Mixing Music
Mixing Music
Download and Listen samples for chapters 6 and 9 are available on www.routledge.
com/9781138218734 and on http://hepworthhodgson.com.
Russ Hepworth-Sawyer has been involved in professional audio for over two
decades and, in the last, mastering through MOTTOsound.com. Throughout his
career, Russ has maintained a part-time higher education role teaching and
researching all things audio (www.hepworthhodgson.com). Russ is a current
member of the Audio Engineering Society and is a former board member of the
Music Producer’s Guild and helped form their Mastering Group. Russ currently
lectures part time for York St John University, has taught extensively in higher
education at institutions including Leeds College of Music, London College of
Music and Rose Bruford College and has contributed sessions at Barnsley Col-
lege. He has written for MusicTech magazine, Pro Sound News Europe and Sound
on Sound, as well as many titles for Focal Press/Routledge.
Mixing Music
Edited by Russ Hepworth-Sawyer and Jay Hodgson
Mixing Music
and by Routledge
2 Park Square, Milton Park, Abingdon, Oxon OX14 4RN
The right of the editors to be identified as the author of the editorial material, and of the authors for
their individual chapters, has been asserted in accordance with sections 77 and 78 of the Copyright,
Designs and Patents Act 1988.
All rights reserved. No part of this book may be reprinted or reproduced or utilised in any form
or by any electronic, mechanical, or other means, now known or hereafter invented, including
photocopying and recording, or in any information storage or retrieval system, without permission in
writing from the publishers.
Trademark notice: Product or corporate names may be trademarks or registered trademarks, and are
used only for identification and explanation without intent to infringe.
George Martin
David Bowie
Prince
Zenon Schoepe
Contents
List of Figures ix
List of Tables xi
Series Introduction xii
About the Editors xiv
Notes on Contributors xv
Acknowledgmentsxx
Introduction 1
RUSS HEPWORTH-SAWYER AND JAY HODGSON
vii
viii Contents
Index279
Figures
ix
x Figures
xi
Series Introduction
xii
Series Introduction xiii
and requirements, and each of these reacts back on, and (re)shapes, the
musical object being produced in turn. Ultimately, it is uncovering and
understanding the broader musical ramifications of these priorities and
biases that comprises this series’ primary analytic concern.
Perspectives on Music Production also looks to broaden methodolog-
ical approaches that currently prevail in music production studies. The
place of traditional academic and scholarly work on record production
remains clear in the field. However, the place of research and reflection
by professional recordists themselves remains less obvious. Though music
production studies tend to include professional perspectives far more con-
scientiously than other areas of musical study, their contributions nonethe-
less are often bracketed in quiet ways. Producers, engineers and recording
musicians are often invited to participate in scholarly discussions about
their work only through the medium of interviews, and those interviews
typically follow more ‘trade’ oriented than straightforwardly academic
lines of inquiry. We thus invite contributions from professional recordists
which elucidate their own creative practice, and in whichever ways they
deem most relevant to scholarly considerations of their work. Similarly, we
hope the series will encourage greater collaboration between professional
recordists and the researchers who study their work. As such, we invite
contributions that model novel and inclusive methodological approaches
to the study of record production, encompassing professional, creative,
interpretive and analytic interests.
It is our sincere hope that Perspectives on Music Production provides a
timely and useful intervention within the emerging field of music produc-
tion studies. We hope each volume in the series will spur growth in music
production studies at large, a more detailed and comprehensive scholarly
picture of each particular procedure in a record production, as well as a
general space for researchers to pause and reflect back on their and their
peers’ work in this exciting new area.
Jay Hodgson and Russ Hepworth-Sawyer
About the Editors
xiv
Notes on Contributors
Chapter Authors
Phil Harding joined the music industry at the Marquee Studios in 1973,
engineering for the likes of The Clash, Killing Joke, Toyah and Matt
Bianco by the late 1970s. In the 1980s, Phil mixed for Stock, Aitken &
Waterman tracks such as ‘You Spin Me Round’ by Dead or Alive followed
by records for Mel & Kim, Bananarama, Rick Astley, Depeche Mode,
Erasure, Pet Shop Boys and Kylie Minogue. In the 1990s, Phil set up his
own facility at The Strongroom with Ian Curnow. Further hits followed
with productions for East 17 (including ‘Stay Another Day’), Deuce, Boy-
zone, 911 and Let Loose. Recent projects include the book PWL from the
Factory Floor (2010, Cherry Red Books) and mixing Sir Cliff Richard’s
2011 album Soulicious. Harding has recently worked for Holly Johnson,
Tina Charles, Samantha Fox, Belinda Carlisle and Curiosity with his new
xv
xvi Notes on Contributors
Alex Krotz has been surrounded by music since a very young age and
has always had a talent and undeterred passion for creating music. He
has worked with a wide range of artists, including some of Canada’s
largest acts (Shawn Mendes, Three Days Grace, Drake), as well as many
up-and-coming bands. He currently is an engineer at Noble Street Studios.
Dean Nelson is a producer and engineer mixer and professor of Music Pro-
duction & Pro Tools at Ontario Institute of Audio Recording Technology
Notes on Contributors xvii
(OIART). Nelson began assisting for producers Neal Avron, Mark Trom-
bino and Ethan Johns. He later moved on to Chalice Studios in Holly-
wood, then Ocean Way Studios. Nelson assisted Jack Joseph Puig for
over a decade, but began a five-year mentor relationship with Jack
that eventually grew from assistant to engineer. During his tenure with
Jack, Dean assisted/engineered projects for the Rolling Stones, Fergie’s
Grammy-nominated hit, ‘Big Girls Don’t Cry’, Mary J. Blige and U2. In
addition to his engineering duties, Dean also assisted Jack and Waves with
PuigChild and PuigTec. Dean was later offered a job as Beck’s engineer. In
this role, he recorded and/or mixed an array of projects including Stephen
Malkamus and the Jicks’s latest Mirror Traffic, Sonic Youth founder Thurston
Moore’s album Demolished Thoughts, ‘The Record Club’ projects, Charlotte
Gainsbourg’s IRM and Stage Whisper, Jamie Lidell’s Compass and the Beck/
Bat For Lashes Twilight collaboration Let’s Get Lost. Upon relocation to Lon-
don, Ontario, he produced and mixed several songs off Buck 65’s Neverlove.
Alastair Sims is passionate about music and has devoted his life to helping
bands and artists recognize their dreams. Working with some of Canada’s
largest bands (Rush, Walk Off The Earth, Three Days Grace) in many of the
greatest studios (Noble Street, Revolution), he has been able to work along-
side and learn from some of the best engineers and producers. Continuing
his work with up-and-coming Canadian artists along with well-established
bands, Alastair is cementing himself as a strong presence in Canadian music.
Adam Marshall (Graze/New Kanada) was born and raised in Toronto and
is now based in Berlin. Adam has been active as a DJ/Producer and record
label boss for ten years.
TJ Train (Room 303) signed his first release under the Room 303 brand
to the UK’s Love Not Money label in 2011. He has since gone on to sign
tracks to other esteemed labels including On the Prowl Records (New
York), Suruba Records (Spain) and Subtrak Records (Berlin). TJ has also
performed alongside Miguel Campbell, Damian Lazarus, Infinity Ink, No
Regular Play, Tone of Arc, Nitin and James Teej, to name a few.
Acknowledgments
The authors would like to thank, first and foremost, all those at Routledge
Press who shared our vision for this project and who were instrumental in
seeing this series to print.
We must, however, extend our thanks to those who have contributed
their research, their conclusions and of course their patience as we’ve
encouraged, edited, disheartened (perhaps) and edited again. The work in
this book, and the subsequent series already in full flow, is wide ranging.
We’d like to thank everyone who has contributed. We’d also like to thank
those who put in a call and were not selected for this book—perhaps we’ll
see you later in the series. Thank you for your support and shared vision
for the project.
Finally, we’d like to thank our families for their patience, too, as we
write and pull together another book.
xx
Introduction
Russ Hepworth-Sawyer and Jay Hodgson
As you will have learned from our series introduction, this is the first in
a series of planned multi-authored books exploring academic interests in
music production from a range of perspectives. This approach enables
us, as a discipline, we believe, to study any topic from differing angles.
Whether that be an angle of perspective, such as the method of approach-
ing a mix, or the way in which you look at or listen to a mix, or to consider
the prism through which you analyze a mix, Perspectives on Music Pro-
duction is a broad church.
Much of our writing to date has been conceptualized and organized
within the adopted term of the production process. For example, From
Demo to Delivery: The Process of Production (edited by Hepworth-
Sawyer, 2010) involved multiple authors discussing phases of that exact
staged production process. From the outset of Perspectives on Music Pro-
duction, our proposal to Routledge, and the invitations to the contributors
here, we were clear what the titles of each of the books would be. True
to form, they are based upon broad segments of the production process.
The output, however, would dispel any such theory. We have welcomed
this. A researcher will take the book out of the library most applicable to
their current investigation. What transpires is a number of discussions that
mixing music cannot be pigeonholed into one specific stage of the music
production process. The mix is within the concept, the seed of the initial
composition, or idea. Reconceptualization can occur again later through
the production of a track, which could of course be equally applied before
going into the studio and after the recording in the mixing stage. Expect,
therefore, to read about recording, production, pre-production and even
mastering within this book, all in reference to the creation of a product.
The focus of Perspectives on Music Production: Mixing Music therefore is
the culmination, the moment if you like, of a discrete number of tracks or
sources coming together to make an exciting, cohesive whole.
Many might query why the first book in a series loosely based around
the production process should launch with Mixing Music. As we researched
and interviewed for our forthcoming book Audio Mastering: The Artists
(due out at a similar time to the release of this book), we noted how little
research had actually been carried out on the mix. We were equally alarmed
1
2 Russ Hepworth-Sawyer and Jay Hodgson
that outside of the popular music technology press, there was little ethno-
graphic, or experiential, discussion of the mix. Additionally, we felt that
the prism by which the mix is viewed should also be considered further.
Contributions have been submitted from both experienced academics
and early career researchers. Contributions have also been accepted from
professionals discussing their work. In addition, interviews have been
transcribed and are presented to you here for your own conclusions. You
will find some pieces based upon discussions and experiences, too. We
welcome these perspectives.
The relatively recent emergence of Music Production as an academic
topic has dawned at a time that coincides with open access journals, the
Internet and considerable change within the audio industry itself. These
have shaped both the academic activities we engage in, but also the indus-
try we write about today. Music Production, we believe, as a field of study
is therefore part history, part present and part future.
Zenon Schoepe, the late editor of Resolution magazine, wrote in one of
his last editorials in the March/April 2016 edition:
At that time, one presumes he was writing about the recent loss of Sir
George Martin, or perhaps David Bowie. He could not have known that
Prince was about to join the reference. However, those of us currently
studying the world of music production can empathize with Schoepe. We
dedicate ourselves to music production’s ‘part history’. Through the lens
of part history, we explore and dissect the historical developments and
undertakings in studios around the globe and judge, compare and root our
knowledge of the present.
As a form of study, our history is actually not that old. The industry
based upon the distribution of recorded audio has only clocked up around
100 years or so, yet there is still so much knowledge still to catalog, ana-
lyze and discuss.
We feel that much present work should be extracting or formalizing
historical or tacit knowledge held within current professionals. As an aca-
demic community we should continue to be noting this living history as
best we can, ensuring the skills, techniques and anecdotes are kept for
future researchers. We hope to capture and analyze as much as is possible
of what Schoepe referred to as ‘what they stand for’ in this and the forth-
coming books in the series.
Chapter Introductions
In the process of constructing this book, we received a number of abstracts
from potential contributors internationally. As we considered each contri-
bution upon its merit, we also considered the tone and flow of the whole
Introduction 3
act ‘The And’, this chapter explores the map that is the mix, concluding
that ‘it’s only rock n roll but we like it’.
Mark Marrington continues the ontological theme in chapter 13 by eval-
uating concepts bestowed upon the art of mixing. Marrington focuses first
on the role of the mix engineer in relation to the production process and
the people within the process. He later addresses the work the mix engi-
neer is undertaking on the music in front of them. Marrington discusses
the importance of the mixer ‘speaking on behalf of the artist’ through the
medium of the mix. Marrington explores other aspects of this rhetorical
art, discussing the constraints mixers are placed within because of expec-
tation or genre.
Jay Hodgson completes the ontology section with chapter 14, titled
‘Mix as Auditory Response’. Hodgson explores the concept of the mix
becoming the one article comprising the multiple multitrack elements.
Joshua Reiss, building upon his work on intelligent systems in music,
has contributed chapter 15 covering the recent emergence of algorithmic
approaches to mixing multitrack content. Using his experience of research-
ing in this area, this contribution explores the current thinking and practice
in intelligent systems that could, in time, transform the workflow of the
mix engineer completely. Reiss explores the factors in achieving a form
of automated mixing, but is keen to express that the creativity in audio
production would not be lost. The intelligent system cannot, at this stage,
replace the artistic decisions. Whilst the end product (if there will ever be
an end) would perhaps remove the monotonous tasks for the sound engi-
neer, Reiss is keen to point out that it will allow musicians to concentrate
on the music more.
As a professional engineer, Gary Bromham considers how academic
practice can inform mixing in chapter 16. Bromham notes that, within
music production, there is a natural thought that the flow of information
about mixing should be from practitioner to academia. Throughout the
chapter, Bromham discusses the potential flow of knowledge and consid-
eration that could be diverted from academia to the professional. Through
a series of interviews and personal experience, Bromham moves to look at
areas of the mix and how the ‘flow’ could benefit both parties. The chapter
does not only elaborate upon the mix and the mixing, but also the mixer as
a ‘sonic trend’ with a ‘sonic signature’.
In the final contributed chapter 17, Rob Toulson considers the final
aspect of mixing—the handing of the audio to the mastering engineer.
Through a series of interviews with mastering and mixing engineers,
Toulson explores the factors, problems and issues that are faced in this
often-silent transaction between the mix and the mastering of an artifact.
The chapter also explores the more modern phenomenon of mixing and
mastering as ‘a single process’. Toulson argues that the possibility to ‘cut
corners’ and put off decision-making until later in the production process
may have led to the processes of mixing and mastering becoming more
‘porous’ over time.
This series is also partly about the future. As described above, Perspec-
tives on Music Production: Mixing Music contains a number of papers and
Introduction 7
interviews that discuss the route map for mixing as we develop technolo-
gies and skills. For example, Joshua Reiss expands on his work on intel-
ligent systems in mixing. Developments such as landr.com have, despite
robust industry concerns of quality, opened up the starting acceptance of
automation. Never before has an automated system been accepted by a
portion of the music production community. Of course we have presets
within, say, Logic Pro, which claim to process a bass drum, but due to the
fact that every bass drum and every studio is different (let alone the drum-
mer), the settings will all need tweaking to get close to what’s required.
Automated services such as landr.com offer, at the time of writing, very
little in the way of honing. It is, however, only a number of years before
substantial portions of our workflow will be automated in mixing, open-
ing up a whole new debate. The final contribution from Russ Hepworth-
Sawyer in chapter 18 explores the future perspectives in mixing music.
1
Introduction
Historical Context
The level of possibility and complexity available to the mix engineer has
increased in steps since the role came into being. To appreciate the signifi-
cance of the increasing finesse that can be applied to the process of access-
ing and manipulating component parts of an audio piece, some historical
perspective is beneficial.
Originally, the mix was intimately connected with the performance,
both of these associated with the immediate, and transitory, fading with
the physical sound vibrations to reside only in the memory of the listener.
The mix was confined to the arrangement and guidance of the parts, a job
sometimes employed by a conductor.
This was changed in 1853, when an endurable artifact that repre-
sented audio was created, committed to a medium of soot on paper.
Parisian Édouard-Léon Scott de Martinville recorded an incomprehen-
sible squawk that is likely a human voice. It would be seven years later,
on April 9, 1860, before he was able to record something intelligible
on his apparatus which he now called a phonautograph, a recording of
someone, possibly himself (MacKinnon, 2012), singing ‘Au Claire de
la Lune’, amidst a sea of noise. Scott de Martinville never was able to
play back his recording or even appeared to contemplate the possibility,
and it would take 147 years before the development of a virtual stylus
by Carl Haber and his team at the Lawrence Berkeley National Labo-
ratory in Berkeley, California, would enable it to be heard for the first
time and open the door for an endurable mix, in this instance between
8
Exploring Potential of the Mix 9
the signal and noise, something that can be addressed with current tech-
nology (Rosen, 2008).
Edison’s sound recordings of a snippet of Handel’s Oratorio in 1877
enabled performances to be played back or, for the first time (Rosen,
2008), recalled for later examination. This necessarily would have brought
with it a new self-awareness, an aural mirror, which gave recording musi-
cians a new ability to refine their own performances. In addition, others
could now access and manipulate the recorded artifact with a degree of
autonomy from the performers.
Around 1920, a young German composer, Stefan Wolpe, created a
Dada provocation by simultaneously playing Beethoven’s fifth symphony
at different speeds on eight separate gramophone players. This was a con-
ceptual development as, for the first time, a piece that combined several
separate previously recorded elements was created. Wolpe possibly missed
the opportunity of exploring the mixing of entirely different recordings,
this being fulfilled three decades later in 1951 by John Cage’s piece for
twelve radios (Ross, 2013).
Cage was in step with the times as by then, the jazz guitarist, inventor
and legend Les Paul was laying the groundwork for combining separate
recordings of musically related parts. This was a significant paradigm shift
as, for the first time, it became practical to consciously, and sequentially,
combine the component parts of a piece.
Les Paul initially achieved this by the process of Sound on Sound
recording. As early as 1949, Paul was getting results by switching off the
erase head of a tape recorder to overlay parts on same piece of tape. An
alternate method involved recording a performance to an acetate (later
tape) and then playing along with it while recording to a second one. Paul
replaced this technique by using a modified Ampex 200 tape recorder with
a fourth head, enabling this process to be done using one, more portable
machine (Snyder, 2003). One mistake in the performance or extraneous
noise, however, and the process would need to be started from the begin-
ning; ‘How High The Moon’, recorded in 1951, had to be recorded three
times as the first two recordings were ruined by first a siren and then an
overflying airplane (Buskin, 2007).
