controls, 57–59, 60–61 hardware/processing options for, 42, insert model as for, 155–156, 156 43–44 metering functions, 59 inserts, 155, 156 operation (mechanics), 57 track naming and scribble strip, 38, 38 plug-ins, EQ and compressor on one See also audio channel strips channel, 155–156, 156 Charles, Ray, “What does it sound like?”, 6–7, plug-ins (soft ware), 57, 62 9, 74, 198 types, 61–64 Cher, “Believe”, 143 See also limiters chipsets, 253 comp tracks, 135, Index computer interfaces, 252–253 channel strips (see channel strips) computer technology. See DAWs (digital audio digital audio formats, 246–247 workstations) editing terminology, 119 condenser mics, 18, 19, 20, 112 as interfaced with consoles, 78–79, 79 power source, 33 I/O routing (see I/O routing) uses, 100, 106, 110–111, 112 microphone connections to, 29 See also large-diaphragm condensers; pencil mixer confi guration, 84 condensers; small-diaphragm condensers; mixer-style interfaces, 29–30 and specifi c mics recording with compression in, 57, 58 congas, recording/miking, 103, 103 screen management, 131–132 consoles. See hardware mixers setup, 78–80, 79, 84–86 consumer digital audio formats, 248 See also control surfaces; plug-ins; Pro Tools; control panel of an EQ, 157, 158 soft ware mixers control-room mixes, in headphone mixes, 89, dbx systems, 73 90, 92 dealers, buying gear from, 17, 244 control-room monitoring, 89–90 Decca Tree confi gurations, 27 setup, 79–80 uses, 116, 117 control room(s) (listening environment) de-essers, 55, 63, 63 acoustics, 14 delay plug-ins, 67, 190 as the mixing environment, 172–173 delays, 67–69, 188–190 recording in, 14, 89–90 (see also control- combining reverbs with, 193–195 room monitoring) musical time delays, 67, 190, 192 control surfaces (digital control surfaces), 29, panning, 189, 190 30, 40, 41 plug-ins, 67, 190 advantages, 79 send and return model as for, 159, 164 capabilities, 43–44 unnatural delays, 188 conversation. See communication in recording See also reverb(s) sessions Del Chiaro, Joe, 114 258 conversion, analog gain to digital gain, 65 deleting audio, 122 conversion boxes, 95 delivery of digital audio. See digital audio copying audio, 120 delivery cowbell, recording/miking, 104, 104 depth of the three-dimensional mix, 198, 198 CPU power usage/availability diff users, 13, 13 for plug-ins, 192 diff usion (of sound) send and return model and, 164 in the control room, 14 creative endeavor in the recording room, 13 economics and, 228 digital audio, 246 mixing as, 170, 173–174, 175, 176, 182, 195 delivery of (see digital audio delivery) recording as, 7, 158 formats, 246–248 creative listening, 6–7, 9 storage, 252–254 criticism in recording sessions, 223, 224 See also audio .topics cross-fades, 128–129, 128, 130, 130, 219 digital audio delivery, 216–217, 249–252 cutting audio, 120 of mixes for mastering, 207–208, 249 digital audio formats, 246–248 D digital audio storage, 252–254 dampening a snare drum, 98–99 digital audio workstations. See DAWs DAO protocol, 216, 217, 250 digital control surfaces. See control surfaces DAWs (digital audio workstations) digital distortion vs. analog distortion, 65 adjusting timing and pitch in, 142–148; digital EQs vs. analog EQs, 49–50, 54 while comping, 136–137 digital gain, analog gain conversion to, 65 audio tracks, 132–133 digital gear vs. analog gear, 3, 6, 28–29, 74, auto-switching in, 167–169 174–175 capabilities, 3–6, 43–44, 71–72, 119, 138, digital mixers, 41, 43 142–143, 145, 146, 148 See also soft ware mixers channels (see channels) digital reverbs, 70, Index digital signal processing (DSP), 5, 41 dynamic mics, 18–19, 20, 96, 97, 112 aux channels in, 39 uses, 19–21, 22, 97, 100, 102, 103, 104, eff ects, 71–75 106–107, 108, 112, 114, 116 routing for (see insert model; send and See also ribbon mics; and specifi c mics return model) dynamic range See also compression; delays; EQ compressing (limiting), 55 (see also (equalization); expansion; pitch shift ing/ compression) correction/adjustment; reverb(s); strip increasing, 66 silence; time compression or expansion; dynamics (audio dynamics), 55 timing adjustment; and also dynamics dynamics processing, 5 processing fade-outs when using on the master fader, digital signal processors, 5, 41, 74 181 two on one channel, 155–156, 156 insert model as for, 155–156, 159 See also dynamics processors; EQs in mixing, 186–188 digital summing vs. analog summing, 42, 175 See also compression; dynamics processors; digital zero, 65, 212 expansion; limiting direct boxes/inputs (DIs), 94–95 dynamics processors, 55, 155 uses, 104–105, 107, 229 expanders/noise gates, 66 directional mics. See cardioid mics; fi gure-8 insert model as for, 155–156, 159 mics See also compressors; limiters direct recording, 104–105, 107 DIs. See direct boxes/inputs E distortion (overload) ear (hearing) digital vs. analog, 65 and EQ-ing, 50–53, 186, 237 preventing (see compression) and mastering, 211, 220 Dolby systems, 73 and mixing, 171, 175, 177 double reeds, recording/miking, 115 sensitivity, 44, 50–51 doubling eff ect, 68, 189 ear fatigue, 177, 234–235 Dowd, Tom, 6–7 early refl ections (reverb), 69, 70–71 259 drive speed, 253 eBay, buying gear from, 245 drum loops, compressing or expanding, echo chambers, 69–70 149–150 echo eff ects, 69–70 drum percussion, recording/miking, 103, 103 repeat delays, 67, 190, 202 drums (drum sets) economic fl ow (in recording sessions), 222, 228 as baseline elements, 182–183 edit functions, 119, 120–122 compression eff ects, 56–57 editing, 119–153 EQ-ing, 53–54, 100 advanced (see advanced editing) group compression on, 156–157, 157 composite (see comping) groups, 178, 179 expansion adjustment, 66 leakage reduction, 66 features (see edit functions; edit modes; edit miking, 19, 95–102, 103; positioning, 100 tools) (see also under specifi c drums) fi lls, 121–122 panning strategies, 183 global editing, 138–142 recording, 95 (see also .miking, above) while mastering, 217–218 sound, 98, 175 miracle edits, 152–153 See also specifi c drums mixing as not, 174 drum tracks, grooving performances to, 145–146 nondestructive editing, 4–5, 138 dryer (term), 207 terminology in DAWs, 119 dry signals. See audio tracks edit modes, 122–125 DSP. See digital signal processing edit points in audio, 128–129, 129, 130, 130 duplicate virtual tracks, 133–134 edits, seamless, 127 duplicating audio, 