Les Paul also explored moving beyond representations of live acoustic
performances, for instance, by recording his electric guitar at half speed so
that, on replay at the correct speed, it would be one octave higher. Record-
ing multiple vocal parts inspired both controversy and inspiration. Rich-
ard Buskin (2013) reports a conversation with Bruce Swedien, destined
later to become a highly acclaimed mix engineer, about How High the
Moon: “Up to that point the goal of music recording had been to capture
an unaltered acoustic event, . . . (it) left no room for imagination, but when
I heard ‘How High the Moon’, which did not have one natural sound in it,
I thought, ‘Damn, there’s hope!’ ”
Sel-Sync (or Selective Synchronous recording), also conceived by
Paul in 1953 (Petersen, 2005) and fabricated by Ross Snyder at Ampex
in 1955, opened up a realm of new possibilities, although initially not
10 Martyn Phillips
Where is all this increased sophistication leading to? Is there some perfect
solution that can be aspired to? Although this term is often used, is it actu-
ally possible to create a ‘perfect mix’?
Plato might have proposed that, like his eponymous solids, the ideals of
which only exist in the transcendent realm of Forms, there exists in there a
divine exemplar of the completed mixed work, a perfect piece of mixed music
that physical reality aspires to but can only approximate (Banash, 2006).
Music experienced in altered states of consciousness, dreams or through
near-death experiences (NDEs) suggests that there might indeed be some-
thing perfect that can be accessed. The latter music has been described as
“transcendental, unearthly harmonic beauty, angelic, sublimely beautiful,
exquisite harmonies, heavenly, a celestial choir of angels, a tone so sub-
limely perfect, joyous and beat-less melody, an orchestra of voices” (Wil-
liams, 2014). Pieces in such states can have the impression of appearing
complete, seemingly before any apparent human input has been done to
create it. The issue of where this music arises—in the mind of the listener
or pre-formed elsewhere and witnessed—opens up fundamental questions
on the nature of consciousness.
Incompleteness
Gödel’s proofs have also been extended beyond the realm of mathemat-
ics to the theory of mind, along with the support of Alan Turing’s ‘Turing
Machine’ thought experiment. One major corollary is that the human mind
can always find some aspect that cannot be contained within any ideal
(machine in this context) (Anon., n.d.).
Gödel’s first theorem is of particular interest, as it might be rephrased
as stating that a self-consistent axiomatic system cannot be completed.
A mixed piece of music might be considered such a self-consistent, or at
least self-referential, system as meaning is derived from within the context
of the combined elements. It might thus be postulated that there is, in fact,
no perfect mix, or indeed any piece of art, because as soon as it is pro-
duced, a new perception can be applied to it, which negates its perfection.
Artistic Exemplar
In the spirit realm, gardens sing and colors can be heard. It is a realm where
light and sound, color and geometrical patterns are all combined into a totality
of harmonic perfection. This is music that is on a level that is beyond hearing.
Holding the idea of what the target artifact might be will generate
responses on mental, emotional or physical levels, and these can enable
appropriate decisions to be made that may get closer to it. The perspectives
chosen against which to assess the mix should be pertinent to the song, the
artist, the genre and the age. The more perceptions that can be considered,
the more refined the potential. Once dealt with, other previously addressed
perspectives should then be checked to see if they are being adversely
affected. There are not-insignificant dangers of dilution of what might be
considered the essence of the piece and a resultant homogenized product
that is a jack-of-all-trades and master of none.
It is tempting to view this as chipping away at the edifice of the work,
as if working the facets of a crystal, incrementally closing in and gradually
revealing the final product. What materializes, however, can end up sur-
prisingly different from that envisaged at the outset. Why should this be?
Illusion
Imbalance
The human trait of exploring that which is off the path of harmony and
venturing into distortion and dissonance, may, at least in part, be attributed
to imbalance in the functions of the two brain hemispheres. Far from being
equal and opposite, there is a considerable variation in aptitude between
them, the dominant left hemisphere being less adaptable, which has been
demonstrated when it ceases to function properly and the right can take
over (Gynn and Wright, 2008: 4–8). The two hemispheres of the brain
can, in fact, operate independently from each other, as patients having
undergone corpus callosotomy, a surgical procedure for the treatment
14 Martyn Phillips
Strange Attractor
The imagined pure form of the mix, the practically perfect mix, exhibits
characteristics of a strange attractor in chaos theory. This can be viewed as
a goal that, as has been shown, can never be reached, only orbited around.
These particular orbits are chaotic, unlike the neat classical models of the
planets orbiting the sun in fixed orbits. This was demonstrated in 1971 by
Edward Lorentz, who showed that with only a few (three) degrees of free-
dom, a never-repeating infinite number of such orbital paths around such
an attractor could be created (Gleick, 1988:139–140).
In considering the mix as such an orbit around the idealized, if not per-
fect, objective, the point at which it is deemed complete, the process of
orbiting the target is stopped, sampled in what is known as a Poincaré, or
return, map. This is the final mix, abandoned in its orbit rather than com-
pleted, the artistic moment crystallized.
Although sometimes close to the objective, sometimes the point of
return, the mix outcome, can be radically different to that initially envis-
aged. The choice of processing used or the inability to match it with the
desired intent may have created unforeseen consequences, but even minor
changes in the choices made in the initial conditions or during the process
can result in a significantly different outcome to that expected: this is like
the popular idea derived from chaos theory of the butterfly flapping its
wings causing a hurricane on the other side of the world. Professor David
Pérez-García, co-author of ‘Undecidability of the Spectral Gap’ (Cubitt
et al., 2015) comments, “the results show that adding even a single particle
to a lump of matter, however large, could in principle dramatically change
its properties” (Knight, 2015).
Limitations
There are, at any particular time, limitations placed on how deep or pro-
found a perception can be achieved in both the technology and the human
use of it.
Technical limitations are imposed by the quantization of audio data.
Although large, the number of possible solutions to an audio problem
using a digital format is finite. This may be familiar to engineers emerging
from working with analog and finding certain sonic subtleties absent in
digital. Advances, particularly with higher sample rates and bit depths,
have ensured that this obstacle is being steadily eroded.
Exploring Potential of the Mix 15
Mixing Guidelines
The red hat deals with emotional responses, such as ‘Hate it, get rid
of it!’ Although termed a hat, the choice of acceptance or rejection orig-
inates in the inductive grey matter of the gut, so reactions in that part of
the body should be noted. This hat is also about shifting emotions: for
instance, one man’s spill is another man’s ambience. Degrees of emo-
tions can be managed with the red hat where language is often inade-
quate in the task.
The green hat is worn for creative thinking. This is the realm of ‘lateral
thinking’, the creation of fresh possibilities, unexpected leaps of percep-
tion. This mode of thinking is also concerned with one of the most import-
ant ingredients in a mix: that of humor, as created by the quantum jump
from one viewpoint to another. Stylistic references can be thought of in
terms of this, as the listener is carried from one musical stream to another.
It could well be argued that green-hat thinking is a part of what differenti-
ates the engineer from the producer.
Time and budget pressures in the mix will often dictate that the problem-
solving, black-hat thinking dominates the mix process. Beyond this, black-
or red-hat thinking advises what balance needs to be addressed. This is
done with additional yellow-hat thinking. Green-hat thinking opens up
new considerations that can then be examined in the same way.
Chance
Po
Choice of Polarity
It can be seen that many of the Oblique Strategy cards deal with finding
an appropriate level or instructing one of some parameter. This can also be
considered as dealing with the balance of a duality, opposite or polarity or
the appearance or state of a monistic aspect.
The Oblique Strategy card instructions that deal with polarities suggest
working with the following: center, accretion, level of structure, comple-
tion, personality, cleanliness, extravagance, cascades, courage, decora-
tion, self-indulgence, desire, time distortion, activity, ease, differences,
repetitions, flaws, ghost echoes, comfort, humanization, glee, gradations,
nobility, humility, credibility, intonation, absence, mechanization, idiosyn-
crasy, wholeness, uniqueness, change, heroism, ambiguities, consistency,
radio-friendliness, insignificance, novelty and note density. This collection
is perhaps an extension on Keene’s nine principles for balancing, but obvi-
ously not an exhaustive list of parameters that can be considered.
In fact, as de Bono suggests with using the word ‘po’, random words
selected for the dictionary can open up a galaxy of other parameters for
consideration that could be argued to offer comparable insights to Eno’s
suggestions. For example, in a test run of twenty random words selected
for this writing (eleven were discarded) the following words were picked:
1) Reluctance
2) Gratify
3) Extremist
4) Flock
5) Histrionic
6) Pusillanimous
7) Insomnia
8) Detraction
9) Distend
In the mix process, the same random selection of words could be reframed
as polarities to be considered within it, for instance:
detail. How much attention should the mix ask of the listener? This will
vary between genres from meditation music to avant-garde. How familiar
should the mix feel? Too much or too little and it will not inspire a need to
purchase it (if that is a source of funding for the project). A sufficient dis-
tance from what chimes with the listener’s experience can inspire a need
to own the product, to include it in their life and inspire others to follow
suit. This is a way to create a hit and an income stream for those involved
in the production.
Calculus
What Is Possible?
Numerology
Conclusion
Inviting in new perceptions on how to assess the mix increases the level
of sophistication available to the mix engineer to refine performances. The
mixer can then more accurately account for imbalances in the human con-
dition and, by more accurately resonating with them, connect with the lis-
tener sufficiently to inspire them to make an investment into the product.
This new power calls for those with the purse strings to allow for new
approaches to the mix, such as those suggested here, to be more fully
explored and not to be satisfied with just the antecedent working meth-
ods. The investment in this part of the production process should result
in a return as the audience engages more fully with the product. Holistic
awareness from mix engineers is also called for so that they may be able to
effectively engage in considering new methods.
Bibliography
5Rhythms. (n.d.). 5Rhythms | Gabrielle Roth’s 5Rhythms. [online]. Available at: http://www.
5rhythms.com/gabrielle-roths-5rhythms/ [Accessed: 22 December 2015].
Accusonics. (n.d.). Achieve the Drum Sound of Your Dreams. Drumatom: The World’s First
Drum Leakage Suppression Tool. [online]. Available at: http://drumatom.com/ [Accessed:
6 January 2016].
Anon. (n.d.). Minds, Machines and Gödel. [online]. Available at: http://users.ox.ac.uk/~jrlu
cas/Godel/mmg.html [Accessed: 3 December 2015].
Banash, D. (2006). Plato’s Theory of Forms. [online]. Available at: http://www.anselm.edu/
homepage/dbanach/platform.htm [Accessed: 28 November 2015].
Bono, E.D. (1990a). PO: Beyond Yes and No, new edition. London: Penguin Books Ltd.
Bono, E.D. (1990b). Six Thinking Hats. London: Penguin Books.
Buskin, R. (2007). CLASSIC TRACKS: Les Paul & Mary Ford ‘How High The Moon’. [online].
Available at: http://www.soundonsound.com/sos/jan07/articles/classictracks_0107.htm
[Accessed: 11 November 2015].
22 Martyn Phillips
Buskin, R. (2013). The Clash ‘White Riot’ Classic Tracks. The Clash ‘White Riot’ Classic
Tracks. [online]. Available at: http://www.soundonsound.com/sos/oct13/articles/classic-
tracks-1013.htm [Accessed: 23 September 2014].
Carr, P. (2012). The Elements of Music: A Good Start to Basic Analysis. Paul Carr. [online].
Available at: http://paulcarr.org/2012/01/27/the-elements-of-music-a-good-start-to-basic-
analysis/ [Accessed: 10 December 2015].
Celemony. (n.d.). Celemony | What is Melodyne? [Online]. Available at: http://www.celemony.
com/en/melodyne/what-is-melodyne [Accessed: 6 January 2016].
Cubitt, T.S., Perez-Garcia, D. and Wolf, M.M. (2015). Undecidability of the Spectral Gap.
Nature 528 (7581): 207–211. [Online]. Available at: doi:10.1038/nature16059.
Gibson, D. (2005). The Art of Mixing: A Visual Guide to Recording, Engineering and Produc-
tion, Second revised edition. Boston, MA: artistpro.com LLC.
Gleick, J. (1988). Chaos: Making a New Science. New York: Penguin.
Gynn, G. and Wright, T. (2008). Left in the Dark. Kaleidos Press. [online]. Available at:
https://books.google.co.uk/books?id=vOj_GZ2–1dEC.
Haggard, P. and Eimer, M. (1999). On the Relation between Brain Potentials and the Aware-
ness of Voluntary Movements. Experimental Brain Research 126 (1): 128–133. [online].
Available at: doi:10.1007/s002210050722.
Hendrikse, W. (2013). David Bowie—The Man Who Changed the World. London: New Gen-
eration Publishing.
iZotope. (2016). RX 5 Audio Editor: Noise Reduction & Audio Repair Software. [online].
Available at: https://www.izotope.com/en/products/audio-repair/rx/ [Accessed: 6
January 2016].
Jowett, B. (n.d.). Ion, by Plato. [online]. Available at: http://www.gutenberg.org/files/1635/
1635-h/1635-h.htm [Accessed: 11 November 2015].
Keene, S. (1981). Practical Techniques for the Recording Engineer, Second edition. (s.l.):
MixBooks.
Knight, G. (2015). Gödel and Turing Enter Quantum Physics. UCL-CS. [online]. Available
at: http://www.cs.ucl.ac.uk/computer_science_news/article/goedel-and-turing-enter-quan
tum-physics/ [Accessed: 23 December 2015].
LANDR. (n.d.). LANDR: Instant Online Audio Mastering Software. LANDR. [online]. Avail-
able at: https://www.landr.com/en [Accessed: 6 January 2016].
MacKinnon, E. (2012). Edison Voice Recording Is Old, but Not Oldest. LiveScience.com.
[online]. Available at: http://www.livescience.com/24317-earliest-audio-recording.html
[Accessed: 6 January 2016].
Petersen, G. (2005). Ampex Sel-Sync, 1955 | Mixonline. [online]. Available at: http://www.mix
online.com/news/facilities/ampex-sel-sync-1955/367111 [Accessed: 15 December 2015].
Roger, C. (1994). ‘Free as a Bird’ Emerging from Beatles’ Studio Reunion—Tribuned-
igital-Thecourant. [online]. Available at: http://articles.courant.com/1994–04–05/fea
tures/9404050734_1_lost-lennon-tapes-homing-bird-mccartney-and-harrison [Accessed:
21 November 2015].
Rosen, J. (2008). Researchers Play Tune Recorded Before Edison. The New York Times, 27
March. [online]. Available at: http://www.nytimes.com/2008/03/27/arts/27soun.html
[Accessed: 20 November 2015].
Ross, A. (2013). Beethoven Dada. The New Yorker. [online]. Available at: http://www.newyo
rker.com/culture/culture-desk/beethoven-dada [Accessed: 20 November 2015].
Schucman, H. (1996). A Course in Miracles Combined Volume, Second edition. New York:
Viking.
Smith, P. (2013). An Introduction to Gödel’s Theorems. Cambridge: Cambridge University
Press.
Snyder, R. (2003). Sel-Sync and the ‘Octopus’: How Came to be the First Recorder to Mini-
mize Successive Copying in Overdubs. ARSC Journal 34 (2): 209–213.
Sound Radix. (2015). Sound Radix. Sound Radix. [online]. Available at: http://www.soundra
dix.com/ [Accessed: 6 January 2016].
Williams, K. (2014). Music and the Near-Death Experience. [online]. Available at: http://
www.near-death.com/science/research/music.html [Accessed: 21 November 2015].
Exploring Potential of the Mix 23
Yonekawa, T., Nakagawa, E., Takeshita, E., Inoue, Y., Inagaki, M., Kaga, M., Sugai, K.,
Sasaki, M., Kaido, T., Takahashi, A. and Otsuki, T. (2011). Effect of Corpus Callosot-
omy on Attention Deficit and Behavioral Problems in Pediatric Patients with Intracta-
ble Epilepsy. Epilepsy & Behavior 22 (4): 697–704. [online]. Available at: doi:10.1016/j.
yebeh.2011.08.027.
Zimmer, C. (2009). The Brain: What Is the Speed of Thought? | DiscoverMagazine.com.
Discover Magazine. [online]. Available at: http://discovermagazine.com/2009/dec/16-the-
brain-what-is-speed-of-thought [Accessed: 21 December 2015].
2
Introduction
As beginners, our first mixes invariably fall short sonically in a great many
ways. We persevere and redo them, or create others, and those next novice
attempts also typically fail, often miserably. Sometimes we recognize the
24
How to Listen, What to Hear 25
Artistic Elements
From this brief examination of what we are trying to hear, let’s move
forward to explore how to listen.
How to Listen
The following sections are for learning to listen. Building our listening
skills will lead us towards hearing—perceiving and recognizing—the ele-
ments that create dimension in the mix. With diligence, we will learn to
listen deeply. In thoughtfully learning to listen, we can learn much about
ourselves, and our unique relationships to sound.
Listening Is Personal . . .
the person who used that term in the first place may not have a clear idea
of why they used the term.
Objective and more functional communication about sound is possi-
ble, however. It does take some knowledge of the physical dimensions
of sound and some focused practice and skill in identifying the charac-
teristics of the dimensions. Meaningful communication is possible, but
it takes attention to develop and some discipline not to resort to old pat-
terns (Moylan, 1992).
The states and activities of the physical dimensions of sound can be
described, as they travel through the duration of the sound. A dynamic enve-
lope can be described by its contour and levels against time; spectral content
can be addressed in specific terms of what partials are present when, and
the contours and levels of their individual amplitudes. All other aspects of
sounds can be addressed with equal precision and detail by describing their
unique dimensions and activities against time (Moylan, 2015).
This brings us to recognize: meaningful communication about sound—
with sufficient accuracy and depth to be of use in learning to listen, in
understanding recordings, and in audio production work—requires con-
siderable skill and knowledge. Thus, talking about sound in a meaningful
way can only be accomplished if one is sufficiently prepared (Schaeffer,
1966).
● Acoustics
● Psychoacoustics
● Dimensions of sound in audio recordings
● Music studies
Beginning to Hear . . .
Learning to hear these sound dimensions will require some new listening
skills. These listening skills will be directed toward identifying:
recorded sound and the mix, need to be found. X-Y graphs can be adapted
for this purpose. Observations of many aspects of sound can be plotted as
X-Y graphs. This can be helpful to track how an individual dimension of
sound might exist over time. In Figure 2.1, we can observe the component
parts of a synthesizer timbre against time. This is the Moog synthesizer
glissando that appears at 0:12, ending the introduction to The Beatles’
‘Here Comes the Sun’ (Abbey Road).
X-Y graphs can now be used to support descriptions of sounds, supple-
menting objective language with visualizations of data. Describing sound
might thus become more direct, articulate, meaningful and effective. Ver-
bal description becomes more focused and universally understood, espe-
cially when supported by the graph’s plotting of the sound’s characteristics.
The X-Y graph is highly adaptable for all of the elements of sound
we might seek to teach. Further, it can be used at any level of perspec-
tive, from showing the smallest detail to depicting the overall shape of any
element’s material. Following is a table of graphs that are central to this
approach; each is dedicated to a single element and focused at a specific
level of perspective (see Table 2.3; Moylan, 2015).
Definition
Pitch
non-pitched
pitched
mf
RDL
Dynamic
Contour
mp
HIGH
Spectral
C6
Content
MID-UPPER
G 5
E5
G 4
MID B4
LOW-MID B3
mf
RDL
Envelope
Spectral
mp
pp
0 .2 .4. .6 .8 1.0 1.2 1.4 1.6 1.8 2.0 2.2 2.4 2.6 2.8
Figure 2.1 Sound quality evaluation graph of a Moog synthesizer sound from ‘Here
Comes the Sun’ (Moylan, 2015)
Copyright © Taylor & Francis, 2015
32 William Moylan
Table 2.3 Primary X-Y graphs for artistic elements at various levels of perspective
Creating these graphs draws one into the listening process directly, and
deeply. It brings detailed discoveries into what makes each mix unique,
and how they are shaped, as well as developing listening skills.
and directly improve recognition of that element, and it will develop the
disciplines of focus and attention. It also embraces the reality that it is
not possible to hear several things simultaneously. Instead, there will be
repeated hearings of the material. Listening repeatedly to the same mate-
rial produces new results, as results are checked and observations refined,
or as the listener’s intention shifts to another element.
We will begin each listening experience with an agenda of intention;
the more specific the intention, the more effective the effort. This involves
directing attention correctly at every stage of the process. Attention can
be directed to a specific level of detail, a specific type of information and
specific aspects of sound. This is part of defining intention.
We listen and process information at various levels of detail. At each
level, sounds have certain characteristics, coalesce into a unique texture
and have certain types of relationships to other sounds. Certain qualities
might appear at one level and be significant, and not be relevant at oth-
ers. Identifying a specific level of perspective for each listening is import-
ant for defining its intention. The level of detail might be most readily
approached, especially during early learning, as ‘middle dimension’ lev-
els and ‘large dimension’ and ‘small dimension’ activities (LaRue, 1992;
Moylan, 2014). Table 2.1 provided some definition to these levels.
As skill develops, the process becomes more involved, and we engage
and contrast critical listening and analytical listening issues. We listen for
several types of information, related to the context of the material. Some
information will be related to the message of the recording (musical mate-
rials and their relationships) and some information will be unrelated to
context (such as the sound characteristics of a microphone or its place-
ment). Knowing that a sound’s qualities are being examined in isolation
(out of context) or in relation to other sounds (in context) brings the learner
deeper understanding and often a more direct way to listen to the material.
Listening with intention continues. This alternation of critical versus ana-
lytical listening practice can strengthen the sense of intention and focus in
the developing listener, particularly once control is becoming established.
Until control of material is gained, listening with intention brings greater
awareness of each aspect studied, with its dimensions and level of perspec-
tive, and develops the listening discipline to not be distracted by or pulled
to other aspects of sound (Moylan, 2015).
As a product of this process, sound memory is improved. Encouraging
listening as a separate process from writing brings us to remember what is
heard. Further, it is simply not possible to write and listen simultaneously.
Separating the two deliberately improves memory and brings greater
understanding to the material.
Once one is able to track each element of sound accurately, it is possi-
ble to move to engaging several elements. This process does not seek to
hear the qualities of several elements simultaneously. Instead, a deliber-
ate alternation between elements brings control to the process and greater
awareness to the material. After gradually improving and engaging more
sophisticated alternations of materials and levels of perspective, a time will
arrive when one can have an open listening field, scanning and listening
34 William Moylan
never before experienced just the same as those known and expected. It
allows for accurately hearing and understanding something new, some-
thing unexpected—or recognizing and evaluating something known.
This openness to all aspects of sound will allow things to be detected and
embraced that otherwise might not have been heard. From this position,
new creative ideas or solutions might be discovered, or minute degrada-
tions in the signal path detected; from this position, a sense of engaging
all dimensions of the recording will be held.
From this position of equivalence, we engage listening to all levels of
perspective with equal attention, to all dimensions of sound as equally
worthy of our attention, to all critical and analytical information as poten-
tially significant in their contributions to the whole.
Listening, that is, the deep listening required of audio, is hard work. It
demands one’s undivided attention, and it requires focused concentration.
We expend energy to keep the mind engaged. As time progresses, attention
is quick to suffer.
Remaining focused on listening is not easy. Our minds wander—a lot.
We can learn to bring awareness to this difficulty by reminding ourselves
very regularly and frequently to remain focused while listening. While we
can be quick to point at the many distractions of modern life making paying
attention more difficult now than ever before, difficulties in keeping focused
attention have always existed as part of the human condition (Nhất H a. nh,
2015). Indeed, it is part of our human nature and conditioning to constantly
shift our attention; we do this in a great many ways, and this often serves us
well—as we suddenly remember to do something important, just in time.
Here, let us remember that sound is a memory.
While we listen during the passage of time, we hear backwards in time,
considering what has happened. We do not know the duration of a sound
until it stops, we do not know the shape of a melodic line until it has
cadenced, we do not know the message of a song until it is over. Accurate
listening engages not judging or reaching conclusions until having listened
completely and considered fully what has happened. All of this happens
as a by-product of engaging an awareness of what is happening now, and
retaining (making a memory of) that experience to form larger experiences
and to compare one experience to others (Snyder, 2000; Moylan, 2015).
The wandering mind steps in the way of this process—in a great many
ways.
Often, our attention strays to switching between listening, music con-
cerns and the production process. This blurs our perception, focus and the
intention of listening—and greatly reduces our productivity. A few com-
mon examples are
brings to mind loudness, but the mix crafts balance within all elements.
Shaping the smallest details of individual sounds contributes to clarity
and blend in a great many ways and begins the pyramiding of perspective
relationships.
We then begin to use our knowledge of the sound to attempt to locate the
sound based on our previous experiences. We use our sense of self, our
sense of occupying personal space, to help determine if the sound is within
our area of proximity (where we might be able to touch the instrument or
slightly beyond our grasp), or a bit farther but still near to us (perhaps in
the same room), or perhaps at a location of considerable detachment, far
or well beyond our sense of space. Levels of timbral detail bring clarity to
these changes of relationship (Moylan, 2012).
Focus your attention on the distance cues of the McCartney’s lead vocal
in ‘Eleanor Rigby’. Hold the vocal in the center of your attention and con-
centrate on the amount of timbral detail; notice how the vocal shifts loca-
tion as low-level spectral information appears or diminishes. Some of these
changes are very subtle, and some will be very pronounced. You may hear
little or no difference at first, but once you are able to experience this qual-
ity and then hold this element within your awareness, these qualities and
their changes will unveil themselves. Listen carefully to the vocal in the
introduction, the first verse and the first chorus; track where McCartney
is in relationship to you, the listener. Once you begin to accurately per-
ceive this dimension, the substantial distance changes between sections will
become clear, and you will become aware of the subtle changes within the
verses. Listening to the two stereo versions (1987 and 2009) will reveal
subtle differences of distance, comparing them to the original mono mix
will create another contrast that should become noticeable (and ultimately
quite obvious) in its differences with repeated listenings. Remember to
keep awareness of timbral detail to determine distance; be aware not to
confuse loudness levels, or the amount of reverb present, with the distance
of McCartney’s vocal to your listening location.
KEY
Voice 1
Electric guitar 2
Maracas 3
Snare 4
Mellotron 5
f
1 2
1 1 1 1 1 1
1 1
1 3 1 2
4 4 2 2
5 5
mf 2 2 4 4 2 4 4
4 4
5 5 3 3 3
mp 2 5
5
2
1 2 3 45 6 7 8 9 10 1112 13 14 15 16 17 18 19 20 21 22
measures: 4 2 4 2 4 3 4
4 4 4 4 4 4 4
Introduction Chorus Verse 1
Figure 2.2 Musical balance graph of the opening measures of ‘Strawberry Fields
Forever’ (Moylan, 2015)
Copyright © Taylor & Francis, 2015
40 William Moylan
note how its loudness is shaped, and notice how your attention is diverted
(perhaps momentarily) when the vocal enters. Repeat this several times
and notice how you can obtain greater clarity in perceiving the changes
and subtleties of characteristics. Now try to shift your attention to the per-
spective where all sounds are held equally in your perception to notice the
musical balance (or loudness balance) of the sounds. Next, follow John
Lennon’s vocal for dynamic shape, notice carefully the changes between
sections; listen again, but shift perspective to observe the vocal’s loudness
in relation to the instruments of this section. Remember to remain focused
on loudness contours, or on loudness relationships and balance; practice
one skill at a time.
Timbres of the instruments and voices that we embraced above are
also present and functioning in this middle dimension. The mix also joins
all sounds by their overall timbre, which is the result of their inherent
sound quality and timbral alterations created by the expression and level
of energy of their performance intensity. The timbre of sounds sets up
a pitch density in the mix, whereby certain frequency/pitch ranges are
emphasized in the mix dependent upon instrument selection and the char-
acteristic timbres of those sounds. Some find it easiest to conceptualize
pitch density as a vertical dimension to the mix, where low to high pitch/
frequency is conceived bottom to top (A. Moore, 2012). Pitch density
exists at this dimension of the individual sound source, and again in the
aggregate mix; it is later reconceived as ‘timbral balance’ in the overall
texture dimension. How timbres are combined in the mix can provide
blend or clarity to sounds and musical ideas, and more. This is a rather
advanced concept that is typically learned after control of loudness and
spatial dimensions.
Beginning mixers are quick to recognize loudness and lateral locations
of sounds in the mix. In stereo, phantom images can span 90 degrees from
a point 15 degrees outside each loudspeaker, when the speakers are at a
60-degree relationship to each other. Instruments and voices are placed in
the stereo field during the mix process, and certain conventions have arisen
concerning where certain instruments and the lead vocal are placed, in cer-
tain types of music or to be most effective. The sound field is not comprised
of placement alone. Images also have width. The width of images coupled
with the location of images in relation to others will play very important
roles in the prominence, clarity and blend of sounds. Image sizes can range
from a very precise point source to occupying the entire width of the sound
stage; obviously most images fall between these extremes.
Figure 2.3 plots the stereo location of the primary sound sources in
the opening of ‘A Day in the Life’; two separate tiers allow us to see the
sounds more clearly. Note the different widths and locations of the sounds,
and how the acoustic guitar and piano images change in width at certain
points during the passage. Bring your attention to Lennon’s vocal, and
follow it as it moves gradually across the sound stage, and how its width
changes, sometimes subtly and sometimes markedly. Listen carefully to
the placement and widths of the percussion sounds, and notice how they
relate to and complement those sounds on the graph.
How to Listen, What to Hear 41
KEY
Voice Maracas
Piano
Bass Guitar
1 2 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31 33
4 7 4
4 8 4
Intro Verse 1 Verse 2 Verse 3 Bridge
measures of 4 (except where marked)
4
Figure 2.3 Stereo location graph of ‘A Day in the Life’ (Moylan, 2015)
Copyright © Taylor & Francis, 2015
C10
F#9
VERY HIGH C9
F#8
C8
4186 Hz
F#7
HIGH
C7
#
F6 1397 Hz F 6
E6 1319 Hz
MID-UPPER
C6
F#5
C4
B4 494 Hz
A4 440 Hz
F#4
MID
C4
A3 220 Hz
G3 196 Hz #
F 3
LOW-MID
C3
F#2
D2 73 Hz
G2 65 Hz
F#1
LOW
C1
F#(23 Hz)
L C R
Figure 2.4 Sound source frequency content against stereo imaging
Allan F. Moore (Rock: The Primary Text) has devised a way of charting
three elements simultaneously. His ‘soundbox’ plots pitch density (ver-
tical), stereo location (lateral) and depth (distance) simultaneously with
a quasi-three-dimensional box. The soundbox can bring visualization to
sections of a mix with a clarity that many have found helpful, and it has
proved a useful tool in both studies of mixes and in practice of planning
and executing mixes.
A recording’s sound stage is the area that encompasses the lateral and
distance locations of all of the instruments/voices of a mix. It establishes
and defines the left/right and front/back boundaries of the illusory stage
from which the ‘performance’ that is the recording emanates. The imaging
How to Listen, What to Hear 43
High Hat
Bass Drum
Perceived
Depth Acoustic Guitar
of
Sound High Kybd Low Keyboard
Stage
Lead Vocal
Background Vocals
Bass
Flute
Tambourine
Perceived Width
of Sound Stage
Listener’s Perceived
Location
Figure 2.5 Sound stage and imaging, with phantom images of various sizes and at
different distances (Moylan, 2015)
Copyright © Taylor & Francis, 2015
44 William Moylan
The sound stage can be located anywhere around the listener. Its size
can encompass any portion of the 360 degrees around the listener. The
position of the sound stage relative to the listener might be reduced to (1)
the extent to which the sound stage resembles traditional stereo, (2) the
placement of ambiance and the perceived performance environment and
of the individual sources, and (3) the presence of an additional sound stage
or of sounds outside of the sound stage.
Figure 2.6 plots the opening of ‘I Want to Hold Your Hand’ (from LOVE).
Here the sound stage is in front, as in stereo but wider, and the ambience is in
the rear. This is a common approach that is related to conventional imaging,
KEY
4 2
9 9
8 1
3
7
10
Figure 2.6 Surround sound stage of ‘I Want to Hold Your Hand’ from LOVE
46 William Moylan
with added control of the environment in the rear. In practice, the front-
center channel may change imaging matters considerably, as can be heard in
this example’s pronounced clarity to the individual musical lines in contrast
to the original mono version. This recording displays a width of sound stage
that extends slightly extending beyond the 90-degree spread of stereo, though
sources are largely grouped as if performing together live. The environment
cues are placed to the rear or to the sides of the listener, and are joined by
screams of crowd noise to depict some semblance of a live performance.
Imaging in surround sound is vastly more complicated than in stereo. In
stereo, phantom images are established by the interaction of sound emanating
from two loudspeakers; the resulting images may appear anywhere between
the two speakers or up to 15 degrees beyond. 5.1 surround-sound imaging
substantially changes the concepts of phantom image size and placement,
image stability and listener’s relationship to the music and the ensemble.
As seen in Figure 2.7, there are five primary phantom image locations
existing between adjacent pairs of speakers in surround. These images tend to
be the most stable and reliable between systems and playback environments.
Many secondary phantom images are possible as well. These can be
created and appear between speaker pairs that are not adjacent. These
images contain inconsistencies in spectral information and are inherently
less stable. Implied are different distance locations for these images, as the
C
* *
L R
× × × ×
* *
LS
* RS
trajectories between the pairs of speakers are closer to the listener position.
These closer locations do not materialize in actual practice. The distance
location of these images are actually pushed away somewhat by the dimin-
ished timbral clarity of these images. This can create contradictory and
confusing sonic impressions.
When we consider locations caused by various groupings of three
or four loudspeakers, placement size and location options for phantom
images get even more complex—and less predictable.
This wealth of opportunity for mixers creates a complex set of possi-
bilities for imaging in surround and generates great potential for other
speaker-combination approaches to imaging. The format for 5.1 surround
provides the opportunity for as many as twenty-six possible combinations
of speakers. Table 2.4 lists the potential phantom images in surround that
may be established by the interaction of any of the five loudspeakers, in
any combination (Moylan, 2017).
As with imaging in stereo, the edges of phantom images are more deci-
sive to defining source locations and size than are their centers. The edges
of images allow us to understand and identify widths of images and where
and how they might overlap. With this awareness of the edges of images,
we can recognize how sounds are separated in space or blended. The sep-
aration of source images in surround can be very marked, and may grow
more pronounced as the width of the sound stage increases.
When using non-adjacent speakers for imaging, sounds blend and fuse
differently. Image placements and size can be unstable and vary quickly
with any listener movement. Sounds can both separate themselves from
others or become ill-defined, almost transparent sonically and largely
masked. When the imaging occurs from speakers on either side of the lis-
tener, the image can be confusing, as it localizes within the conceptual lis-
tener location, yet there is the impression of detachment due to diminished
Table 2.4 Possible combinations of speakers for generating phantom images in 5.1
surround sound; combinations exclude potential inclusion of the subwoofer
KEY
2 3 7
5
8
1
9
9
4
6
2 3
Figure 2.8 Surround sound stage of ‘Eleanor Rigby’ from LOVE, 0:00–1:09
How to Listen, What to Hear 49
Hearing the mix at the highest level of perspective, we engage the overall
sound—the sound of the mix as one, blended entity. Final shaping of this
level of perspective often is the role of the mastering engineer, though the
mix engineer fundamentally establishes these qualities and relationships.
The primary elements of this perspective (the overall musical texture) are
timbral balance, program dynamic contour and perceived performance
environment.
We hear one overall environment for the mix as the PPE. This is the
‘performance environment’ of the sound stage, or the world within which
the performance that is the mix takes place. The PPE can be crafted for the
mix, but it is most often perceived through the interactions of the environ-
ments of the mix’s instruments and voices. Its dimensions shape the entire
recording and bind all individual sound sources and their spaces into a
single performance area (Moylan, 1992, 2015).
Timbral balance might be considered the frequency response of the
mix, and it represents the recording’s ‘spectrum’. It is the distribution and
50 William Moylan
are very rich in detail and vast in their influences on the mix and on the
music. How the mix, or more fully how the recording, transforms music
is examined deeply in the author’s new book: Recording Analysis: How
the Record Shapes the Song (Routledge, 2017).