121, 121 edit tools, 125–130 DVD digital audio format, 251 eff ects (FX) (wet signals) dynamic EQs (multiband compressors), 55, adding impact to mixes, 187–188 63–64, 64 aux sends for, 35, Index eff ects (FX) (wet signals) (continued) EQs (equalizers), 44 balancing audio tracks with, 160 control panel, 157, 158 digital signal processing eff ects, 71–75 de-essers, 55, 63, 63 echoes, 67, 69–70 digital vs. analog, 49–50 panning between dry signals and, 161–164 dynamic EQs (multiband compressors), 55, send and return model as for, 159, 164 63–64, 64 sharing among audio channels, 159 graphic EQs, 49, 49 tremolo, 74, 75, 203, 203 insert model as for, 155–156, 155, 156, 159 vibrato, 74–75 for mastering, 220 See also compression; delays; EQ (equali- parametric settings, 45–46, 45, 46; control zation); expansion; reverb(s); stereo eff ects panel, 157, 158 elastic pitch capability, 148 plug-ins, 45, 155, 155; compressor and EQ electric bass, recording/miking, 104–105, 105 on one channel, 155–156, 156 See also bass instruments presets, 186 electric guitar equal-gain cross-fades, 130, 130 fi nger vibrato/tremolo eff ects, 75 equal-power cross-fades, 130, 130 panning reverb returns, 193 equipment. See gear recording/miking, 74, 106–107, 106 eSATA protocol, 252–253 See also lead guitar expanders, 66 electronic instruments insert model as for, 155 recording/miking, 118 (see also direct expansion (of dynamic range), 66 recording) See also time compression or expansion short delay eff ects, 189 experience Electrovoice RE-20 mic, 20, 97 in EQ-ing and compression, 94 Elvis eff ect, 190 familiarity with gear, 17, 22 engineering as a guest in commercial studios, in mastering, 211, 220–221 239–241 in miking, 22–23 engineers. See recordists in mixing, 171, 172, 173, 175–176 260 ensemble recording in pitch adjustment, 146 hardware mixers for, 42 in setting levels, 87 including a piano, 110–111 in troubleshooting, 88 microphones for, 21, 22, 24–25; placement, external routing (interface routing), 32, 32, 26–27 35–36, 160 See also orchestral recording environment. See ambience; control room; F recording room fader (channel/output fader), 34, 35, 37, 79, 127 EQ (equalization), 42, 44–55, 185 See also master fader communication about, 50, 206–207 fades (fade-outs), 127–130, 200, 218, 219 fi ltering with, 47–49, 48, 54–55 on the master fader, 181, 181 insert model as for, 155–156, 155, 156, 159 fades menu, 129 parameters, 44–49 familiarity with gear, 17, 22 parametric EQ, 46 far mics, 106–107 and phase, 50 farther (term), 207 shelving EQ, 46–47, 47 feedback loops (talkback system), 230 side eff ects (unintended eff ects), 52–53 fi delity, microphones and, 19–21, 22–23, 28–29 See also EQ-ing; EQs (equalizers) fi gure-8 mics (bi-directional mics), 21 EQ-ing placement of, 27–28 the ear (hearing) and, 50–53, 186, 237 fi le management, 84–85 fi ltering, 47–49, 48, 54–55 fi les. See audio fi les learning process, 53 fi lls while mixing, 53, 179, 185–186 editing, 121–122 multiple tracks, 179 moving, 126–127 the overall mix, 157 fi lm, digital audio delivery for, 251 while recording, 53–54, 94, 100 fi ltering sounds best vs. fi ts best confl ict, 185–186 with EQ, 47–49, 48, 54–55 EQ presets, 186 with noise reduction processors, 73, Index fi lters, high/low-pass fi lters, 47–49, 48 grid(s) fi nger vibrato, 75 adjusting timing with, 144–146 fi rst run-throughs, recording, 87 moving fi lls with, 126–127 fi ts best vs. sounds best confl ict, 185–186 musical time settings, 122, 123, 203 fl anging eff ect, 68, 189 regions on, 123 fl ash drives, 253 repositioning audio on, 137 Fletcher-Munson curve, 50–51, 177 groove templates, 145–146 fl ute, recording/miking, 115 group compression, 156–157, 157 “fooling the automation”, 204 group controls, 37 foreground elements in mixes, 182, 198 groups (of channels), 37–38, 178, 179 free-hand automation, 203, 203 designation/categorization of, 178 frequency (EQ parameter), 44 inserts on, 156–159 high/low-pass fi lters, 47–49, 48 subgroups, 179, 180 range, 197, 198 guest engineering in commercial studios, shelving a starting bandwidth frequency, 239–241 46–47, 47 guests at recording sessions, 224 terminology for, 206–207 guiding principle of audio production, 6–7, 9, frequency-conscious compression, 62–63 74, 198 frequency response (of the recording room), guitar amps, microphones for, 19 11–13 guitar amp simulation/simulators, 71, 74 frequency sensitivity of the ear, 44, 50–51 guitar fi lls FX. See eff ects editing, 121–122 moving, 126–127 G guitars, recording/miking, 74, 106–107, 106, 107 gain See also acoustic guitar; electric guitar adjusting while comping, 136, 137 adjusting while mixing, 182–183 H boost and dip (EQ parameter), 44 half-normaled patch points, 82, 82 gain control (on compressors), 59, 61–62 hand drums, recording/miking, 103, 103 261 gain conversion (analog to digital), 65 hard drives, 252–254 gated reverbs, 188, 192 hard-knee settings, 61 gear (equipment) hard left /hard right rule, 24, 28 analog gear, 3, 28–29, 174–175 hardware mixers (consoles), 30, 40–41 buying (see buying gear) advantages, 78 familiarity with, 17, 22 analog mixers, 30, 41, 42–43, 175, 200 mastering tools, 219–220 channels, 39 mixing tools, 174–176 digital mixers, 41, 43 outboard gear, 33, 209 as interfaced with wall panels and DAWs, talkback systems, 229–230 78–79, 79 See also amplifi ers; compressors; DAWs; onboard mic preamps, 34, 41 EQs; expanders; headphones; inserts; setup, 78–80, 79 limiters; microphones; mixers; monitors vs. soft ware mixers, 42–43 (studio monitors); plug-ins; presets; signal talkback system, 229–230 processors; speakers headphone mixes, 89–93 “Genie in a Bottle” (Aguilera), 143 amplifi cation for, 91 global editing, 138–142 asking about, 89, 93 additions, 139–142, 142, 143 aux sends for, 35, 91 cuts, 138–139, 140 control-room mixes in, 89, 90, 92 global time compression or expansion, 151 helping musicians create, 92–93 glossary link online, 254 musician control of, 89, 92 grabber tool, 125 separate mixes, 90–91 grand piano, recording/miking, 109–111, setup, 79–80 110 submixes, 91–92 graphic EQs, 49, 49 headphones graphic mode of pitch correction, 147 amplifi er and mixer options, 91 grid mode, 122–123, 123 checking, 86–87, Index headphones (continued) instrument channels, 40 listening with while mixing, 197 instruments mixes for (see headphone mixes) levels (see audio levels) vs. monitors, 15–16, 89 recording/miking, 93–117 as needed, 15–16 sound elements, 93 as not needed, 89 See also specifi c instruments power requirements, 78 interconnecting session elements (patching), setup, 77–78 80–83 volume controls on, 87 interface routing. See external routing hearing. See ear interfaces height of the three-dimensional mix, 197, 198 computer interfaces, 252–253 hiding audio tracks, 178 DAW—console interface, 78–79, 79 high-frequency buzzes, 48 DAW mixer-style interfaces, 29–30 high-level listening (loud listening), 237–238 internal routing (buss routing), 32, 32, 35–36, and ear fatigue, 234–235 160 hearing high and low frequencies with, 235, Internet 236 buying gear on, 244, 245 as masking pitch and rhythm, 235–236 researching gear on, 172–173, 243–244, while mastering, 216 254 while mixing, 177, 197 Internet applications, digital audio delivery for musicians, 236 for, 250–251 as not always appropriate, 237 “in time” delays, 190 high-pass fi lters, 47–49, 48 intonation, adjusting, 146, 148 high-pitched percussion, recording/miking, invisible track, 219 103, 104 I/O connections for plug-ins, 158–159 hi-hat I/O routing, 7–8 panning strategy, 183 analog routing, 29 recording/miking, 99, 99 aux send outputs, 34, 35 262 hi-hat fi lter, 49 external/interface vs. internal/buss, 32, 32, honesty in recording sessions, 224 35–36, 160 horn sections, recording/miking, 115–116 fundamentals (see auto-switching; insert humor in recording sessions, 224–225 model; send and return model) hums, 88 multiple tracks to an aux track, 179, 180 fi ltering out, 54–55 in parallel/serial, 194, 195 side-chaining, 63–64, 64 I I/O settings (on channel strips), 31–32 impact, adding to mixes, 187–188 mono/stereo confi gurations, 32–33 impulse-response reverbs, 70, 70 isolation incoming audio (in talkback systems), of microphones, 97, 110 230–232 of the recording room, 10–11 input levels. See audio levels ISRC codes, 217 input lists (mic plots), 76–77, 77 “It could have happened” approach, 136 input-only mode (of monitoring), 165 vs. auto-switching, 165, 166 J not using, 166, 167 Jagger, Mick, 95 input/output, connections for plug-ins, 158–159 K See also I/O routing; I/O settings keyboard-like instruments, recording/miking, insert model, 154–155, 155–156, 155, 159 111–112, 112 See also inserts kick drum (bass drum) inserts (in soft ware mixers), 34, 154–159, 155, as a baseline element, 182–183 156 panning strategy, 183 on groups and stereo buss channels, recording/miking, 96–98, 96, 97 156–159 kick drum fi lter, 49 on individual channels, 155–156, 155, 156 knee characteristics (of compressors), 61 See also plug-ins Kubota, Reiko, 117, Index L longer reverbs, 191, 192 large-diaphragm condensers, 19, 20, 21, 22, 97, look-ahead operation (of soft ware com- 112 pressors), 62 uses, 22, 97, 101, 102, 103, 106, 107, 108, looping audio, 121 109–110, 110–111, 112, 116 cross-fades for, 130 See also specifi c mics recording multiple takes while, 121 large monitors, 17–18 loops, compressing or expanding, 149–151 latching talkback systems, 230, 231–232 loud listening. See high-level listening lead guitar loudness curve (smile curve), 51–52, 52 combining a reverb with a delay, 194 low-frequency oscillators (LFOs), 68, 68 two delayed and pitch-shift ed signals eff ect, low-frequency rumble, 48 189 low-level listening (quiet listening) See also electric guitar for detecting pitch and rhythm, 235 lead sheets in walls, 10–11 while mastering, 216 lead vocals while mixing, 177, 196–197 compression of, 187 low-level noise, stripping (see strip silence creating, 135–136, 137, 138 function) panning strategy, 183 low-pass fi lters, 47–49, 48 two delayed and pitch-shift ed signals eff ect, 189 M leakage reduction, 66 magnifying glass tool, 131 Leslie speakers, 75 make-up gain, 59 levels. See audio levels; listening levels managing fi les, 84–85 LFE channels, 251 marimba, recording/miking, 109, 111, 112 LFOs (low-frequency oscillators), 68, 68 mashups, 72, 172 limiters master auxiliary track (SUB), 181, 181 brickwall limiters, 60, 65–66 master fader (channel), 31, 39–40, 157, vs. compressors, 60 180–181 See also compressors fade-outs on, 181, 181 263 limiting (of dynamic range) stereo buss processing on, 180 brickwall limiting, 188 two-buss level on, 182 when recording, 94 mastering, 5, 210–221 See also compression; limiters balancing levels/elements, 212, 214–215 linear fades, 129–130 basics, 212–219 line testing, 86–87 brickwall limiting in, 188, 212–214, 213 listening delivery of mixes for, 207–208, 249 creative listening, 6–7, 9 editing while, 217–218 while mastering, 214–215, 216 experience in, 211, 220–221 while mixing, 176–177, 196–197 goals/tasks, 211, 215 See also listening environments; listening listening environment, 220–221 levels listening while, 214–215, 216 listening environments, 197 mixing while, 218–219 for mastering, 220–221 with multiband compressors, 64 for recording and mixing (see control room) requirements, 210–211 listening levels separation mastering, 219 and conversation, 237 tools, 219–220 for detecting pitch and rhythm, 235 mastering tools, 219–220 for hearing high and low frequencies, 236 masters for mastering, 216 burning, 217 for mixing, 177, 196–197 delivery of, 216–217, 249–252 See also high-level listening printouts, 217, 218 live recording, 15–16, 89 master submix (SUB), 181, 181 fi le management, 84 maximum volume/level, 64–65 locked audio region, 124 medium delays, 67–68, 188–189, 189–190 locking audio in place, 124–125 melodic loops, compressing or expanding, long delays, 67, 188–189, 190 150–151, Index Melodyne program, 147, 147 I/O routing (see I/O routing) mic clips, 26 types, 41 mic plots (input lists), 76–77, 77 See also DAWs; hardware mixers; soft ware mic preamps, 29, 33–34, 33, 41, 42, 43 mixers DI function, 95 mixers (recordists), experience with the mixing setup, 78–79, 79, 81–82, 82 environment, 173 microphones, 18–29 mixes brighter-sounding mics, 108, 112 adding impact to, 187–188 cardioid (see cardioid mics) compression of the overall mix, 157, centering mics, 109–110 187–188 condensers (see condenser mics; large- delivery of for mastering, 207–208, 249 diaphragm condensers; pencil condensers; EQ-ing the overall mix, 157 small-diaphragm condensers) multiple mixes, 208–209 connections to DAWs, 29 overloading, 204 directional (see cardioid mics) recall of, 199, 204–205 dynamic (see dynamic mics) revising, 198–199 and fi delity, 19–21, 22–23, 28–29 rough mixes, 79 fi gure-8 mics, 21, 27–28 saving under diff erent names, 199 isolation of, 97, 110 stems, 219, 252 omni-directional mics, 21, 21, 26 as three-dimensional, 197–198, 198 overhead mics, 95–96, 100–101, 101 See also headphone mixes; submixes pickup patterns (see pickup patterns) mixing, 5–6, 170–209 placement of, 14, 22–23; with musicians, 77; art of, 176 omni mics, 21; stereo techniques, 24–28 authority for, 174 (see also under specifi c instruments) automation of, 5–6, 199–205 room mics, 102, 102 balancing levels/elements, 182–183, selection of, 19–21, 22; musician’s 197–198 preferences, 116 (see also under specifi c basic operations, 181–188 264 instruments) in the box vs. out of the box, 174–175, 199, setup, 76–77 205 signals (see signals) brickwall limiting in, 188, 249 speakers as, 97–98 collaboration on, 174, 205, 206 stereo (see stereo mics) communication about, 206–207 types, 18–21, 20, 25, 105–106 compression in, 187–188 warmer-sounding mics, 108, 112 as a creative endeavor, 170, 173–174, 175, See also specifi c microphones 176, 182, 195 MIDI channels, 40 defi nition, 171–172, 176 MIDI quantizing function, 144 delays as used in, 188–190, 193–195 mid/side technique (M/S technique), 27–28 dynamics processing in, 186–188 miking instruments, 95–117 the ear and, 171, 175 close miking, 69 environment, 172–173 with musicians, 77, 116 EQ-ing while, 53, 179, 185–186 stereo techniques, 24–28; three-mic experience in, 171, 172, 173, 175–176 technique, 109–110, 110, 111, 111, 112 gain adjustment, 182–183 (see also under specifi c instruments) goal, 182 miracle edits, 152–153 listening while, 176–177, 196–197 mixers (mixing boards), 29–44 while mastering, 218–219 analog mixers, 30, 41, 42–43, 175, 200 as not editing, 174 channels (see channels) panning strategies, 183–185 channel strips (see channel strips) playback system, 172–173 in DAWs (see control surfaces; soft ware preparing fi les for, 177–181 mixers) procedures, 195–199, 200, 209 defi nition, 29 recall, 5, 204–205, 209 digital mixers, 41, 43 (see also soft ware remote mixing, 205 mixers) requirements, 171–176 headphone amplifi er and mixer systems, 91 reverbs as used in, 191–195, Index setting levels, 182, 196–197; volume sensitivity to, 225, 225–226 automation, 201–202, 202 setting input levels with, 87 tools, 174–176 muting audio, 121–122, 121 mixing environment, 172–173 mixing tools, 174–176 N modulation, 68–69 naming mixes, 199 by pitch-shift ing devices, 72 near-coincident pair technique, 25–26, 26 momentary talkback systems, 230–231 near-fi eld monitors, 15–16, 172 monitoring, 165 placement of, 16–17, 16 input-only mode, 165, 166 selection of, 17 See also auto-switching; control-room Neumann KM-84 mic, 20, 101, 105, 107 monitoring Neumann KM-184 mic, 110, 117 monitors (studio monitors), 14–18 Neumann M-49 mic, 108 vs. consumer speakers, 15 Neumann U-47 mic, 97 vs. headphones, 15–16, 89 Neumann U-87 mic, 20, 102, 110, 111, 112, large monitors, 17–18 114 mixing speakers, 172–173 nitpicking by musicians, avoiding, 235 powered monitors, 16 noise timbre characteristics, 15 fi ltering out, 54–55 See also near-fi eld monitors; speakers reducing, 48, 54, 66, 73 mono in/mono out reverbs, 193 stripping low-level noise (see strip silence mono input, panning in stereo, 36 function) mono in/stereo out confi guration, 32, 33 noise gates, 66 reverbs, 161, 162, 163–164, 163, 193 noise reduction processors, 73 mono output(s), 33 nondestructive editing, 4–5, 138 stereo output as playing in mono, 36 nondestructive recording, 4 monophonic summing, 27 nonlinear automation, 203, 203 Moorhead, Michael, 108 normaled patch points, 81–82, 82 moving audio, 122–123, 125, 137 normalizing audio, 213 265 moving fi lls, 126–127 nudging audio, 126–127, 143–144 mp3 digital audio format, 248, 250 M/S technique (mid/side technique), 27–28 O multiband compressors (dynamic EQs), 55, oboe, recording/miking, 115 63–64, 64 off -line automation, 200, 201–203, 202, 203, multiple mixes, 208–209 206 multiple takes off -line processing, 147–148 managing, 85–86 omni-directional mics, 21, 21 recording, while looping, 121 placement of, 26 See also virtual tracks online audio store, 254 mults (in patch bays), 82 online automation, 199–200, 201 musical fl ow (in recording sessions), 222, online glossary link, 254 227–228 optical-type compressors, 61, 62 musical time orchestral recording, microphones for, 21, delays based on, 67, 190, 192 24–25; placement, 14, 26–27 grid settings, 122, 123, 203 See also ensemble recording musical understanding needed by recordists, organizing audio tracks, 177–181 227–228 ORTF confi guration, 25–26, 26 musicians outboard gear, 33, 209 avoiding nitpicking by, 235 output ceiling control (on brickwall limiters), headphone mix control, 89, 92 65–66 helping create headphone mixes, 89, 91–92, output fader. See fader 92–93 overdubs high-level listening for, 236 setup, 80 miking instruments with, 77, 116 in vocal recording, 90 overplaying by, 225 overhead mics, 95–96, 100–101, 101 putting at ease, 224 panning strategy, 183, Index overload. See distortion pitch overloading mixes, 204 increments, 146 overplaying by musicians, 225 listening level for detecting, 235–236 overtone series, 51 pitch shift ing/correction/adjustment, 71, Oxford chipsets, 253 142–143, 146–148 while comping, 136 P devices, 72–73 packing blankets, 110 graphic mode, 147 panning (stereo output), 36, 184, 197–198, 198 by sight vs. by ear, 148 auto-panning, 185, 203, 203 vibrato, 74–75 basic positions, 183 pitch-shift ing devices, 72–73 delays, 189, 190 placing audio, 122–125 between dry signals and eff ects, 161–164, planning for recording sessions, 76 162 plate reverbs, 70 mixing strategies, 183–185 playback system(s) reverb returns, 192–193 for mastering, 219–220 parallel eff ects/routing, 194 for mixing, 172–173 parallel wall/fl oor/ceiling surfaces, 12 volume control, 234–238 parametric EQ, 46 playback volume, 234–238 passive DIs, 94–95 varying, 196–197 pasting audio, 120 See also listening levels patch bays, 80–81, 83, 83 playing outside (the rules), 7, 158 mults, 82 plug-in parameters, automating, 204 patching (interconnecting session elements), plug-ins, 41, 154–155 80–83 Auto-Tune, 73, 73, 146–147 patch panels, 84 