The relationships between the dimensions of the mix and the creative
process of crafting the mix are also examined throughout the author’s
Understanding and Crafting the Mix: The Art of Recording (Focal Press,
2015). A significant portion of the book is dedicated to developing lis-
tening skills; some of which were introduced above. Understanding and
Crafting the Mix also contains greater detail on the materials presented in
this chapter, and applies it to analyze mixes for artistic elements and much
more. It explores mixing as a creative process, in other words.
To conclude, let us remember that Mixing Music begins with listening.
That listening is a focused activity to be learned; it is a skill that can be
developed. Listening is paying attention, and remembering to pay atten-
tion. Listening is bringing attention to a specific aspect of sound, or hold-
ing an open awareness for what might arrive. Listening can be seeking
specific information, and can bring awareness to any level of detail and any
element of sound. We listen without preconceived ideas or expectations,
and hold the possibility of the unknown arriving in the next instant.
Listening also requires knowing what to listen for—what to seek to
hear. Listening and hearing work in tandem. Listening takes place during
the passage of time, an event unfolding moment by moment in the present.
Hearing takes place in memory; considering what has happened, we hear
backwards in time.
If hearing is perceiving and recognizing, hearing only really arrives
with knowledge and experience; these inform and refine perception to
bring realization and understanding. It is for this reason that we carefully
examine the unique qualities of recordings and refine our skills at hearing
subtle details—so we might become aware of how those details are shap-
ing our recordings, and so we might embrace mixing as a creative process.
We can learn to conceive and compose our own mixes, and establish
our own musical voices—once we are in control of the recording and mix
processes, and once we know what to hear, and learn how to listen.
Bibliography
Butler, David (1992). The Musician’s Guide to Perception and Cognition. New York: Schirmer
Books.
Clifton, Thomas (1983). Music as Heard: A Study in Applied Phenomenology. New Haven,
CT: Yale University Press.
Deutsch, Diana (2013). The Psychology of Music, Third edition. Orlando, FL: Academic
Press.
Handel, Stephen (1993). Listening: An Introduction to the Perception of Auditory Events.
Cambridge, MA: The MIT Press.
Holman, Tomlinson (2008). Surround Sound: Up and Running, Second edition. Boston, MA:
Focal Press, Elsevier.
LaRue, Jan (1992). Guidelines for Style Analysis, Second edition. Detroit: Harmonie Park
Press.
52 William Moylan
Levitin, Daniel J. (2006). This is Your Brain on Music: The Science of a Human Obsession.
New York: Plume.
Moore, Allan F. (2002). Rock: The Primary Text. Aldershot: Ashgate.
Moore, Allan F. (2012). Song Means: Analysing and Interpreting Recorded Popular Song.
Burlington, VT: Ashgate.
Moore, Brian C.J. (2012). An Introduction to the Psychology of Hearing, Sixth edition. Lon-
don: Elsevier Academic Press.
Moulton, David (1995). Golden Ears: Know What You Hear. Sherman Oaks, CA: KIQ Pro-
duction, Inc.
Moylan, William (1992). The Art of Recording: The Creative Resources of Music Production
and Audio. New York: Van Nostrand Reinhold.
Moylan, William (2012). ‘Considering Space in Recorded Music.’ In The Art of Record Pro-
duction: An Introductory Reader for a New Academic Field, eds. Simon Frith and Simon
Zagorski-Thomas. Surrey: Ashgate Publishing Limited, pp. 163–188.
Moylan, William (2014). ‘Pathways Through Recording Analysis.’ Delivered to the 137th
Audio Engineering Society International Convention, Los Angeles, October 9. Paper
Proceedings.
Moylan, William (2015). Understanding and Crafting the Mix: The Art of Recording, Third
edition. Burlington, MA: Focal Press, Taylor & Francis Group.
Moylan, William (2017). Recording Analysis: How the Record Shapes the Song. New York:
Routledge, Taylor & Francis Group.
Nhất H anh,
. Thích (2015). Silence: The Power of Quiet in a World Full of Noise. New York:
HarperCollins.
Polanyi, Michael (1962). Personal Knowledge: Towards a Post-Critical Philosophy. Chicago:
University of Chicago Press.
Roederer, Juan (2008). Introduction to the Physics and Psychophysics of Music, Fourth edi-
tion. New York: Springer.
Schaeffer, Pierre (1966). Traité des objets musicaux. Paris: Editions du Seuil.
Schaeffer, Pierre and Guy Reibel (1966). Solfège de l’objet sonore. Paris: Editions du Seuil.
Snyder, Robert (2000). Music and Memory: An Introduction. Cambridge, MA: The MIT
Press.
Tenney, James (1986). Meta ≠ Hodos and META Meta ≠ Hodos. Oakland, CA: Frog Peak
Music.
Thompson, William Forde (2015). Music, Thought, and Feeling: Understanding the Psychol-
ogy of Music, Second edition. Oxford: Oxford University Press.
Discography
Beatles, The. ‘A Day in the Life’, Sgt. Pepper’s Lonely Hearts Club Band, EMI Records Ltd.,
1967, digitally re-mastered 1987. CDP 7 46442 2.
‘Eleanor Rigby’, LOVE, EMI Records Ltd., 2006. 0946 3 79810 2 3/0946 3 79810 9 2.
‘Eleanor Rigby’, Revolver, EMI Records Ltd., 1967, digitally re-mastered 1987. CDP 7
46441 2 0.
‘Eleanor Rigby’, Revolver, EMI Records Ltd., 1967, stereo digital re-mastered 2009. 0946
3 82417 2.
‘Eleanor Rigby’, Revolver, EMI Records Ltd., 1967, mono digitally re-mastered 2009. LC
0299 509999945823.
‘Here Comes the Sun’, Abbey Road, EMI Records Ltd., 1969, digitally re-mastered 1987.
CDP 7 46446 2.
‘I Want to Hold Your Hand’, LOVE, EMI Records Ltd., 2006. 0946 3 79810 2 3/0946 3 79810
9 2.
‘I Want to Hold Your Hand’, Meet The Beatles! EMI Records Ltd., 1964, re-mastered 2004.
CDP 7243 8 66878 2 1.
‘Strawberry Fields Forever’, LOVE, EMI Records Ltd., 2006. 0946 3 79810 2 3/0946 3 79810
9 2.
‘Strawberry Fields Forever’, Magical Mystery Tour, EMI Records Ltd., 1967, digitally
re-mastered 1987. CDP 7 48062 2.
3
53
54 Ruth Dockwray
Hall, 1969: 126–127).2 Indeed, it is worth noting at this point that each
type of interaction as discussed by Hall that takes place within each zone
is not universal and, as he states, “proxemic behaviour . . . is culturally
conditioned and entirely arbitrary” (Hall, 1969: 115).
With Hall’s zones, he offers a specific way of discussing space and differ-
ent types of distances, which can be applied to recorded song. More specif-
ically, the interpersonal relationship and zone location that is focused on is
that between the listener’s perceptual position and the position of the song’s
persona, as modified by its (personic) environment (see Moore, 2005).
The use of Hall’s theory enables the listener to determine elements and
qualities of each zone through the concept of ‘personal space’ (the individ-
ualized space that surrounds a person) and the concept of ‘interpersonal
distance’ (the related distance between two persons). To identify the per-
ceived distance between the listener and the persona, the listener needs
to engage in the overall scene of the track, which offers aural clues as to
spatial placement and perceived proxemic distance.
Drawing on Hall’s bodily dimensions, some important qualities refer
directly to the analysis of recordings (‘aural’ and ‘oral’) and others are
merely implied (‘vision’) when articulating the zone placement of the
Proxemic Interaction in Popular Music 55
persona. The oral and aural aspects of Hall’s receptors in proxemics per-
ception can be directly applied to recordings, such as whispering and soft
voice in the closer zones and loud voice and full speaking voice in the
distant zones. These vocal distinctions can be identified as ‘qualifiers’ and
form part of the paralinguistic features that Lacasse (2010) discusses in his
work on phonographic staging, which enable the aural interpretations of
the interpersonal distance and placement of the persona to be made by the
listener. Lacasse also discusses ‘alternants’ (vocal sounds that can include
“inhalations and exhalations and gasps”), which as he mentions are often
used in a musical context, but when highly audible may also be identified
as being part of the intimate or personal zone.
While these paralinguistic features may go some way in articulating the
persona, the loudness or intensity of the voice needs to be taken into account
and is again relative to the personic environment. As Moylan (2007: 140–
142) discusses, the “perceived performance intensity” and the “dynamic
level” are both important in the listener’s perception of the persona’s prox-
emic placement and the relationship between the persona’s performance
environment. The perceived performance intensity is the original level of
the sound source, which may be altered in the overall dynamic level of the
track. This allows, for instance, a vocal whisper to be heard at a higher level,
relative to the rest of the sonic sources in the recording. While the perceived
loudness has been manipulated, the sonic properties of the whispered voice
have not changed and therefore give the listener clues as to perceived proxe-
mic location of the persona. In this sense, the loudness can serve to enhance
the interpersonal distance between the listener and the persona, particularly
when analyzing the intimate zone. In terms of the intimacy that can be cre-
ated between then listener and person, Nicola Dibben (2009) discusses the
effect of the voice within the mix and the importance of vocal characteristics
in the articulation of the persona and represented emotions, in terms of “per-
forming intimacy” (p. 319). In this sense, amplification “creates intimacy
between listener and singer, and communicates the ‘inner’ thoughts of the
song character and/or performer” (Dibben, 2009: 320).3
Kylie Minogue’s track ‘Slow’ presents an example of this type of inti-
macy, which can be perceived by the listener. Kylie’s persona is presented in
an intimate zone as evidenced by her audible vocal qualifiers and alternants,
such as her whispered voice and clarity of breath intakes. Her intimacy and
close proximity to the listener is also made evident through the lyrics that
also situate the persona in the intimate zone. Hall’s close-intimate phase
descriptors highlight the potential of physical contact and “the high possibil-
ity of physical involvement is uppermost in the awareness of both persons”
(Hall, 1969: 117). What is also interesting about this example is how Kylie’s
persona occupies a wide area of the sound-box, which suggests closeness to
the listener and emphasizes Hall’s ‘vision’ aspects of each zone. Vision, in
terms of head size and peripheral vision, can be perceived in terms of sonic
placement within the sound-box; however, in order to illustrate this notion,
the concept of the ‘framing variable’ needs to be discussed.
The relationship between the perception of interpersonal distance and
the framing variable considers the notion of para-proxemics, a term coined
56 Ruth Dockwray
(semi-whisper) and audible vocal breaths due to the close miking of the
sound source. His lyrics do not suggest intimacy, but rather are centered
on himself and his search for something. The lyrics are also indicative
of his close proximity, as he makes clear that “nothing satisfies but I’m
getting close, closer to the prize at the end of the rope”. Additionally, the
58 Ruth Dockwray
guitar in the introduction also plays a key role in the proxemic positioning
of Grohl, as it provides a point from which the listener can perceive the
vocal’s relative distance from the listener and its distance from the guitar.
The wide sound source of his persona, which spans the entire scene width
and high definition of the vocals, brings Grohl into focus and perceptually
closer to the listener.
The ensemble width, or rather, the perceived width of the main group of
sound sources, allows for the differentiation between other sound sources
that are spatially detached from the rest of the ensemble. Source separa-
tion and increased source definition can also provide the listener with an
increased perception of the degree of distance the persona is from the rest of
the ensemble and from the listener. The sonic attributes of each sound source
provide aural clues to spatial positioning; for instance, the lower dynamic
and narrow source width of the guitar immediately places the guitar behind
the vocals.
Half a minute into the track ‘All My Life’, the group width is increased
by the entering of the rest of the sound sources and now seems to encom-
pass the entire scene width. The persona is no longer detached from the
environment and is enveloped by the individual sound sources of the
ensemble, reducing its definition from being in focus and detached to
appearing as part of the ensemble and less in focus. The difference here is
that the guitars panned to each side appear to be closer to the listener, and
the persona has changed in terms of vocal delivery (now shouting) and is
dynamically louder in an attempt to be heard over the rest of the ensemble
(or personic environment)—placed in a social zone.
Interestingly, the speed with which the persona moves from one zone to
another highlights the possible discomfort Grohl feels about being in an
intimate zone and in such close proximity, disclosing personal facts. His
need to quickly come out of that zone into a more social zone suggests
there has been an invasion of personal space, which he clearly is uncom-
fortable about. He later returns to the intimate zone before the chorus,
albeit briefly.
Changes of proxemic zone are perhaps among the most interesting uses
of space within recordings. Leona Lewis’s cover of Snow Patrol’s ‘Run’ is
one such example and is representative of the gradual movement from one
zone through the others. The track begins with Lewis in an intimate zone,
backed by a resonant piano very much in the background. The piano is
joined at 45" by mid-range sustained strings, providing a carpet of sound.
At 1'08", both strings and Lewis achieve an upper range in terms of reg-
ister: while the degree of reverb has not altered, it is as if she has stepped
back, as the greater range opens her words to a wider audience. By 1'38",
she is no longer foregrounded by her environment, but the whole texture
has receded into a social space, or rather the environment has moved
towards us. In the second verse at around two minutes, the drum kit (which
had previously consisted simply of on-beat ride cymbals, which had been
crucial in bringing the environment to the fore) falls into a conventional
pattern, and any remaining illusion of intimacy is lost. (She now seems
proud of her profession of constancy.) By 3'20", both she and the strings
Proxemic Interaction in Popular Music 59
appear with full force and are subsequently joined by a full gospel choir,
any sense of restraint lost, as if she no longer cares who hears. Lewis’s
persona is fully enveloped by the environment. Thus, the increased envel-
opment of the persona by the environment clearly marks the change in
proxemic zone from intimate to public space.
Moore (2005) discusses how the persona can be articulated through the
lyrics and the importance of lyrics in determining proxemic placement and
the degree of interpersonal distance. As Table 3.1 illustrates, the lyrical
content also gives the listeners clues as to the person or persons being
addressed and the location of the persona with a particular proxemics zone.
One such distinguishable feature is the use of pronouns, which articulate
who is being addressed. The combination of vocal attributes, lyrics and
personic environmental features enable the listener to identity the different
zones, as evident in this next example.
With regard to the social and public zones, the perception of proxe-
mic placement and interpersonal distance relies not only on the level of
personic environmental envelopment but on lyrical modes of address.
Hall’s descriptors for the social and public zones state that other people
are seen and are important in peripheral vision, alluding to the notion of
co-presence and group communication, as opposed to the intimate zone,
where communication is to an individual. Group communication relies on
public modes of address, increased voice level and the use of full voice.
While the voice level is “a significant variable in judging distance” (Hall,
1963: 1016), it is the lyrical content and public mode of address in this next
example that places the persona in a more social/public zone.
In the rock anthem ‘We Will Rock You’ by Queen, the listener locates
Mercury’s persona as being in a social space. The full-voice vocals, a
medium distance away from the listener, are evidently communicating to
a group of people through Mercury’s persona being situated within the
environment consisting of multiple hand claps, suggesting co-presence.
The social zone location remains throughout the track, with little differ-
ence between the verse and chorus in terms of Mercury’s vocals. The only
change occurs with the personic environment, which thickens in texture
during the chorus due to the addition of multi-backing vocals, thus wid-
ening the overall width of the sound source and engulfing the lead vocals.
The public mode of address as evidenced by use of the pronoun ‘we’
and the interchangeable use of ‘you’ from the singular to plural serve to
emphasize the social/public proxemics placement and the increased inter-
personal distance from the listener. Such is the nature of rock anthems that
they rely on group participation and a sense of co-presence. Rock anthems
essentially induce gestural and vocal participation, which creates a sense
of community through this interaction with the listener and the audience
in live contexts (Dockwray, 2005).
There are examples of rock anthems where several different zones can
be heard across the track. The different audible zones are particularly
significant for rock anthems as they act as a cue for audience participa-
tion, which is typically during the chorus. In another of Queen’s tracks,
‘We Are the Champions’, the verses situate Mercury in a personal zone,
60 Ruth Dockwray
Notes
1 See Moore (2002: 120–126) on the concept of the sound-box; Moore and Dockwray (2008)
and Dockwray and Moore (2010) for the uses of the sound-box and taxonomies of sound-
box uses (Moore, 2012).
2 The application of proxemics is used in the work by Maasø (2008), which focuses on the
mediated spoken voice in a film sound context. Interestingly, he notes that the near-far
phase seems applicable in a mediated voice context, as opposed to recorded tracks, where
the different levels of intimacy can be differentiated, through vocal qualities and lyrics.
3 For more analysis on the application of proxemics in terms of intimacy and the intimate
singing voice, see Dibben (2013).
Bibliography
Bregman, A.S. (1990). Auditory Scene Analysis: The Perceptual Organization of Sound. Lon-
don: MIT Press.
Clarke, E.F. (2005). Ways of Listening: An Ecological Approach to the Perception of Musical
Meaning. New York: Oxford University Press.
Dibben, N. (2009). ‘Vocal Performance and the Projection of Emotional Authenticity.’ In The
Ashgate Research Companion to Popular Musicology, ed. D.B. Scott. Farnham, Ashgate,
pp. 317–334.
Dibben, N. (2013). ‘The Intimate Singing Voice: Auditory Spatial Perception and Emotion in
Pop Recordings.’ In Electrified Voices, eds. D. Zakharine and N. Meise. Göttingen, Nied-
ersachs: V & R Unipress, pp. 107–122.
Dockwray, R. (2005). Deconstructing the Rock Anthem: Textual form, Participation and Col-
lectivity. PhD thesis, University of Liverpool.
Proxemic Interaction in Popular Music 61
Discography
Foo Fighters. (2002). “All My Life,” One by One. CD, RCA 82876 50523 2.
Lewis, Leona (2008). “Run,” Spirit (deluxe edn). CD, J Records 88697 46016–2.
Minogue, Kylie (2003). “Slow,” Body Language. CD, Capitol CDP 7243 595645 0 0.
Queen. (1977). “We Will Rock You” and “We Are the Champions,” News of the World. CD,
EMI Parlophone—CDPCSD 132 (1993, Remastered Reissue).
4
62
Top-Down Mixing—A 12-Step Mixing Program 63
For my first step, I would set the main lead vocal fader to zero—that can be
the group master or the individual fader if I have more than one lead vocal
track. Also, I set my master fader to zero. If the vocals are well recorded,
this will be a good start point and hopefully a good null-point setting for
the mix overall.
well behind the lead vocal in a supporting role, making sure that it is not
fighting for space with the vocal; it could always be raised later in the mix
process.
a. A feed from the internal guitar pick up via the direct injection (D.I.) box.
b. A small diaphragm condenser microphone over the acoustic hole,
angling in from the fretboard—typically a Neumann KM84.
c. A large diaphragm condenser microphone on the body of the acous-
tic guitar, below the guitarist’s strumming hand but out of the way of
being hit by that hand—typically a Neumann U87 would do this job
well, as would an AKG C414.
Now I would start processing the lead vocal. Below is my standard set of
vocal mix techniques:
a. Insert a vocal compressor starting with a 3:1 ratio and the threshold set
so that the gain reduction meter is only active on the louder notes.
b. Set up an auxiliary send to a vocal plate reverb with a decay of about
three seconds and a high-pass filter (HPF) up to 150 Hz.
c. Set up an auxiliary send to a crotchet mono delay effect with around
35% feedback and 100% wet.
d. The equalizer settings are entirely dependent on how the vocals sound,
but typically if the vocals were recorded flat (and I will not know unless
I have recorded them), then I would boost a few decibels (dB) around
10 kHz or 5 kHz and consider a 4 dB cut at 300 Hz to 500 Hz and also
an HPF up to 100 Hz, provided this doesn’t lose the ‘body’ of the vocal
sound. If the vocal is already sounding too thin, then I would try a boost
from around 150 Hz to 250 Hz, but no lower than that, as I would want
to save any boost of 100 Hz downwards for kick drum and bass only.
e. My final ‘go to’ on a lead vocal, and I apply this later in the mix,
is the Roland Dimension D at its lowest setting—Dimension 1. The
Universal Audio plug-in virtual copy of this piece of hardware is a
good replacement and again this would be on an auxiliary send, in
addition to keeping the original signal in the stereo mix. The effect on
the Dimension D is still set to 100% wet and balanced behind the lead
Top-Down Mixing—A 12-Step Mixing Program 65
The next step is for me to bring in the rest of the vocals behind the lead
vocals—any double- or triple-tracked lead vocals would be 5 dB below the
lead vocal. This will give the effect of ‘fattening’ the lead without losing
66 Phil Harding
the character of the main lead that we are likely to have spent hours editing
and tuning after the recording. Typically, this works very well in a pop cho-
rus. All of the processing I have described for the main lead vocal would
generally go onto the double-track lead vocal except for the crotchet delay.
I tend to leave that just for the single-lead track, otherwise it can sound
messy if it is on the double as well. Next would be any harmony vocals
recorded to the lead vocal, generally a third up or maybe a fifth below. If
the harmonies are single-tracked, then they would remain panned center.
If they were double-tracked, then I would pan them half left and right, or
even tighter. Processing on these would be similar to the lead vocals but
with no delay effects. Finally, to complete the vocal stage of the mixing,
we move onto the chorus backing vocal blocks, which would often start
with double- or quadruple-tracked unisons to the lead vocal in the chorus.
This is to add strength and depth, as well as a stereo image with these
panned fully left and right. From there, all of the other harmonies in the
chorus would be panned from the outside fully or, for instance, half left
and right for the midrange harmonies, tight left and right at 10 o’clock
and 2 o’clock for the highest harmonies. All of these need to be at least
double-tracked once to achieve a true stereo. The processing would be
applied on the stereo group fader these vocals are routed to. This saves
the computer system DSP by not processing the individual tracks. Typical
backing vocal processing would be compression first, set similarly to the
lead vocal, equalization, again similar to the lead vocal but less low mid
cut and minimal HPF. The vocal reverb would stay the same, though it’s
worth considering a longer reverb time, four seconds or higher to place
the backing vocals farther back from the lead vocals. I would not put the
crotchet delay on the backing vocals except for a special, automated effect
on one or two words. I send the backing vocals to a small amount of quaver
delay overall to give them a different and tighter perspective to the lead
vocals. Multiple tests and use of this methodology since the 1990s have
proven to me that this is a repeatable formula for all pop and dance mixes.
I may wish to vary my iteration of this with more delays and processing
for extended and club mixes (especially by more use of the crotchet delay
on the backing vocals), but for a radio and video mix, the above techniques
almost guarantee an industry standard and accepted sound.
Pianos and keyboards add more musicality and support underneath the
new vocal stereo sound spectrum that has been created. In terms of time,
steps 1 to 5 could take half a day of your time; so if I have not already done
so, I would take a break before step 6. I would advise taking a break every
two hours to rest your ears and equally so these days the eyes, which have
been constantly staring at computer screens throughout this process. Some
people say that our ears are only good for four hours work per day. I am not
sure I agree with this, especially if regular breaks are taken. Often I will
be happy to stop my day’s work after step 5 and come back completely
Top-Down Mixing—A 12-Step Mixing Program 67
Other keyboard parts really have to be treated on their own merits, and
ideally one should find a space for them that fits behind either the main
piano or pad—or they jump out as a feature of their own. If there were
programmed string and orchestra parts, I would deal with them next.
Typically, for strings I like to have a longer plate or hall reverb setting,
at around four to five seconds in length and ready to go on an auxiliary
send. I would have an HPF on the strings up to 150 Hz and definitely no
LPF; I prefer a completely flat top end. For brass, programmed or live,
I would send them to a small plate or a room setting at around 1.5 sec-
onds in length. Other rhythmic keyboard or sequence parts I leave until
I get the drums in, but certainly I would check them at this stage for any
need to add compression and equalization, or possibly rhythmic quaver
delay to bed them into the track. It’s not unusual to have strings and
orchestra on pop productions, whether they are real or programmed
samples—the mix processing on both are very similar. For the orches-
tra hall reverb, where I’ve already explained the settings, I would only
send the violins (first and second) to this, plus the harp (often arranged
to work with the strings), but a lesser amount on the violas. I would
generally keep celli and double basses dry. If there were woodwinds as
well, I would send them to the vocal reverb plate so that they are a little
tighter than the violins. My ideal string recording setup is described in
Howard Massey’s excellent book The Great British Recording Studios
(Massey, 2015: 178). My equalizer recommendations on a live orches-
tra would be:
a. An HPF on the violins to cut out celli and double bass spillage up to
300 Hz. Then a small boost between 8 kHz to 12 kHz.
b. For the violas, I would use an HPF up to 100 Hz and a small top end
boost at 5 kHz.
c. For celli, I would generally leave them flat other than a small boost at
3 kHz if they need it to cut through the balance.
d. The double basses I would keep flat.
e. For the harp, I would consider an HPF up to 200 Hz, a small cut at 300
Hz to 900 Hz and small boost at 8 kHz to 10 kHz.
f. For the woodwinds, a small boost of around 5 kHz to 10 kHz and an
HPF up to 100 Hz.
All of this should help the orchestra to blend together and to blend
into the track. Notice there has been very little low middle cutting on the
orchestra in my loathed 300 Hz to 900 Hz frequencies.
Top-Down Mixing—A 12-Step Mixing Program 69
I would generally deal with acoustic guitars before electric guitars. If the
production were centered on an acoustic singer/songwriter, then the artist’s
main guitar part would have been my first support instrument of choice
while processing the lead vocal. I described in step 3 how I would record an
acoustic guitar for a singer/songwriter-orientated production, and I would
now record any acoustic guitar part the same way for safety, but in a
double- or triple-tracked guitar backing I would only use the microphone
that was on the guitar sound-hole (Neumann KM84). The equalization
choices for the acoustic guitar on the mix would typically be HPF up to
100 Hz; a 2 dB to 6 dB cut at 300 Hz to 900 Hz, and if it sounds too dull
then apply boosts at 5 kHz and 10 kHz. All of this is for the multitracked
one microphone acoustic. If you have just one main acoustic and you are
using the multi microphone technique and direct injection that I described
in step 3, then I would pan the ‘body’ condenser signal hard or half left and
leave it virtually flat other than a 300 Hz to 900 Hz cut, I would even con-
sider a 150 Hz–200 Hz boost or 2 dB to 4 dB to give more depth. I would
then pan the sound-hole microphone hard or half right and duplicate the
equalization described for the multitrack acoustic but possibly with the
HPF up to 200 Hz. The direct injection signal would feed in behind the
stereo microphones in the middle, probably kept flat, but at this stage you
should check the phasing of the three signals combined and finally con-
sider a tiny bit of the vocal reverb plate on the right-hand signal to help
blend the acoustic guitar into the track. Certainly, I would avoid room,
ambience or hall reverb.
Electric guitars are so technically varied these days I would again record
three signals if I were involved at the recording stage:
In the hope that the guitar sounds are well sourced and well played—I
do little or nothing at the mix stage. I generally do not touch the low fre-
quencies or lower mid-frequencies, deliberately leaving them in because
I am cutting them so much elsewhere. If the guitar sounds at all dull,
I only boost around 4.5 kHz, as I am trying to leave 5 kHz and above to
the vocals, piano, acoustic guitars and cymbals on the drums. I would
only consider any reverb or delays if the guitarist has not used any pedals
or guitar amp effects. Thankfully, the technology of guitar effect ped-
als and the effects on guitar amps is good enough now, I believe, for
engineers to trust their quality and low noise ratios. I generally trust the
guitarist to deliver the sound that feels right to them on their amplifier
to suit the track. If the guitars have been recorded flat and dry, I would
add some short plate or ambience reverb between one to two seconds
70 Phil Harding
in length and some quaver delay with around 30% feedback for rhythm
guitars. Usually I would apply some longer reverb and crotchet delay for
solo and lead guitar parts.
Step 9. Bass
Much like the electric guitars, I do very little to the bass. I would hope to
have D.I. and amplifier signals and I would use a single large diaphragm
condenser microphone on the bass amplifier at the recording stage (Neu-
mann U87). I compress the signals at the first stage of the mix insert chain,
being careful to only use 2:1 ratio, as anything above that can destroy the
low frequencies. I generally boost both signal paths at 100 Hz (I have been
saving this frequency in this ‘top-down’ mix method exclusively for the
bass). Finally, this is where I would try the Roland Dimension D using
the same settings and auxiliary send setup as I did on the lead vocal. This
will add a stereo perspective to the bass and some more warmth. I also add
a small boost around 1 kHz to 2 kHz (the only time I use these frequen-
cies), after the drums are in, if the bass is not cutting through the mix or
sounds as if it needs more edge. Regularly now I will also experiment with
the Sansamp on the D.I. signal—this is a very useful amplifier simulator
plug-in that comes free with all versions of Pro Tools.
Finally, we get to the drums, which for the traditional rock, pop and dance
genre mix is usually the first step. What I have found strange and yet
enlightening in my 40-year career is the similarity in the methods and
sometimes even sounds for drums, certainly the processing of them, which
can work for all three genres: rock, pop and dance. Generally metal rock
drums will need a different specialist approach, but it is astounding that
similar compressions, gating, equalization and ambiences will all work
fantastically to give drums the same amounts of power required. At this
point, I really have only very roughly balanced the various elements from
the vocals down to the bass, but the important thing is that I will have
looked at everything individually. I will regularly need to mute the vocals
to achieve some of the things I have described here, but I will put the
vocals back in as I move through each step. Now for the drums, I need to
focus mainly on bass, drums and guitars in the balance. It is tough to start
gating, compression and equalization accurately on drums when the vocals
are still in the mix and prominent. As with a ‘bottom-up mix’, I would
start with the kick but I usually bring the whole kit in roughly and quickly
under the current balance to remind myself of what the drums are doing.
In my chain of kick inserts, it would be gate first, then compression, then
equalization. I prefer to see inserts in that order, as I do not want the equal-
ization to affect my gating and compression settings. I would set up a side
chain if I felt that was required. Copying the kick track before you start
Top-Down Mixing—A 12-Step Mixing Program 71
the processing is a good idea, because the gating and equalization can
end up being quite drastic (especially on live drums) and you may wish to
balance in the unprocessed original track. The same can be done for the
snare track. Typical equalization on the kick would be a 2 dB to 6 dB cut
at 300 Hz to 900 Hz and a boost of 2 dB to 6 dB at 50 Hz; if more ‘slap’
or pedal-beater is required, then boost 2 dB to 6 dB at 3 kHz, but no higher
than that as we are saving 4 kHz and upwards for the guitars, keyboards,
strings, vocals and cymbals. Typical snare equalization would be a 2 dB to
6 dB cut at 300 Hz to 900 Hz, boosts of 2 dB to 6 dB at 4.5 kHz and 2 dB
to 6 dB at 7 kHz or 8 kHz. The kick and snare gating would aim to isolate
them from the rest of the drum kit, and typically you would hope to elim-
inate snare spill on the kick signal and hi-hat spill onto the snare signal,
always being careful to allow as much release and ‘hold time’ to allow the
drum decay to breathe before the gate closes. Compression on the kick
and snare should be minimal, as they will be likely to drive compression
on the overall mix at the end of the session. An individual compressor on
the kick should start at a 2:1 ratio for the same reason as mentioned on the
bass; higher compression ratios can have the effect of losing low frequen-
cies. A major part of the snare sound will be the choice of room ambience
effect, even something like the D-Verb in Pro Tools, set to medium room
and below one second in length, can give you a good room ambience for
drums and instantly the snare will be more powerful and sound as though it
is part of the track. I would also send the tom-toms to the same room ambi-
ence effect, then gate the tom-toms to eradicate spill from the rest of the
drum kit or, more efficiently, you could edit them to only remain where the
tom-toms play throughout the track. Any equalization on tom-toms would
be minimal, a moderate boost around 4.5 kHz is all I would do if they are
lacking attack or sounding too dull. Overhead cymbals and hi-hats should
all have similar equalization to bring out the cymbal sound and to eradicate
kick, snare and tom-tom spillage. My equalization for these would be an
HPF up to 200 Hz to 300 Hz, a cut of 4 dB to 6 dB at 300 Hz to 900 Hz,
a small boost at 5 kHz and 10 kHz. That concludes the drum processing.
An important decision is panning perspective, as was the case with the
piano. We need to decide on panning to the ‘player perspective’: hi-hat left,
tom-toms left to right, overheads left to right—or the ‘audience perspec-
tive’: hi-hat right, tom-toms right to left and overheads right to left. I much
prefer the player version on drums and piano. If I have some stereo room
microphones recorded for the drum session, I make sure that their panning
matches the close microphones and consider a similar equalization to the
overhead cymbals. I often experiment on the room microphones with some
heavier compression at a 6:1 ratio or more and an overdriven threshold
setting. You will need to raise the compression output or make up gain,
but this can sound very powerful. My personal microphone choices when
recording drums are
a. Kick—Electrovoice RE20
b. Snare—Shure SM57
c. Hi-hat—AKG C451
72 Phil Harding
d. Tom-toms—Beyer 101s
e. Overheads—AKG C451s
f. Room microphones—Neumann U87s
g. Ride cymbal AKG C451 on the bell of the ride cymbal and route that
to its own track to enable control of this during guitar solos and so on
Everything has now been individually processed and each step has dealt
with a group of overdubs or a single instrument at a time. After the drum
step, the main lead vocals should still be prominent in the balance when
I bring them back in. Now is the time to achieve the overall mix balance, and
if I have been listening on large studio monitors or quite loudly on nearfield
monitors—one often has to when processing individual sounds, I would
consider taking a break. I believe it our duty as creative technicians to make
sure that the overall mix is playable without ear damage at high volume. This
is because many listeners, who we hope are excited by the music we create,
want to listen loud. But at this stage of the mix you should come down to a
quiet or medium volume on your nearfield speakers to achieve the final ste-
reo balance. In an ideal world, I would approach step 11 on a fresh day—or
at least after an extended break. I find that these final tweaks can take up to
half a day, maybe checking on headphones, flipping between speakers and
volumes and possibly running a test mix wav file or CD to listen to in differ-
ent environments, such as your home hi-fi system or car stereo. Also, at this
final step, I would send the lead vocals and bass to the Roland Dimension
D, as mentioned earlier, to help the lead vocal stand out from the track; the
Dimension D has an effect of bringing the vocal forward. While we do not
want the bass to be as loud as the lead vocal, the Dimension D effect brings
it forward and gives a stereo perspective that is generally a nice touch.
Figure 4.1 Typical Maxim settings, inserted on the stereo Master Fader in Pro Tools
The idea with this method of mixing is that one is concentrated on the
song from the minute of starting the mix, and although I have described
the process from the vocals down, it would also work for instrumentals and
other non-vocal-orientated tracks where you would start from your lead
instrument or ‘theme’. Fauconnier and Turner (2002) talk about “concep-
tual blending” as an unconscious human activity, and that’s exactly what
I seem to do in my mixing approach. Short ambience on the drums are
blended with medium- to long-plate reverb on the vocals, then those vocal
reverbs are blended with my crotchet mono delay as highlighted in step 4.
These are good examples that describe the types of technology we engage
while mixing a record using today’s technology. Another common industry
74 Phil Harding
practice in 2016 that I have not mentioned is the dropbox file sharing sys-
tem. This is useful when either engaging others to perform a mix for you
or also if you have been hired for a mix. Exchanging multitrack files via
dropbox has become common practice, and I recommend signing up for an
account to enable participation in this.
Figure 4.2 Logic Pro: export ‘All Tracks as Audio Files’ menu choice recommendation
Figure 4.3 Pro Tools Mix Template edit window
Figure 4.4) and then import all of the audio tracks, also at bar 1 beat 1. The
final task before starting the 12-step mixing program is to position all of the
auxiliary ‘group VCAs’ to the right of the cluster of audio tracks you want
them to control. Route those audio tracks, via buses, to the corresponding
input buses of the VCA groups, for instance, all drums and percussion can
go to bus 11–12 (we have saved buses 1–10 for auxiliary effects sends). See
the Pro Tools mix template screenshots in Figures 4.3 and 4.4 to see how this
final process should look.
Now refer back to step 1.