auto-tuning devices, 73, 73 patch points, 81–82, 82 compressors, EQ and compressor on one patterns. See pickup patterns channel, 155–156, 156 266 peak-level detection (by compressors), 62 compressors (soft ware), 57, 62 peak normalization, 213 CPU power availability for, 192 pencil condensers, 21, 25 delay plug-ins, 67, 190 uses, 22, 110 EQs, 45, 155, 155; compressor and EQ on pencil tool, 126 one channel, 155–156, 156 percussion instruments, recording/miking, I/O connections, 158–159 102–104, 103, 104 on the master fader, 180 See also drums (drum sets); keyboard-like mixing gear, 175–176 instruments; pianos pitch-shift ing devices, 72–73 percussion loops, compressing or expanding, reverb plug-ins, 70, 161, 191 149–150 Stereo Tools VST plug-in, 28 percussive sounds trying before buying, 244 compression of, 60–61 See also inserts with cross-fades, 130, 130 polarity (of signals), 24, 24, 106 See also drums popping sounds in audio, 127 performing, mixing as related to, 173–174 portable hard drives, 253 phantom power, 18 positive feedback in recording sessions, 224 supply source, 33 post-fader aux sends, 34–35, 35, 194–195 phase relationships/coherency (of signals), power, phantom. See phantom power 23–24, 23, 106 powered monitors, 16 EQ and, 50 power requirements for headphones, 78 phasing eff ect, 68 practical aspects of audio production, xi pianos, recording/miking, 109–112, 110, 111 essential question, 6–7, 9 pickup patterns (of microphones), 21, 21 predelay times for reverbs, 71 See also cardioid mics; fi gure-8 mics; omni- pre-fader aux sends, 34–35, 35, 91 directional mics reverb only eff ects, 195, 196 pickups for acoustic bass, 105–106 presets ping-ponging delays, 188 for EQ-ing, 186, Index for reverbs, 71 recording digital audio formats, 246–248 for timeline views, 131 recording room(s) Presley, Elvis, slapback eff ect, 190 acoustics, 10–14 primary input/output (of audio channels), ambience, 13–14; minimizing, 22–23 31–32 isolation of, 10–11 printouts from masters, 217, 218 predelay times, 71 processors. See signal processors shooting the room, 18 project genres, and mixing, 171 recording sessions, 76–118 See also CD projects communication in (see communication in protocols for audio fi les, 247 recording sessions) Pro Tools (DAW), 240 fl ow in (see session fl ow) features, 120, 122, 147–148, 178, 200, 201; guests at, 224 channel strips, 31, 179 planning for, 76 terminology, 119 setup, 76–88 proximity eff ect (directional mics), 109 See also headphone mixes; miking pumping (compression artifact), 61 instruments; recording; recordists punching-in, 4, 165 recordists auto-switching and, 166–167, 168 best practices, 222–238 (see also playback run-up time for, 227–228 volume; session fl ow; talkback system(s)) punk rock, short delay eff ects, 189 communication by (see communication in recording sessions) Q compression by, 56, 62 quantizing audio, 144–145, 146 creativity (see creative endeavor) questions in recording sessions, 223 EQ-ing by, 52–53, 186 asking about headphone mixes, 89, 93 experience (see experience) quiet listening. See low-level listening familiarity with gear, 17, 22 fundamentals, 7–9, 154–169 (see also auto- R switching; insert model; send and return rack toms. See tom-toms model; signal path) 267 ratio setting (on compressors), 57, 58–59, 59 job description, 93 real estate. See screen mixing preferences, 172, 195–196, 200, 209 reamp boxes, 95 musical understanding needed by, 227–228 reamping, 95 pitch correction by, 148 recall (in mixing), 5, 204–205, 209 sensitivity to musicians, 103, 224, 225, recombining recorded elements, 138, 172 225–227 recommendations and reviews of gear, 243, 244 record producers, 224 recorded elements. See audio regions/ redrawing waveforms, 126, 126 sub-regions reed instruments, recording/miking, 114–115, recording 114 in commercial studios as a guest engineer, refl ection (of sound), 14 239–241 and frequency response, 11–13 with compression, 57, 58, 94 isolation and, 11 in the control room, 14, 89–90 monitor placement and, 16–17 as a creative endeavor, 7, 158 regions of audio. See audio regions/ digital audio formats, 246–248 sub-regions direct recording, 104–105, 107 release times on compressors, 61 EQ-ing while, 53–54, 94, 100 remixing, 172 fi rst run-throughs, 87 remote mixing, 205 instruments, 93–117 (see also under specifi c repeat delays/echoes, 67, 190, 202 instruments) repeating audio, 121, 121 line testing, 86–87 repositioning audio. See moving audio multiple takes, while looping, 121 research options for buying gear, 172–173, nondestructive recording, 4 243–244, 254 setting levels for, 87 returning audio to original place, 124 troubleshooting, 88 reverberation. See ambience; reverb(s) See also audio production reverb only eff ects, 195, 196, Index reverb plug-ins, 70, 161, 191 send and return model, 159–160, 161 reverb(s), 69–71, 191–195 and CPU power usage, 164 combining with delays, 193–195 with stereo eff ects, 161–164, 162, 193, control parameters/qualities, 70–71, 191 194–195 devices, 69–70; plug-ins, 70, 161, 191 uses, 159, 164, 192, 194–195 gated reverbs, 188, 192 sending audio to reverb, 159–160, 161 individual vs. common settings, 192 sends. See aux sends listening to, 177 Sennheiser 421 mic, 20, 97, 100, 100, 103, 104 mono in/stereo out reverbs, 161, 162, sensitivity 163–164, 163, 193 of the ear, 44, 50–51 panning returns, 192–193 of recordists to musicians, 103, 224, 225, plug-ins, 70, 161, 191 225–227 predelay times, 71 separated audio region, 125 presets, 71, 192 separation mastering, 219 selecting (choosing), 191 sequencing songs, 215–216 send and return model as for, 159, 164, 192, serial eff ects/routing, 194, 195 194–195 session fl ow, 222–228 sending audio to, 159–160, 161 and mic placement, 23 slapback and, 190 setting levels stereo (see stereo reverbs) aspects, 212 time/length, 69, 70–71, 191–192 when mastering, 212 types/timbre, 70, 191 when mixing, 182, 196–197; volume reverb tails, 69, 70–71 automation, 201–202, 202 reverb time/length, 69, 70–71, 191–192 for recording, 87 revising mixes, 198–199 sharing eff ects among audio channels, 159 rhythm, listening level for detecting, 235–236 shelving EQ, 46–47, 47 rhythm-altering soft ware, 72 shooting the room, 18 ribbon mics, 21, 112–113 short delays, 68–69, 188–189 268 uses, 106–107, 107, 112–113 short fades, 127–128, 127 ride