Notes
1 Please note that throughout these steps I refer to the 300 Hz to 900 Hz frequencies as dis-
liked or loathed by myself. I have never found an occasion when it is useful to boost this
frequency range. I find it to be ‘muddy and boxy’ when boosted and will generally recom-
mend either cutting these frequencies or leaving them flat.
2 There is a technical appendix at the end of this chapter that describes my audio and MIDI
file export procedure from Logic Pro to Pro Tools. You may want to read that before
embarking on step 1.
Bibliography
Cauty, J. and Drummond, B. (1988). The Manual (How to Have a Number One the Easy Way).
London: KLF Publications/Ellipsis.
Fauconnier, G. and Turner, M. (2002). The Way We Think (Conceptual Blending and the
Mind’s Hidden Complexities). New York: Basic Books.
Harding, P. (2010). PWL from the Factory Floor. London: Cherry Red Books.
Massey, H. (2015). The Great British Recording Studios. San Francisco, CA: Hal Leonard
Books.
Moylan, W. (2015). Understanding and Crafting the Mix. Oxford: Focal Press.
Zagorski-Thomas, S. (2014). The Musicology of Record Production. Cambridge: Cambridge
University Press.
Discography
76
5
Introduction
77
78 Justin Paterson
Overlaps
While mixing per se has always been a profoundly creative act, mixing ITB
has amplified this potential and expanded it to overlap with other major
aspects of the creative process: composition, arrangement, pre-production,
editing, production and mastering. The DAW has presented the integrated
environment that allows these aspects to be incorporated (at least in part)
into the mixing stage, or indeed for the mix to start to coalesce before
its technical inception. Moorefield (2010) presented his argument for the
producer as composer, underpinned by considering both the bidirectional
interface of illusion and reality, as well as auteurism in music produc-
tion. While composition and arrangement could be guided from a DAW
Mixing in the Box 79
arrange page with the mix itself performed on hardware, it is the integra-
tion of mixing into the computer domain of ITB that facilitates hybrid-
ization of the creative process. “It’s not surprising that far from removing
‘sacred cow auteurs’, modern technology has simply shifted the metaphor
from exceptional accomplishment on paper by ‘composers’ to exceptional
accomplishment on hard disk by ‘producers’ ” (Moorefield, 2010: 111).
Although Moorefield refers to the producer in this context, mixing ITB
allows the metaphor to be extended from producer to mix engineer. Ever
more often, the two roles tend to overlap, and it is ITB that offers the
increasingly common scenario where the producer and the mix engineer
are one and the same. Assuming they are not, consider two workflows: the
producer might elect to print an effect that will then be subjected to the
mix process, or the mix engineer might apply an equivalent effect of his or
her own volition. While either mode is valid, the former could be said to
offer the ultimate auteurism to the producer, but of course it can also bring
its own disadvantages to bear. The production aesthetic becomes entirely
constrained by the subjectivity, stylistic preferences and limitations of the
individual; as such, auto-editorial control becomes an increasingly rele-
vant skill. The latter offers its own considerable advantages by bringing
the objectivity and specialist skills of a professional mix engineer to build
upon the creative foundation laid by a producer,4 but it is the mode of
producer/mixer working in the integrated environment of ITB that offers
a greater level of dynamism and the ultimate flexibility. For many (at all
levels), mixing ITB has also now facilitated a broadening from Moore-
field’s auteur to atelier—in which the neo-craftspeople of music produc-
tion might compete with the established orthodoxy or simply enthuse in
their newly found capabilities. The apogee of their aspiration is ever more
accessible.
Traditionally, the mix engineer would work with a number of relatively
fixed performances, optimizing these spatially, spectrally and dynami-
cally. Digital editing brought an enormous fourth dimension to mixing—
temporal adjustment. Preceding this and building on the earlier tape-based
work from musique concrète, John Cage (1952) introduced the notion of
(what is now often referred to as) hyper-editing5 in his piece, Williams
Mix, which featured hundreds of random tape edits of short fragments of
sound. Although aleatoric, in principle this deviation from the assumed
linearity of music in Western traditions was to have a profound impact on
future studio approaches. Such detailed editing might be termed ‘micro
editing’. On the other hand, ‘macro editing’ might refer to editing longer
sections of music. One example of this is how, in jazz, Rudy Van Gelder
pioneered tape splicing as a creative tool in the 1950s (Skea, 2015) to con-
catenate sections of ensemble performances into an idealized composite
rendition of a given tune.
The advent of non-linear recording and editing made both micro and
macro modes of adjustment more accessible. Micro editing came to be
used for phase-aligning multi-microphone recordings such as drum kits—
time-slipping the different channels in order that they most precisely
align with a key element such as the snare-drum-microphone recording.
80 Justin Paterson
John Webber (2013) suggests that if the intention is ultimately to make the
master very loud, then the mix engineer preempts this by monitoring with
a limiter on the mix bus, but removing it before printing. Webber recom-
mends that the limiting be applied only at the end of the actual mastering.
The Interface
One problem that can be encountered when mixing ITB is the lack of pro-
prioceptive control. When using a traditional hardware interface, the user
can typically operate a fader or knob without looking at it, using a combi-
nation of proprioception and aural feedback to exert control. When using a
mouse, trackpad or touchscreen, parametric control is much less intuitive,
and it can be difficult to control the rate of parametric change or even
remember the correct mode of interaction with a given plug-in (e.g., rotary
versus linear action)—each of these presents a momentary impediment to
the primary aural intention. Further, typically only one parameter can be
changed at a time,10 which is often less than optimal when, for instance,
applying EQ. Although a dedicated hardware DAW-control-surface (or to
some extent, one of the increasingly popular tablet-based remote-control
apps, which are of course multi-touch) can mitigate such awkwardness,
such systems can introduce further complications such as orientation when
using a subset of soft controls that frequently change assignment.
Of DAWs, Mycroft et al. (2015: 687) found that
and that
Reducing the need to navigate the interface to find visual information can sig-
nificantly improve the user’s abilities to hear concurrent audio changes to the
programme material.
This provides a logical rationale for the intuitive preference for multiple
displays when working ITB, and helpfully sets a path towards further con-
sideration of the influence of the interface upon workflow.
It is commonly understood that the multimodal layers of information
provided by the DAW extend the operator’s sonic perception beyond typ-
ical hardware metering (which is often purely gain based). This is exem-
plified, for example, by ITB representations of pitch and timing. There is,
however, a trade-off. Ihde (2013) discusses the broader phenomenology of
instrumentation, and while acknowledging that instruments might embody
human experience and therefore extend our senses, he emphasizes that
the use of such tools is non-neutral. The typical ITB interface attempts
to seduce the user with its feature-set and encourages certain behavior.
Of course, the hardware studio also does this, but its configuration and
82 Justin Paterson
There is a balance between a visualization that aids, reinforces and expands our
innate aural abilities, and one in which the instrumentation translates all other
aspects into visible results.
There might appear to be a useful trajectory forming for faster, more ergo-
nomic and intuitive mixing ITB in the future. The latest GUI additions
to Apple Logic Pro X11 appear to reflect such a conscious move towards
finger-sized simplicity, although more likely driven by the iPad ‘Logic
Figure 5.1 Top: the older ES2 Synthesizer from Apple Logic Pro; Bottom: the newer
Clav plug-in. Both images are to scale, revealing the difference in GUI style
Images captured from Apple Logic Pro X
Mixing in the Box 83
model. Although, again speaking more generally, Ihde (2013: 40) states,
“it is also possible to ‘objectify’ the instrument such that it is understood to
be animated, to have its own ‘ghost’, and thus be reified”. While this obser-
vation could be applied to hardware, software users are often (fiercely)
patriotic and frequently relish championing their preferred system to users
of others. It is perhaps symptomatic of an established familiarity, yet fra-
gility of mastery over a given highly complex system, and the naiveté to
the feature set of other apparently arcane systems that give rise to this par-
tisanship. As such, with regard to mixing ITB, Ihde’s (2013) objectifica-
tion could be considered beyond reification and towards deification. While
console users frequently sit at different models in different studios (‘latest’
plug-ins aside), it is rare for the ITB disciple to change deity.
Tools
Acquisition of new software is part of daily life in the world of music pro-
duction and mixing. It is often exciting to first launch a new tool and explore
its capabilities, very likely before consulting its manual. The naiveté that
one might hold when first approaching such a tool can lead to immediate
gratification, and often with the application of the tool to a greater extreme
than the experienced user might employ. Thus, it is a raw creative process
free from preconceptions, quite akin to Hans-Georg Gadamer’s (2004)
being “lost in play”,12 which is an approach to recognition and interpre-
tation, as related to musical improvisation and beyond, first suggested by
Ramshaw (2005). Once experience with the tool accrues, it is likely that
preferred modes of working will emerge, very often resulting in a rather
more subtle application. While such a pattern equally applies to more gen-
eral software processes, it is the frequency with which this situation arises
in the mixing ITB environment (particularly through plug-ins) that causes
it to exert more impact on this workflow.
Once such a tool proliferates and it is heard in a number of differ-
ent contexts with different interpretations, that initial excitement might
typically wane. Should the tool have a powerful unique identity, it may
become meme, perhaps even to the point of cliché. A good example
of this was the definitive early ‘extreme’ application of Antares Auto-
Tune in Cher’s (1998) Believe. Such an application falls into the cate-
gory of creative abuse (Keep, 2005), and it is notable that the producers
Mark Taylor and Brian Rawling published an exposé that originally
attempted to hide their new tool (Sillitoe, 1999) in order to maintain
the mystique of their technique. The application of Auto-Tune (and its
equivalents) is currently all but ubiquitous in pop music, both as an
effect and when applied more transparently, perhaps as the tool was
originally intended. Perhaps the example of Auto-Tune is one that is
simultaneously both meme and cliché.
As in the above example, in the mode of ITB, innovation is often driven
by technology, in that periodically it is emergent technologies that facili-
tate new modes of mixing13 and corresponding new sounds. Further, these
Mixing in the Box 85
Of course, the professional user might still find time for all of this, but
such a pattern might impact more on the amateur. Such a paradigm is not
just a function of the physicality of the hardware tool, since a parameter-light
vintage compressor will impact upon such situations in a different way
than would a complex Eventide harmonizer.
Perhaps it is the presence of the embedded software operating system
that makes the difference in this example. Such operating systems became
increasingly ubiquitous in the sampler/workstation era,15 vastly increas-
ing potential functionality, but also making its intended boundaries more
nebulous. The prevalence of user-installed software on a computer greatly
multiplied the complexities that were typically dealt with, multiplied again
by the transition from MIDI sequencer to DAW plus third-party systems,
and ever more time was spent reading manuals and solving problems just
to stay abreast of expected operation. As the palette of functionality con-
tinued to multiply exponentially, the depth of engagement with the tools
diminished further, a simple function of available time versus the need to
complete certain tasks. To some, this was amplified as both the intrinsic
and financial value of software decreased.
Regardless, the contemporary ITB engineer needs to deal with an
ever-larger toolset. While many neophiliacs might use this to pursue a
sonic avant-garde, pragmatism and focus must remain. As Skea (2015:
1) observed of Rudy Van Gelder, who worked long before the software
age, “the quality of Van Gelder’s output rests not necessarily on tech-
nical innovation but on determination to master successive waves of
state-of-the-art technology available to him and a legendary degree of
perfectionism”.
The use of preset parameter sets can make the array of tools much
more accessible, although often at the expense of precise function or
even appropriateness. Expert users might tend to shy away from presets
in equipment with which they are very comfortable, yet still employ them
in the interests of pragmatism with less familiar systems (Paterson, 2011).
The sheer number of such presets is increasingly difficult to navigate, yet
the range of their functionality can be increasingly flexible and exotic.
Manufacturers are starting to respond with more intelligent preset options
that simplify user engagement or standardize parameter sets across a
range of equipment. In part, this is an attempt to keep (less involved) users
abreast of the equipment, but it also introduces a danger of homogeneity
and ‘user irrelevance’ when working ITB. Naturally, there are still a great
many practitioners who actively pursue the novel, and there are ever-larger
opportunities to implement this, but as with Van Gelder, adaptability is the
key in the fast-paced fluxive arena.
mixMacros
of melody, harmony, and rhythm” (Cope, 2004: 109), over a small num-
ber of beats. Such signatures provide a useful metaphor for many mix
processes, but especially when mixing ITB. In the world of hardware,
Michael Brauer’s trademarked ‘Brauerize’ multi-bus compression tech-
nique (Tingen, 2008) is a notable example. Brauer assigns groups of
instruments to one of a number of buses, each of which features a chain
of compressors and equalizers. The console faders function pre-compres-
sor chain, and everything is mixed through one of the chains. Although
the specific configurations can be variable depending on the situation,
the effect of this process produces a replicable sonic signature, derived
from what are effectively templates. Each of these might here be coined
in a more general sense as a ‘mixMacro’.
While Brauer’s large collection of hardware facilitates this, it is
beyond the reach of many. In the software world, however, not just exten-
sive compressor chains, but mixMacros of automation curves, complex
signal routing configurations, groove templates, all manner of plug-in
combinations, and more, are easily configurable in sets, and importantly
replicable via copy/paste or embedding in DAW template sessions. The
deployment of mixMacros allows the ITB engineer to form a custom tool
kit from which to draw, to form signatures of sonic mediation. Counter
to such convenience, such an approach can create an interesting ten-
sion for those who wish to avoid ‘preset’ sounds in the name of bespoke
treatment or integrity. The nub is perhaps dependent on whether a man-
ufacturer’s effect chain is used in the mixMacro, or whether the engineer
develops a wholly bespoke and considered one, perhaps even over many
years, as was the case with Brauer. Either way, if the pre-configured mix-
Macro is consciously allowed to influence the artistic direction, it gains a
degree of autonomy, something that Zagorski-Thomas (2014) describes
as its residing within the actor-network theory (ANT) of social theory.
The mixMacro is an actor concatenating with Ihde’s (2013) non-neutral
instruments.
So, in mixing ITB, mixMacros are autonomous objects that might
be assembled in a near-infinite order, in an actually infinite number of
contexts (tunes), yet each always contributes its own unique ‘signature’.
This allows an extension to a further metaphor from visual arts, specifi-
cally the actual implementation of these components could be regarded
as parallel to collage or montage. As Cutler (2004: 145) notes, “with so
many precedents in the world of the visual arts . . . it does seem surpris-
ing that it took so long for there to be similar developments in the world
of music”.
Conversely, Theberge (1997: 206) states, “the artistic practices of col-
lage, assemblage and montage used in popular music virtually destroy the
organic integrity of ‘the work’ ”. He is speaking specifically of songs,16
and clearly does not hold that aspect of them in high regard. Of course,
the semantics of “organic integrity” might be analyzed; surely, such a
statement must relate to the singer-songwriter or the paper-based com-
poser. It is interesting that in the intervening years since that was written,
both modes of creator have increasingly turned to computer assistance to
88 Justin Paterson
Groove
Sidechain
template
Insert Chain
DAW Track or sub-group Audio
inc. EQ
Group, multi-bus
or other routing
Sends, inc.
Crossfades
parallel FX
Figure 5.2 A schematic of a mixMacro for one or a group of tracks, illustrating some
of the factors that define its ‘signature’
Mixing in the Box 89
Recall
One of the biggest advantages to mixing ITB is total recall, and as such,
it deserves specific mention here. Opening a DAW session re-creates the
exact state that it was in when saved, whereas almost any analog part of
the signal path would have to be manually restored. Manual restoration
can virtually never be totally exact (to phase accuracy). More importantly,
it is the time taken to re-patch the hardware and set all the parameters that
makes the difference. Mix engineer Mark ‘Spike’ Stent attests:
In addition, producers, record companies, and artists are used now to the fact
that they can call you, even two months after your mix, and request a change,
and you just bring up the Session and five or 10 minutes later the change is
made. So mixing in the box is about time and being flexible, and of course it
also saves on the budget.
(Tingen, 2010)
However, there are also a number of reasons why mixing ITB can be more
time consuming than legacy hardware approaches, at least during the pri-
mary mix. When performing a given operation such as compression, the
mixer must decide whether to opt for a favorite plug-in or perhaps try out a
number of options based on prior recommendations or curiosity. Typically
the number of plug-ins available to the mixer will be an order of magnitude
greater than those in a hardware studio, and many will offer unusual modes
of operation that require considerable experimentation and/or reading of
the manual. As previously mentioned, it is easy to become fixated on micro
editing, not just of audio, but also automation and numerical parameters—
such operations often require highly repetitive and multi-stage workflow.
Despite this, an increasing number of professional mix engineers are turn-
ing towards an ITB approach in the interests of pragmatism.
Future
The current paradigm of mixing ITB is almost entirely based around the
metaphor of the traditional mixing console. New modes of mixing are now
being proposed. There are innovative designs, such as that of Cartwright
et al. (2014), which presented a 2D map around which a cursor could be
moved to influence the EQ and gain settings of every channel simultane-
ously on a holistic like or dislike basis. Automated mixing is being devel-
oped as alluded to in chapter 15 by Joshua Reiss, and this is likely to
realize many procedural tasks for the professional, and perhaps one day
precipitate user irrelevance with regards to technical operations, allowing
the user to focus on artistic and strategic choices.
Butterfield’s (2002) model is yet to be fully adapted and specified to
define specific ITB approaches, but perhaps if combined with Moylan’s
(2007) mix descriptors, then a taxonomy of such approaches could be
90 Justin Paterson
formed for the digital era. This could be augmented with Zagorski-Thomas’s
(2014) sociological approaches to align with the emergent musicology of
record production.
It is inevitable that further new and surprising tool sets will evolve and
apply artificial intelligence to many of the processes, extending Moore-
field’s (2010: xiii) “reality of illusion”. Perhaps the trajectory of illusion
can be illustrated even more clearly if parallels to other art forms are
made. Whereas once, audiences were amazed by ghostly projections of
actors onto pieces of glass on the theatre stage, we currently have mark-
erless facial motion-tracking (Faceshift, 2015) in cinema and games, and
this technology is still in its nascency. In this context, the physical charac-
teristics of an actor are now irrelevant, and further, the very performance
can be retrospectively adjusted. Intel RealSense (2015) is poised to bring
similar 3D sensing and beyond to the consumer, with many developers
yet to identify potential markets. It takes little imagination to translate
this into an equivalent form of sonic control, as is already being demon-
strated by the augmented performance capabilities of Zoundio (2015),
where lesser-skilled performers can deliver virtuosic performances, yet
still control the artistic direction. A Guitar Hero (Activision, 2016) met-
aphor for real instruments might yet become the norm in the future, with
long-honed traditional dexterity as obsolete as touch typing in the current
voice-recognition era.
In addition to new 2D mixing-interface paradigms, 3D gesture spe-
cifically for musical control is already at hand with the Leap Motion
controller and the Geco MIDI app, and the emergence of augmented
reality for the DAW is likely to extend this, making the action of
mixing ITB ever more intuitive and powerful. At the time of writ-
ing, Bullock (2015) is developing algorithms for two-handed sonic
manipulation in free space, supported by a 3D real-time visualiza-
tion. Such approaches might develop into a new mode of HCI when
mixing ITB and be applied in numerous contexts. Retina tracking and
augmented-reality workflows are also likely to emerge commercially
in the near future, and all the time, brainwave-based control systems
become ever more feasible.
As component modelling develops to the point where software is ubiq-
uitously accepted as being genuinely capable of accurate emulation of
consoles and hardware outboard, workflow will be the most compelling
demarcation, although this too will evolve in parallel to the growing demo-
graphic that feels independent of a console/outboard-based workflow.
Mixing ITB will become ever more powerful, autonomous and indepen-
dent of its legacy, allowing more of us to reach the current zenith more
rapidly. If reaching an ideal takes less time, then reaching a number of
equivalent and musically interesting ideals in the same time might become
possible. The emergence of new dynamic or interactive playback formats
that require multiple mixes will be empowered by this. This mode might
go beyond the zenith, and for a given tune offer multiple new points on the
celestial sphere of music.
Mixing in the Box 91
Notes
1 It is particularly pertinent in a chapter on mixing in the box to emphasize that the content
of a ‘track’ might be either recorded audio or a MIDI sequence.
2 For example, recorded to ½" magnetic tape, digital audio tape (DAT) or CD. Such mixes
were often recorded, and artists would decide which they preferred for the final release by
retrospectively listening to each.
3 This text will not concern itself with room acoustics, which of course influence the size
and shape of any mixing environment.
4 Who will generally create a rough mix that includes printed effects for guidance.
5 A term derived from intensive MIDI editing in the 1990s, employed by artists such as
Aphex Twin.
6 Sometimes, this alignment was done with a bespoke hand clap before the music com-
menced. It is perhaps interesting to speculate that if the drummer were to clap in a natu-
ral position, one foot above the snare, then the synchronization would be susceptible to
approximately 2 ms error; 1 ms farther from the (assumed close) snare drum microphone
and 1 ms closer to the overhead microphones.
7 Of course, this is actually the primary function of quantization.
8 For instance, with compression attack and hi-mid EQ.
9 As discussed, for instance by Vickers (2010).
10 Although this is currently being challenged by emergent multi-touch systems such as the
Slate Raven (Slate Pro Audio, n.d.).
11 Logic Pro has long sported screen-space efficient GUIs that packed many small fields into
a limited area, e.g., the ES2 synthesizer.
12 First published in 1960.
13 As well as, of course, the greater field of music production.
14 Cher, most likely, since they will not have heard of a niche company. Perhaps the scholar
might prefer that the debate on such a musical gesture be between Homer Dudley and
Sparky the Magic Piano.
15 Which commenced with the launch of the Fairlight CMI in 1979.
16 It would be unfair to presume his holding any objection to mixing as collage, but it serves
as a useful foil for this metaphor.
17 It must be so, given that the DAW tool set is ubiquitous across genres.
Bibliography
Activision. (2016). Guitar Hero Live The Game | Official Site of Guitar Hero. Activision.com.
Available at: https://www.guitarhero.com/uk/en/game [Accessed: 18 April 2016].
Bullock, J.B. (2015). Transforming Transformation: 3D Models for Interactive Sound Design.
Gateway to Research. Available at: http://gtr.rcuk.ac.uk/project/23DDE2F7-3184-4D94-
97C6-85B50B09602C [Accessed: 6 January 2016].
Butterfield, M. (2002). The Musical Object Revisited. Music Analysis 21 (3): 327–380. Avail-
able at: http://www.jstor.org/stable/3840795 [Accessed: 14 March 2013].
Cage, J. (1952). Cage: The 25-Year Retrospective Concert of the Music of John Cage by
Various Artists. iTunes. Available at: https://itunes.apple.com/gb/album/cage-25-year-ret
rospective/id521720437 [Accessed: 24 June 2014].
Cartwright, M., Pardo, B. and Reiss, J.D. (2014). ‘MIXPLORATION: Rethinking the Audio
Mixer Interface.’ In 19th International Conference on Intelligent User Interfaces, Haifa,
Israel, 365–370.
Cher. (1998). Believe. Available at: https://itunes.apple.com/gb/album/believe/id73273758
[Accessed: 20 October 2015].
Clifford, A. and Reiss, J. (2010). ‘Calculating Time Delays of Multiple Active Sources in Live
Sound.’ Audio Engineering Society. Available at: http://www.aes.org/e-lib/online/browse.
cfm?elib=15580 [Accessed: 7 December 2015].
92 Justin Paterson
Cope, D. (2004). Virtual Music: Computer Synthesis of Musical Style. Cambridge, MA: MIT
Press.
Cutler, C. (2004). ‘Plunderphonia.’ In Audio Culture: Readings in Modern Music, eds. C.
Cox and D. Warner. New York City: Continuum International Publishing Group Ltd.,
pp. 138–156.
Faceshift. (2015). Faceshift. Available at: http://www.faceshift.com/ [Accessed: 14 July 2015].
Gadamer. H.-G. (2004). Truth and Method, New edition. London; New York: Continuum.
Ihde, D. (2013). Technics and Praxis: A Philosophy of Technology: Volume 130, 1979 edition.
New York: Springer.
Intel. (2015). Intel® RealSenseTM Technology. Intel. Available at: http://www.intel.com/
content/www/uk/en/architecture-and-technology/realsense-overview.html [Accessed: 15
April 2016].
Johnson, S.A. (1999). Interface Culture: How New Technology Transforms the Way We Create
and Communicate, New edition. New York: Basic Books.
Keep, A. (2005). ‘Does Creative Abuse Drive Developments in Record Production?’ In First
Conference of the Art of Record Production, London, UK. Available at: http://www.artof
recordproduction.com/index.php/arp-conferences/arp-2005/17-arp-conference-archive/
arp-2005/72-keep-2005 [Accessed: 7 May 2014].
Moorefield, V. (2010). The Producer as Composer, First edition. Cambridge, MA: MIT Press.
Moylan, W. (2007). Understanding and Crafting the Mix: The Art of Recording, Second edi-
tion. Burlington, MA: Focal Press.
Mycroft, J. and Paterson, J.L. (2011). ‘Activity Flow in Music Equalization: The Cognitive and
Creative Implications of Interface Design.’ In Proceedings of the 130th Audio Engineer-
ing Society Convention. Available at: http://www.aes.org/e-lib/browse.cfm?elib=16568
[Accessed: 19 March 2013].
Mycroft, J., Reiss, J.D. and Stockman, T. (2013). ‘The Influence of Graphical User Interface
Design on Critical Listening Skills.’ In Proceedings of the Sound and Music Computing,
Stockholm, Sweden, 146–151.
Mycroft, J., Stockman, T. and Reiss, J.D. (2015). ‘Audio Mixing Displays: The Influence of
Overviews on Information Search and Critical Listening.’ In Proceeding of the 11th Inter-
national Symposium on CMMR, Plymouth, UK, 682–688. Available at: http://cmr.soc.
plymouth.ac.uk/cmmr2015/downloads.html [Accessed: 26 November 2015].
Orban, R. (1979). FM Broadcast Quality. Stereo Review.
Paterson, J.L. (2007). Phase Experiments in Multi-Microphone Recordings: A Practical
Exploration. Journal on the Art of Record Production 1. Available at: http://arpjournal.
com/1803/phase-experiments-in-multi-microphone-recordings-a-practical-exploration/
(Accessed: 19 March 2013).
Paterson, J.L. (2011). ‘The Preset Is Dead; Long Live the Preset.’ In Proceedings of the 130th
Audio Engineering Society Convention, London, UK. Available at: http://www.aes.org/e-
lib/browse.cfm?elib=16569 [Accessed: 19 March 2013].
Phillips, M. (2012). Interview with Martyn Phillips.
Ramshaw, P. (2005). ‘Lost in Music: Understanding the Hermeneutic Process in Cre-
ative Musical Composition.’ In Vilnius, Lithuania. Available at: https://www.academia.
edu/172365/Lost_in_Music_understanding_the_hermeneutic_process_in_creative_musi-
cal_composition [Accessed: 5 January 2016].
Rothwell, R. (1994). Towards the Fifth-Generation Innovation Process. International
Marketing Review 11 (1): 7–31. Available at: http://www.emeraldinsight.com/doi/
abs/10.1108/02651339410057491 [Accessed: 7 December 2015].
Sillitoe, S. (1999). Recording Cher’s ‘Believe’. Available at: https://www.soundonsound.com/
sos/feb99/articles/tracks661.htm [Accessed: 20 October 2015].
Skea, D. (2015). Rudy Van Gelder in Hackensack: Defining the Jazz Sound in the 1950s |
Jazz Studies Online. Jazz Studies Online. Journal. Available at: http://webcache.google
usercontent.com/search?q=cache:xWTfR-HmMo4J:jazzstudiesonline.org/resource/rudy-
van-gelder-hackensack-defining-jazz-sound-1950s+&cd=1&hl=en&ct=clnk&gl=uk
[Accessed: 8 July 2015].
Slate Pro Audio. (n.d.). RAVEN MTX: Slate Pro Audio. Available at: http://www.slateproau
dio.com/products/raven-mtx/ [Accessed: 13 December 2013].
92
Mixing in the Box 93
Audio editing has played an important role in record production since razor
blade first touched tape, as it were. That said, modern computer-based modal-
ities comprise an entirely new and crucial genus of that aesthetic species. So
crucial has this new genus become to record production in general, in fact, that
few working recordists would blink if I suggested to them that modern record
production is defined or characterized almost entirely by it. Indeed, production
styles are no longer marked by the amount of audio editing they encompass, as
they once were, simply because every style of professional music production
now entails the same high degree of editing. Those who know what to listen
for can hear audio editing permeate every musical nook-and-cranny of mod-
ern record production, regardless of genre. And yet, despite the tremendous
growth in research on record production in the last few decades, very little
research focuses directly on this crucial new musical competency.
What follows is meant to address two of the lacunae I identify above. It
provides a broad description of audio editing as a musical competency in
and of itself, and it elucidates its position within—and with regards to—
modern record production at large. An overview of common editing techniques
is then provided, and musical examples supplement the discussion to help
readers hear those techniques in action. These audio examples can be down-
loaded at http://ww.hepworth-hodgson.com, and it is strongly recommended
that readers audition each example precisely where indicated in the text.
The manner of explanation modeled below would best be considered
a methodological hybrid, mixing ethnographic interview techniques with
traditional musicological analysis. It is my firm contention that, to truly
understand modern professional production practices, a methodological
broadening is required that allows practitioners themselves to speak in
academic contexts. And I believe the hybrid below models one among
many possible such ‘broadenings’. I chose a series of questions, in a par-
ticular sequence that I felt addressed the lacunae comprising the primary
subject of this chapter. And I sought the expertise of Alastair Sims, one
of Canada’s most successful audio editing engineers currently at work,
to provide those answers.1 Sims and I spoke for a few hours, a transcript
of our conversation was made and edited, and then Sims was given the
transcript to approve and edit further however he saw fit. I identified areas
94
Audio Editing In/and Mixing 95
where I thought listening examples might help concretize some of the con-
cepts and techniques Sims discussed, and Sims provided them in turn. The
Download and Listen tracks are listed later in the chapter and are available
on www.routledge.com/9781138218734 and www.hodgsonhepworth.com.
It is ultimately our hope that, in doing all this, we have provided analysts
with a useful toolbox that will help them hear audio editing permeate the
modern recorded soundscape completely. In turn, the musical role that
audio editing plays in modern record production should clearly emerge to
the analytic fore.
it would only take me about thirty minutes to edit. But back then it took
me about three days! I was nervous and so I questioned everything I did.
I’d think I was finished, and the other editors would come and listen to my
work, and find more and more mistakes. It was brutal!
silence left over between where the cut was made and where the note was
moved to (see Figure 6.2), as well as clicks at the cut points (see Figure 6.3).
Normally, when working in a clip/region-based DAW, you pull the
beginning of the second clip back to meet the end of the first (see Fig-
ure 6.4), and place a crossfade there, roughly five to eight milliseconds’
duration, or more, as a starting point (see Figures 6.5 and 6.6). The shorter
the crossfade the better, as a general rule. Occasionally, this process will
cause some audio to repeat itself (see Figure 6.7) which can be addressed
by adjusting the length of the crossfade (see Figure 6.8). Sometimes,
though, the fade needs to happen earlier into the note.
Failing these options, you might see if the same note is played some-
where else in the song, and use it in place of the problem note. You might
also use time compression or expansion to smooth edits, digitally length-
ening a note to fill the gap. You might say there are three primary ways to
approach editing timing. Each one of these approaches—(i) using fades,
(ii) replacing notes and (iii) time compression and expansion—are essen-
tial techniques or tools for us and can be applied whether you are quantiz-
ing to a grid or nudging notes manually.
Pitch correction and tuning is another aspect of editing, usually a little
more automated than timing. You would generally use a plug-in (Melodyne,
Auto-Tune) to achieve the edited result. Just like timing, pitch correction
can be applied to the whole track or simply a few off notes depending on
the desired aesthetic for the song.
of the recording at large in mind, after all. It’s not like I change pitches
around on performers to what I like instead, or needlessly quantize an
already amazing piece of music or take of a song. Editing engineers do
their very best to ‘serve the performance’, the same way producers do their
very best to ‘serve the song’.
Song > Player > Instrument > Microphone > Equipment >
Processing (Editing)
Figure 6.9 Order of importance
first chorus, the guitar player was really laying into the guitar, therefore
hitting the amp harder giving a more distorted, edgy sound. In the second
chorus, though, he was relaxing and playing smoother, perhaps improving
the timing and tuning of the chorus but taking away the energy. Now take
the edited (timed and tuned) first chorus and paste it to the other choruses
so they maintain the tone and energy of the first and you’ve just edited the
tone or timbre of a track.
The first, a filled gap in the audio, is the most common. When a piece
of audio is played out of time, cutting at the beginning of the note or the
transient of a hit and quantizing those pieces of audio is the most common
practice. As a consequence of the quantization process, though, two pieces
of audio might be moved apart from one another, leading to a gap in the
audio. The simplest way to remedy this gap is to pull back the beginning
of the second note or hit until the gap is filled. Because you are pulling out
108 Alastair Sims with Jay Hodgson
the beginning of the second note, you will be repeating a section that fills
the gap twice. Once in the first note and again in the part of the second clip
used to fill the gap. This will cause an audible effect of
doubling or stuttering that part of the audio file.
Download and Listen
To solve this repeat artifact, you simply extend
to tracks 6.18, 6.19
and 6.20
the length of the fade back into the first note
(extending it towards the second note would
mean the transient or beginning of the second note
would be repeated, which only increases how audible the edit
is), blending the two notes together making it one smooth, longer note.
AUDIO EXAMPLES Electric guitar stutter, filled and crossfaded with fades adjusted
Audio Editing In/and Mixing 109
The second type of ‘fade artifact’ you can hear is a fade that is either
in phase or out of phase. Imagine again two notes that have been cut
and quantized and they move very little or exactly the length of half
or one of the wavelengths of the note being played. In the instance that
the two are moved only a very small amount when you make a fade,
particularly a longer fade, you will hear comb filtering. The best way
to avoid comb filtering is by making the fade as short as possible. If
the audio clips are moved over by either one full wavelength or half a
wavelength of the note played, you will have a fade that is either per-
fectly in phase (one wavelength), or out of phase (half a wavelength).
This will lead to a quick volume increase in the case of in-phase audio
or a volume decrease in the case of out-of-phase audio. To avoid these
changes in volume, you need to change the type of fade you’re using.
There are two basic shapes of fades, equal gain and equal
power. Equal gain is a linear fade, while equal
power is logarithmic. To avoid a volume
increase in the case of in-phase audio, you Download and Listen
would use an equal gain fade, and to avoid to tracks 6.8 through 6.14
a volume decrease you would use an equal
power fade. While the chances are small that
the two audio files will phase match perfectly in an
additive or subtractive way, it will happen from time to time. I suggest
starting always with an equal power fade that is short, around five to ten
milliseconds, and changing them as needed.
The third type of ‘fade artifact’ is when the sustain of a note is shortened
because you’ve moved two notes closer together. Imagine a piano with the
sustain pedal held down, allowing for notes to ring out under one another. If we
were to then cut up passages played on this piano and move them closer together
then fade them, you would have a quick decrease in
volume between the sustain of the first note to
Download and Listen
the now-moved forward second note, which has
to tracks 6.15, 6.16
a fixed amount of sustain bleeding over from and 6.17
the first note. This causes an interesting effect
on the source being edited; it makes for a disjointed
and jumpy sound. To fix this type of edit you adjust the length
of fade, making the fade longer, even to the point of it being most of the length
of the first note—the jump between different sustain levels is smoothed out
across the fade, making for a natural-sounding note transition.
While editing, these three types of fade happen all over the place and all
together. You might have an out-of-phase fade that has been pushed closer
together, so you’re also hearing a jump in sustain volume. This is just what
happens. You have to spend time adjusting the size, shape and location of
the fade. If that doesn’t work, find the note later in the take; if the note is
only played once, you can try adjusting the length of the note using time
expansion or compression tools, or if the problem is really that bad even
record another pass of the song.
Note
1 Sims has worked as an editor on recent releases from Rush, The Tragically Hip, Barenaked
Ladies, Three Days Grace, and many more household names.