cymbal, recording/miking, 96, 101, 102 short reverbs, 191 riding gain, 61–62 shuffl e mode, 123 right-angle wall/fl oor/ceiling intersections, shuffl ing/sliding audio, 123, 123, 125, 143 11–12 Shure SM57 mic, 20, 98, 98, 106 RMS-level detection (by compressors), 62 Shure SM81 mic, 111, 112, 113 room mics, 102, 102 side-chain routing, 63–64, 64 rough mixes, creating, 79 signal path, 7–9, 8 routing. See I/O routing basics, 154–169 (see also auto-switching; rules, breaking, 7, 158 insert model; send and return model) run-up time for punching-in, 227–228 for de-essing, 63 DIs and, 94–95 S talkback system feedback loops, 230 sample rates for digital audio formats, 246, 247, testing, 86–87 248 troubleshooting, 88 SATA drives, 252–253 signal processing, 5, 41 saving mixes under diff erent names, 199 See also digital signal processing; dynamics saxophones processing combining a reverb with a delay, 194 signal processors, 5, 41 recording/miking, 114–115, 114 See also digital signal processors Schoeps CM-5 mic, 110 signals screen polarity, 24, 24, 106 managing, 130–131 splitting, 82, 82 setup recall options, 131–132 See also phase relationships/coherency scribble strip, 38, 38, 81 simple patching, 80 seamless edits, creating, 127 slapback delays, 67–68, 189–190 selected audio region, 125 sliding/shuffl ing audio, 123, 123, 125, 143 selecting audio (selector tool), 125 slip mode, 122, Index small-diaphragm condensers, 19, 20, 98, 98, mixing speakers, 172–173 112 monitors vs. consumer speakers, 15 uses, 98, 99, 100–101, 101, 105, 106, 107, timbre characteristics, 15 109, 110–111, 112, 113–114, 116 See also monitors (studio monitors) See also specifi c mics speaker trick, 97–98 smile curve (loudness curve), 51–52, 52 splitters, 95 SMPTE timecode, 123, 251 splitting signals, 82, 82 snare drum spot mode, 123–125 adjusting a hit, 144, 145 Spot mode dialog box, 124 brickwall limiting and, 219 spreads panning strategy, 183 between songs, 216 recording/miking, 98–99, 98 stereo spreads, 189 stripping of noise, 152, 152 spring reverbs, 70 soft -knee settings, 61 SSDs (Solid State Drives), 253 soft synths, 31, 40 SSL mixers/consoles, 30, 42–43, 175 soft ware compressors, 62 automation systems on, 200 soft ware mixers, 29–30, 40–41 stage plots, 77 vs. hardware mixers, 42–43 standing waves, 12 inserts (see inserts) starting elements in audio events, 128 I/O routing within, 32 stems (of mixes), 219, 252 See also control surfaces; DAWs stereo buss channels, inserts on, 156–159 soft ware plug-ins. See plug-ins stereo buss processing, 180, 219 soft ware synthesizers. See soft synths stereo eff ects Solid State Drives (SSDs), 253 send and return model with, 161–164, 162, songs 193, 194–195 adding identifying information to, short delays, 189 216–217 stereo input(s), 32–33 adding verses, 139–142, 142, 143 stereo mics, 25 with adjustments in timing and pitch, 143 uses, 96, 101, 102, 102, 109 269 removing verses, 138–139, 140 stereo miking techniques, 24–28 sequencing, 215–216 three-mic technique, 109–110, 110, 111, spreads between, 216 111, 112 transitions between, 219 stereo output See also vocals panning (see panning) sonic characteristics and consistency, adjusting, as playing in mono, 36 215 See also mono in/stereo out confi guration; soprano saxes, recording/miking, 115 stereo eff ects; stereo reverbs sound, 3 stereo reverbs, true, 161–163, 163, 164, 164, cascading sounds, 190 193, 194 of digital gear vs. analog gear, 6 stereo spreads, 189 of drums, 98, 175 stereo stems, 252 importance (“What does it sound like?”), Stereo Tools VST plug-in, 28 6–7, 9, 74, 198 stopping elements in audio events, 128 instrumental sound elements, 93 storage of audio fi les, 252–254 percussive sound compression, 60–61 stores for buying gear, 244–245, 254 quality (see timbre) strings (stringed instruments), recording/ refl ection of, 14 miking, 22, 116, 117, 117 of speakers/monitors, 15 string sections, recording/miking, 116–117 transients, 60–61 strip silence function, 66, 151–152 See also acoustics studio monitors. See monitors sound leakage, preventing, 10–11 studios, commercial, engineering as a guest in, sounds best vs. fi ts best confl ict, 185–186 239–241 spaced pair confi guration, 26–27 See also control room(s); recording room(s) speakers stutters, 190 Leslie speakers, 75 subgroups, 179, 180 as microphones, 97–98 sub-master (SUB), 181, 181, Index submixes, 179, 180 speaker/monitor characteristics, 15 for headphones, 91–92 time compression or expansion of, 150–151 master submix (SUB), 181, 181 time compression/expansion menu, 150 stems, 219, 252 time compression or expansion, 71–72, 72, sub-regions of audio, 120, 120 148–151, 150 See also audio regions/sub-regions timeline (in DAWs) subwoofers, 17 managing, 131 summing managing multiple takes on, 85–86 analog vs. digital summing, 42, 175 regions on (see audio regions/sub-regions) monophonic summing, 27 time stamps/timecodes, 123, 124, 247, 251 surround sound, 33, 248, 251 timing adjustment, 142–143, 143–146 digital audio formats, 251 while comping, 136–137 Sweetwater online audio store, 254 by sight vs. by ear, 148 See also time compression or expansion T timing codes. See time stamps/timecodes Takahashi, Brandon, 113 timing run-up for punching-in, 227–228 takes, multiple. See multiple takes Tom Dowd & the Language of Music, 6–7 See also virtual tracks tom-toms talkback button, operating, 232–234 panning strategy, 183 talkback level, checking, 87, 232 recording/miking, 95, 100, 100 talkback system(s), 228–234 stripping of noise, 152, 152 built-in systems, 229–230 tone controls, 44 button operation, 232–234 See also EQs bypassing, 17–18, 229 touch mode (online automation), 201 feedback loops, 230, 232 T-Pain, 143 level checking, 87, 232 track names and scribble strip, 38, 38, 81 operational types, 230–232 track names/notes, 38, 39 tambourine, recording/miking, 103, 104 tracks. See audio tracks; virtual tracks 270 TAO protocol, 216, 250 transients (in sounds), 60–61 tape-based editing, 5 analyzing programs based on, 145 tape hiss, fi ltering out, 73 transitions between songs, 219 tap tempo function, 190 transposing, 146 technical aspects of audio production, xi tremolo eff ect, 74, 75, 203, 203 essential process, 9 triangle, recording, 54–55 technical fl ow (in recording sessions), 222, triangle symbol, 202 225–227 trimmed audio region, 126 templates, 83, 84, 85 trimmer tool, 125–126 groove templates, 145–146 trimming audio, 125–126, 201, 204 tempos, assigning, 149 trim mode (online automation), 201 terminology troubleshooting, 8–9, 88 for ambience, 207 trust in recording sessions, 223 for editing (in DAWs), 119 “tunnel of love”, 97 for frequency, 206–207 two-buss level on the master fader, 182 glossary link online, 254 two delayed and pitch-shift ed signals eff ect, 189 See also communication in recording sessions testing the signal path, 86–87 U thickening eff ects, 68, 189 unattended mixing, 205 three-dimensional mix, 197–198, 198 unnatural delays, 188 three-mic technique, 109–110, 110, 111, 111, unrestricted editing mode, 122 112 upgrading the weakest link, 242–243 three-to-one rule (3-to-1 rule), 26 upright piano, recording/miking, 111, 111 threshold control (on compressors), 57–58, 59 USB-1 interface, 253 on brickwall limiters, 65–66, 65 timbre (sound quality), 51 V EQ and, 52 VCA-type compressors, 61–62 of reverb types, 70, 191 verbal fl ow (in recording sessions), 222, 223–225, Index verses (of songs) reduction of (see compression) adding, 139–142, 142, 143 See also audio levels; dynamic range; removing, 138–139, 140 playback volume vibraphone, recording/miking, 109, 111 volume automation, 201–202, 202 vibrato eff ect, 74–75 volume controls on headphones, 87 video, digital audio delivery for, 251 volume scale for the human ear, 44 video games, digital audio delivery for, 252 virtual tracks, 132–138, 134, 135 W comping using, 135–138 wall intersections/parallel surfaces, 11–12 as comp tracks, 135 wall panels, as interfaced with consoles, 78–79 duplicate tracks, 133–134 wall treatments, 10–11, 12–13, 12 managing multiple takes on, 85–86 warmer-sounding mics, 108, 112 vs. multiple individual tracks, 135 Wave fi les, 247–248 new tracks, 134–135 waveforms pop-up menus, 133 redrawing, 126, 126 See also multiple takes reliability, 148 vocabulary. See terminology wet signals. See eff ects vocal comping, 135–136, 137, 138 wetter (term), 207 vocal recording, 108 “What does it sound like?” (Charles), 6–7, 9, in the control room, 14, 89–90 74, 198 microphones for (see under vocals) width of the three-dimensional mix, 197–198, overdubs, 90 198 vocals Wiesendanger, Beth, 112 “ad-libbed vocal vamps”, 138 wind instruments, recording/miking, 112–116, automating, 203–204 113, 114 combining reverbs with a delay, 193–194 window dubs, 251 compression of, 56 woodwinds. See reed instruments creating (see vocal comping) workarounds, 88 miking, 19–21, 22, 108–109, 108; placement, workspace, managing, 130–131, 131–132 271 14, 108, 109 See also background vocals; lead vocals; X songs; vocal comping; vocal recording X/Y confi guration, 25, 25, 101 volume (of sound) absolute and maximum values, 212 Z EQ and, 52 zero, digital, 65, 212 panning and, 185 zero crossing point, 127, 128, 128]

Similar documents

CAMBRIDGE TESTS 8-1 ENGLISH GRAMMAR - REZOLVARE Choose the correct word (s) to complete each sentence. CAMBRIDGE 1. You should try to avoid foods which …A… a lot of fat. A. contain B. contains C. are containing D. is containing 2. Can you remember what you were doing when …D… the news? A. hearing B.
The Beatles Complete Songbook A collection of Beatles’ chords and tabs by Sergio Palumbo
The Beatles Complete Songbook A collection of Beatles’ chords and tabs by Sergio Palumbo email is hidden Naples, Italy “Release” 1 - July-August 2001 Index INTRODUCTION 9 PLEASE PLEASE ME 10 I SAW HER STANDING THERE 10 MISERY 12 ANNA (GO WITH HIM) 12 CHAINS 13 BOY
The Immortals: Brian Eschbach
The Immortals: Brian Eschbach Trevor Strnad Ryan Knight Max Lavelle Alan Cassidy Live Sound: Matt “Woody” Woods Tour Management/Merch: Michael MacDonald Backline Madman (EU): Patrick Uhlmann Management: EJ Johantgen for Prosthetic Publicity: Kelli Malella for Metal Blade Records Merchandise Director
PINK FLOYD – THE WALL Christmas comes but once a year For every girl and boy The laughter and the joy They find in each new toy l'll tell you of a little boy Who lives across the way This little fella's Christmas ls just another day lt was just before dawn One miserable morning ln black '44 When the
Ritual (pron. rich-oo-uhl): An established or prescribed procedure, rite, ceremony, proceeding or
Ritual (pron. rich-oo-uhl): An established or prescribed procedure, rite, ceremony, proceeding or service...Any practice or pattern of behavior regularly performed in a set manner...A prescribed code of behavior. The Black Dahlia Murder does not exactly seem like a band that's easily ritualized. The
The BIOS Optimization Guide rev. 5.7
The BIOS Optimization Guide rev. 5.7 All trademarks used are properties of their respective owners. Copyright © 1998-2001 Adrian Wong. All rights reserved. BIOS Features Setup (Part 1) Virus Warning / Anti-Virus Protection Options : Enabled, Disabled, ChipAway When Virus Warning is enabled, the BIOS
FOURTH EDITION THE ARCHITECTURE OF COMPUTER HARDWARE, SYSTEM SOFTWARE, AND NETWORKING AN INFORMATION TECHNOLOGY APPROACH Irv Englander Bentley University John Wiley & Sons, Inc. Vice President & Executive Publisher Don Fowley Executive Editor Beth Lang Golub Marketing Manager Christopher Ruel Market
Please use CPS: orgchem/0010004 in any reference to this article
The chemistry of isatins: a review from 1975 to 1999 Joaquim Fernando M. da Silva,* Simon J. Garden and Angelo da C. Pinto Departamento de Química Orgânica, Instituto de Química, Universidade Federal do Rio de Janeiro, RJ, 21945-970, Brazil Isatina (1H-indol-2,3-diona) é um composto de grande versat
Hitler: The Greatest Story Never Told
Hitler: The Greatest Story Never Told 00:00:01,021 -> 00:00:03,061 part 12 Adolf Hitler - the gretest story never told! 00:00:08,703 -> 00:00:10,855 On 25 July 1943, italien leader Benito Mussolini is overthrown and imprisoned. 00
The Adventures of Sherlock Holmes
The Adventures of Sherlock Holmes by Arthur Conan Doyle A Scandal in Bohemia The Red-headed League A Case of Identity The Boscombe Valley Mystery The Five Orange Pips The Man with the Twisted Lip The Adventure of the Blue Carbuncle The Adventure of the Speckled Band The Adventure of the Engineer's T
CEMETARY 1213 LYRICS The Beast Divine (2000) 1. Firewire 2. Union Of The Rats 3. Silicon Karma 4. Antichrist 3000 5. The Carrier 6. Linking Shadows 7. Sunset Grace 8. Dead Boy Wonder 9. Empire of the Divine 10. Anthem Apocalypse 1. Firewire out of the flames burned black from shame we rule again no
The Drughouse vol.12
The Drughouse vol.12 After more then 100.000 downloads on volume 11, 4 months of collecting the best tracks and a crazy release-party, we proudly present you The Drughouse vol.12! Mister Artistic Raw has put together the best exclusive house-tracks in a banging new mix to give you that crazy Drughou
THE INFLUENCE OF VEDIC PHILOSOPHY ON NIKOLA TESLA'S UNDERSTANDING OF FREE ENERGY "The first thing to realize about the ether is its absolute continuity. A deep sea fish has probably no means of apprehending the existence of water; it is too uniformly immersed in it: and that is our condition in rega
The Outer Limits - Episode Guide
The Outer Limits - Episode Guide Page 1 of 118 The Outer Limits - Episode Guide Season: 1234567All The Outer Season 1 Limits 1. The Sandkings (1) Show gs: Helen Shaver (Cathy Kress) Home Beau Bridges (Simon Kress) Episode Lloyd Bridges (Colonel Kress) List Dylan Bridges (Josh Kress) Goofs Kim Coates
The audiomachine LABYRINTH – THE PLATINUM SERIES IV release contains: one
The audiomachine LABYRINTH – THE PLATINUM SERIES IV release contains: one DVD and two audio CDs. For your convenience, the DVD includes 16 bit AIFF files in 48 kHz format along with corresponding mp3 files and an accompanying PDF file with helpful writer/publisher information. We have also included
The Joker Is Wild (1957)
The Joker Is Wild (1957) Frank Sinatra plays Joe E. Lewis, a famous comedian of the 1930s-50s. When the movie opens, Lewis is a young, talented singer who performs in speakeasies. When he bolts one job for another, the mob boss who owns the first speakeasy has his thugs try to kill Lewis. Lewis surv
The Memory Book Harry lorayne and jerry lucas
The Memory Book Harry lorayne and jerry lucas BALLANTINE BOOKS • NEW YORK Copyright© 1974 by Harry Lorayne and Jerry Lucas All rights reserved. Library of Congress Catalog Card Number: 73-90705 SBN 345-24527-X-195 This edition published by arrangement with Stein and Day Publishers First Printing: Ju
The Mechanical Design Process
ullman-38162 ull75741_fm December 18, 2008 16:19 The Mechanical Design Process ullman-38162 ull75741_fm December 18, 2008 16:19 McGraw-Hill Series in Mechanical Engineering Alciatore/Histand Heywood Introduction to Mechatronics and Measurement System Internal Combustion Engine Fundamentals Anderson
THE STRANGE LIFE OF NIKOLA TESLA Editors Note, August 28, 1995 This text has been entered by John R.H. Penner from a small booklet found in a used bookstore for $2.50. The only form of date identification is the name of the original purchaser, Arthua Daine (?), dated April 29, 1978. The book appears
The Truth About Creation & Evolution More Than 75 Ways Science Supports Creation and Disproves Evolution. By Robert Knopf
The Truth About Creation & Evolution More Than 75 Ways Science Supports Creation and Disproves Evolution. By Robert Knopf ((1st inside page – right facing)) Copyright 2005 Robert Knopf All rights reserved. No part of this book may be reproduced or transmitted in any form or by any means, electronic
THE WATCHMEN This is a work of fiction. All the characters and events
THE WATCHMEN This is a work of fiction. All the characters and events portrayed in this book are fictional, and any resemblance to real people or incidents is purely coincidental- Star Watchman copyright © 1964 by Ben Bova. The Dueling Machine copyright © 1969 by Ben Bova. All rights reserved, inclu
The Essential List Of Foods Dogs Cannot Eat
The Essential List Of Foods Dogs Cannot Eat Mild Danger • Avocado. The problems with avocado are twofold. Firstly, avocado contains a compound called persin, which is toxic to many animals, including rabbits, birds, horses, cats, and dogs. Avocado rarely causes serious illness in animals like dogs,
The Grand Deception A Second Look at the War on Terrorism
The Grand Deception A Second Look at the War on Terrorism © 2001 by G. Edward Griffin The concepts I would like to share with you today were set to paper three days after the terrorist attack against the World Trade Center and the Pentagon on September 11. I printed about a dozen copies and gave the
THE MARIJUANA GROWER'S GUIDE by Mel Frank and Ed Rosenthal Typed by Ben Dawson Revised 1992 NOTE:- Footnotes have been placed in double brackets (()). Numbers throughout refer to bibliography and are sometimes in brackets, sometimes they aren't. All dates are for northern hemisphere only. Comments o
A Mythic Vistas Campaign Setting for the d20 Modern Roleplaying Game Design: T.S. Luikart and Ian Sturrock Editing: Michelle Lyons Development: Robert J Schwalb
A Mythic Vistas Campaign Setting for the d20 Modern Roleplaying Game Design: T.S. Luikart and Ian Sturrock Editing: Michelle Lyons Development: Robert J Schwalb Art Direction and Graphic Design: Hal Mangold Cover Art: Christian Gossett, Snakebite and Paul Schrier Interior 2D Art: Christian Gossett,
part 1 CharaCters
THE ART OF part 1 CharaCters Characters Kate Walker riding a snow ostrich clinic outfit *draw by Amanda Goengrich connections to oscar necklace oscar’s heart start-up - radiation Goggles* - *draw by Amanda Goengrich Characters the Youkol people owl Ayawaska Kurk Characters Nic Cantin simon steiner *
JoWooD Productions Software AG, Pyhrnstraße 40, A-8940 Liezen, Austria
JoWooD Productions Software AG, Pyhrnstraße 40, A-8940 Liezen, Austria Homepage JoWooD:, Homepage “The Guild 2”: © 2006 by JoWooD Productions Software AG, Pyhrnstraße 40, A-8940 Liezen, Austria. Developed by 4Head Studios. Developed with the support of the MEDIA Prog
This is Freeware Numerous Input Formats
Intro video: This is Freeware offers absolutely free programs which have been developed as alternatives to paid ones. Make sure that our free video converter is freeware in the truest sense of the word. Screenshots: Numerous Input Formats Import videos from video cams (avi, mpg, tod, mo
Keyboard Commands Movement and In-game Actions Bard Song Left click on button to play Bard Song. New Mage Spells
Keyboard Commands Movement and In-game Actions Bard Song Left click on button to play Bard Song. New Mage Spells Pause/ Press [ Space Bar ] or to pause play and issue commands to any of your characters. Walk to location Left click on ground to have selected character(s) walk there. Nahal's Reckless