7
Pre-Production in Mixing
Mixing in Pre-Production
Dylan Lauzon
Delineating roles and rules in music creation isn’t something I often run
into in the field, working professionally in the music industry. While cer-
tain tasks are saved for certain phases of the recording process, there is
rarely as wide a gulf between phases like production and pre-production
as is often portrayed in music academia. Jay Hodgson discusses it further:
From a practical perspective, such divisions will always be artificial. Each time
recordists select a particular microphone to record a particular sound source,
for instance, they filter the frequency content of that sound source in particular
ways; in so doing, they equalize and mix their records, even at this very early
stage. Recording practice is an entirely holistic procedure, after all. Tracking,
signal processions, mixing and mastering cannot be separated—not in practice,
at least. They are easily excised in theory, though, because each procedure is
tailored to produce a different result. During the tracking phase, for example,
recordists capture raw audio singles which they later massage into final form
using a variety of signal processing, mixing and mastering techniques. During
the signal processing phase, recordists filter and refine the “raw audio signals”
they collect during tracking; and, moreover, we will see that many kinds of
signal processing are done during tracking. Mixing is done to spatially orga-
nize the component tracks in a multi-track production into well-proportioned
shapes, and during mastering recordists apply a finishing layer of audio to var-
nish their mixes, to ensure they sound at their optimal best on a variety of
playback machines and in a variety of different formats.
Mixing isn’t a process that begins entirely in the formal mixing phase
of the recording process. In practice, mixing begins in pre-production—
when the first note is recorded. To reiterate Hodgson, every decision
throughout the creation process has a profound effect on the mixing pro-
cess. To further explore this phenomenon, I intend to thoroughly explore
my own personal pre-production process and attempt to describe how
mix decisions are occurring throughout. Further compounding the con-
fusion is modern recording technology. Recordists at all levels now have
access to sample libraries recorded in rooms like Abbey Road and digi-
tal emulations of synthesizers and guitar amplifiers that were historically
114
Pre-Production in Mixing 115
What Is Pre-Production?
The Process
i) Idea Generation
It’s beyond the scope of this chapter to discuss the creative process and
how ideas are generated from scratch, but we can discuss how best to cap-
ture this stage. Everyone has their own method, but at this stage simple and
transparent is better. It’s about capturing the content, rather than capturing
Pre-Production in Mixing 117
tones and creating productions. The simplest in most cases is to use the
dictaphone built into most modern smart phones. The purpose here is to
capture the idea for later use in the pre-production studio.
I pick presets, and I really make sure not to over analyze or second guess
myself. I often just pick loops that sound good in the context of things, and
move on not changing anything. It’s more about momentum than it is about
having the perfect or ideal part.
https://www.youtube.com/watch?v=2Bxzr9n_HK8
The insight here is that the focal point needs to be on the development of
the song rather than the tones or actual part played often. Mix issues can be
linked to structural and musical problems within the song as often as they
can be linked to actual technical and engineering issues.
The common next step is to develop the bulk of the musical elements of
a track; this often includes bass, guitar, synthesizers, banjo, kazoo. Really,
anything other than vocals and drums. As always, a huge amount of vari-
ety of processes exists at this part of the pre-production process. As in the
session set-up step, it’s very important here to capture a tone that’s easily
reproducible and can be used as an element in a mix. That’s not to say to
shy away from unique or interesting tones, but rather make sure to docu-
ment your processes, so that decisions can be recreated.
As an aside, this is a good moment to begin developing a collection of
production moves if it’s something you don’t have as a producer. A collec-
tion of presets can really facilitate flow and momentum in songwriting and
idea creation. That is, have sounds prepared ahead of time for particular
scenarios; a Mellotron pad patch for dramatic moments, a clean, dirty and
distorted guitar tone, a solid funky bass tone.
It’s important to make mix decisions as these ideas are developed. Com-
mitting equalization curves and compression settings not only allow one
to begin creating mix-ready tones early in the process, it sits elements
in their proper places, encouraging the inclusion of additional elements
that wouldn’t become relevant until the production and mixing phases of
recording.
While tracking in pre-production, it’s also important to track to the
same extent one would in studio. Specifically, things like guitar doubling
are important. Mix decisions based on a mono guitar will be very different
from those based on a wide and full stereo guitar, as will mix decisions
based on a simple single performance versus a powerful quad-tracked
heavy performance. Sometimes, underdeveloped elements can lead to an
overpopulation of elements that otherwise don’t need to be in a mix.
In the video linked above, you see Devin Townsend taking these steps
in creating his own production. Each step taken is with the final mix in
mind. As he captures even the least reliable instruments, he’s always con-
sidering the mix. For example, during the video he is seen recording an
acoustic guitar that has less than desirable intonation. Though the guitar
Pre-Production in Mixing 119
v) Vocal Capture
The message here isn’t to obtain the best equipment possible to capture the
vocal, but rather to make decisions. High pass where necessary, compress
and equalize to balance with the tracks captured in previous stages. The
tracking process shouldn’t be unlike one that would be taken in studio:
capture it part by part or line by line if necessary, and create a near-perfect
composite of all the takes. It’s even viable to automate the vocal as one
would during the mix stage. A vocal that is too full and too loud can make
necessary backing vocals and harmonies less obvious, while a mix-ready
vocal necessitates backing vocals at key moments.
Following lead vocal, backing vocals are fully fleshed out, almost beyond
true necessity at mix level. While I’ll dig into this more deeply in the next
phase of the process, the excess harmonies will be muted in the editing
portion of the process. The simplest method is to do blocks of harmony for
120 Dylan Lauzon
each part—thirds, perfect fifths, low and high octaves. This gives clarity as
to what more interesting and complex harmonies are necessary.
The goals of the studio should be to capture tones that are impossible in
a project’s setting stage, i.e., instrumentation that require immense input
lists like drums, or perhaps a more exciting vocal tone using rare and vin-
tage microphones. Even choirs, strings or horns if they are right for the
song are a great thing to capture in a commercial studio.
As outlined in the introduction, this process truly does gray the line
between traditional notions of production and pre-production. In many
cases the writing process never stops, even when the pre-production pro-
cess is arguably over and the song has entered the studio. Modernization
and availability of recording equipment has truly blurred the lines between
all roles within the recording process. It’s as difficult to pin down where
engineering begins and production ends as it is to pin down where pre-
production ends and mixing begins.
As you step into a studio or a mix environment following your pre-
production sessions, you really do come to realize how valuable it can be
to make conscious mix decisions in pre-production. Lacking components
become much more apparent, while conversely overproduction is easily
visible, saving valuable resources and, more importantly, time and creative
energy.
Academically, this needs to be explored further. I believe that it is
apparent that pre-production is, today, an integral part of mixing as a
whole. To put it simply, one needs material to mix in the first place, and
pre-production provides some of that material, thus informing the mix as
a whole. Even though pre-production processes are individual and varied
in nature, they will always affect the mix in that sense. In closing, I hope
that outlining my process has successfully illuminated my connections
that stretch between pre-production and mixing.
Bibliography
Hodgson, Jay. (2010). Understanding Records: A Field Guide to Recording Practice. New
York: Bloomsbury.
8
The recording community of today has the most superior equipment, stu-
dios and processing power. Almost everyone now has access to pieces that
were once only accessible to those who were lucky and talented enough to
work in the great rooms around the world such as Abbey Road, Olympic,
United Western or Sunset Sound, to name just a few. Anyone can load a
$20k Fairchild 660 to every single channel of the DAW. While I believe the
digital emulations are somewhat ‘apples and oranges’ compared to their
analog counterparts, I do somehow prefer the consistency and reliability
of digital plugins.
So, I ask myself, are we then creating, recording and mixing history’s
most epic records? We should be, and I think we are for the most part!
However, I believe other mixers can become sidetracked by the latest piece
of gear, plugin or update. They sometimes appear to be some holy carrot
hung before us. It’s so tempting to endlessly research all the facts and
figures, including the gear used on so many great records, and for me the
result is that I’m buried in a referential foundation. I too am a victim of
these distractions and they do frustrate me. I believe in a healthy embrace
of the past but with a strong ear towards the future. Depending on the proj-
ect, I value a blend of past and present. I’d rather be forging into the future
and not wasting energy.
There is no question about it, we have a bounty of resources, and
I believe in knowing recording history. However, that only gets one a
hair closer to completing a mix from beginning to end. Once we place
ourselves between the speakers, the ultimate goal is to make the speak-
ers move and elicit an emotional response. That’s where I want to be,
‘between speakers’. Ideally, the mix should be a journey that wraps
around each ear, through the eardrum to the inner ear and thus setting
off emotional neurons. I’d like to share a few tips to make this journey
from speaker to ear happen with speed and efficiency, and find the ceil-
ing while embracing the liberation of limitation in the areas of tones and
textures. I’ll share some of my mixing staples. I’ll explain my fondness
for the woolliness of my blue sweater, i.e., the classic Neve with loads of
iron-wrapped transformers, and how I establish a bold creative point of
view while serving the song and production.
122
Between the Speakers 123
Each of the aspects that will be examined include both the technical and
artistic side of mixing. Additionally, while keeping the practical in mind,
areas such as time and budgetary constraints in the realm of prepping,
working from presets and templates from the beginning to finishing a mix
will also be considered. The digital world provides the ability to work with
amazing speed and efficiency, thus allowing for stellar records to be made
for a budget that would have been considered a demo budget twenty years
ago. I began in the industry at the end of the 1990s/2000, during what
I would say was a sea change of the dominant analog recordings to the
digital world, with Pro Tools being the leader in DAWs. This allowed me
a great vantage point of the two worlds of analog and digital. There are
many positive workflow and efficiency aspects to DAWs, but I do encoun-
ter how easy it is to overdo it due to the unlimited ability of recalls, which
can jeopardize the end result by overprocessing sonically. I’ve learned the
beauty of chance and happy accidents provided by the analog world when
mixing. I have found further delaying decisions can be counterproductive.
The mixer oftentimes can wait for the next day and the mood has changed,
and a vicious cycle is entered of unlimited recalls and a never-ending mix.
The way I’ve learned to circumvent infinite possibilities is with limita-
tions. Limitations for me are a way to direct and harness my full potential
of creativity. In the coming pages, I will explore the virtues of narrowing
down the playing field in the area of workflow, which includes session
preparation of routing and layout, to the beginning and end of the mix.
While I’m not suggesting that new plugins shouldn’t be tried, I do find
boundaries are a helpful place to start. Just because we can eat food from
across the globe doesn’t mean we can throw together a mixture of ingre-
dients on a plate with a hodgepodge of spices. As a mixer, you’re trying
to obtain a balance of flavors. My spice cabinet is a blend of tubes, trans-
formers and transistors. A way to manage these choices and to reach for
the appropriate sound is to know to which group the EQ or compressor
belongs. As I stated, there is no shortage of gear and equipment choices
today, which means that knowing how and where to start can be daunting.
As with cooking, you become handy with certain methods, for example,
sautéing, frying and developing a competency and dexterity with certain
knives: paring, boning and chef’s knives. When mixing, we can use a com-
pressor for controlling dynamics and elevating and embellishing the feel,
but we can also establish the mood based on the type of compressor: tube,
transformer or transistor. Having these go-to staples creates a familiar
landscape for you to work within. These are the pigments of your palette,
your familiar tones, and the compass to use throughout your mix.
Before I start with pre-mixing details, let me give my thoughts on what
I think of the role of a mixer. First, I definitely never elevate my role as a
mixer above the producer, band or, most importantly, the song. My job is to
serve the song and the production. I believe I’m hired for the outsider per-
spective and as a second set of ears. I believe it can become even more of a
challenge if you also recorded the record. Sometimes it becomes apparent
that you can’t see the forest through the trees. The goal is taking what’s
there and highlighting and magnifying all of the beautiful moments. At
124 Dean Nelson
times, it can be the last 25% which may be felt and not heard. That’s what
I believe mixing to be in a nutshell.
I remember, when I was just starting out, how intrigued I was by how
people made their way into the mixing bowl, so I share my brief route.
My path came by way of a recording school called the Ontario Institute
of Audio Recording Technology in London, Ontario, Canada. After grad-
uation, I got an offer via a past grad at the school to be a tech at a studio
in Burbank, California. After a week-long cross-country drive from North
Carolina, I met with the studio owner and we quickly surmised that the
tech position was not for me. I was terrible at soldering. However, the
current assistant was looking to move on, which meant that there was
the chance at assisting, but I’d have to start out as a runner, learn the room
and get thrown into a couple sessions to see if I sank or swam. I definitely
swam, although the gracefulness can be questioned. The studio was Ocean
Recording, and when I started in June 2001, they had recently installed
and refurbished two old 80 series Neves. They essentially wired the two
together for a higher channel count. But, while the place was amazing for
tracking, there was no automation, which meant not too much mixing.
They did eventually install Flying Faders sometime after I left. The great
thing about starting off in a tracking studio with a healthy amount of clas-
sic mics, outboard gear and vintage Neves is that it allowed me to have a
strong reference point of understanding the most sonically bare state that
a record could be. Enter ear training 101. I think the biggest instruments
I got to know were the drums in their raw state. The industry was starting
to feel the early effects of file sharing and Napster, thus the loss of record
sales. In turn, the budgets were being cut, which meant the majority of
bands were coming in to track drums or beds and then moved on to cheaper
studios for overdubs. So, I got to hear some of the finest drummers and kits
(vintage and modern) in Los Angeles, as well as top-notch engineers and
their various approaches to miking. This was amazing ear training for me,
as drums typically play a large part in the mix foundation. After a couple
years, I felt I had a pretty good grasp on the foundational side of tracking,
for example, the placement of the musicians in the room and mic and com-
pressor choices, as well as basic EQ’ing approaches. After bands finished
up their sessions, I would often visit them when they moved onto other
studios for mixing. What I heard at the mixing point seemed miles beyond
the tracking stage. In the early 2000s, mixing in the box was still in its
infancy, mainly due to processing power and, more specific to Pro Tools,
the lack of delay compensation. Mixing as you go had not yet reached its
stride. I decided to move on from Ocean and ended up at a studio called
Chalice, where there was a decent amount of mixing going on. The cool
thing about Chalice was the mixture of various genres of music from rock,
urban, RnB and hip hop. It was great to see the different approaches based
on genre. I think the reason they catered to the variety was that they had an
SSL 9000J and a Neve 88R. My time there was a bit short-lived, as Chal-
ice was just getting going, and when the bookings slowed down around
Christmas, the owner freaked out and fired most of the staff. Eventually,
I ended up at Ocean Way, which has now been changed back to the original
Between the Speakers 125
It’s no secret that there is an inherent conflict between art and commerce,
and it is within that conflict that a challenging divide in record making
falls. A mixer’s goal is to excel and execute all that is needed creatively,
yet the end goal is also to produce an album that sells. Anyone can trim
extraneous time spent on a mix so that there will be money saved for you
and the band. One area that can burn time away unnecessarily is getting
started. For example, when you receive a session and can’t start imme-
diately, it can be a creative damper. In my time with Jack, I learned the
importance of sitting down and going from zero to sixty mph in a couple of
seconds. In other words, hitting the spacebar and unmuting and bypassing
126 Dean Nelson
some channels and plug-ins to see the tonal and sonic possibilities of the
mix come to life. This ability comes from one of the benefits of the DAW
world. Starting a mix again (prep work of routing, choosing plugins, figur-
ing out naming, etc.) can take from thirty minutes to a couple hours. That
extra time, however, adds up over the course of a couple weeks, months or
years. And one must ask, “who’s paying for that and how it will it affect
the ultimate course of the record?” Therefore, I offer some insight on how
to expedite and streamline this process to allow you to dive into what you
really want to be doing, which is the creative side of mixing.
During the first hour of any project, I like to find my bearings by starting
to address points noted: the main areas of timing and tuning to the spe-
cifics of tones along with any bad recording anomalies. These undesired
artifacts can be clipping, distortion, edits without fades or improper fades.
If there is an excess of timing or tuning issues, I’ll get in touch with the
band/producer. If it’s a small thing, I’ll make the edit and try to accommo-
date and work around the issue. It’s ideal for a mixer to receive sessions
in which all tracks are cleaned up with completed edits, fades applied and
consolidated clips, and that playlists are cleaned out unless there is need
for an alternate take. Markers should be located in the correct spot, and all
tracks should be labeled with the name of the instrument. The comments
section is a great spot for information such as the mic, pre and part that the
instrument is used in. I can’t stress that enough. Somebody’s name is not
an instrument, for example. Lastly, there should not be any unused audio
tracks, and if any virtual instrument tracks have been used those should
be printed as audio tracks. Sessions sent having the playlist and clip list
cleaned out helps with file size in that you are not sending any unused
parts, which allows for faster uploading and downloading time. Plus, it
allows for lower chances of confusion in regards to any unnecessary clips
or tracks which may be questioned for the intended use in the mix.
Other important considerations are needed when beginning a mix. If you
have the luxury, decide whether you will mix completely in the box or use
a console/summing mixer. My preference is a hybrid of the two: part in the
box and part on the desk. This allows some variation in the tone as well as
the tactile and ergonomic variety of not looking at a screen. This hybrid
approach is how I mixed Buck 65’s last record, Neverlove. I discuss this
record more at the end of the chapter. One reason that I like to mix on a
desk, and without total recall, is that when the day or session is finished, the
mixes were either right or not. I like that challenge—it forces me to make
decisions in the moment. Buck’s Neverlove was mixed on a desk without
recall, as was what I mixed for Beck. There’s a certain amount of fear that
pushes me in that type of situation by knowing that I’ll never have that time
again and I can’t just go home and open the session to make changes. Thus,
the safety net of recall creates sterility and mediocrity. There are times,
however, when I don’t have the luxury of mixing on a desk and will use an
old Yamaha summing mixer from the 1970s. It is a simple six-channel mixer
with mic/line selection and three band EQs. Usually, I’ll send the drums
and bass to channels 1 and 2, the remaining music to channels 3 and 4 and
all the vocals to channels 5 and 6. This is a basic type of stem processing;
Between the Speakers 127
it glues the major sections together. Next, I may use a compressor on the
output before I print back into Pro Tools. My go-to bus compressors are the
SSL, SmartC2 and the Neve 33609. With this setup, I like to have a solid
set of analog-to-digital convertors going back in, because if the convertors
are average, it seems to cancel out the benefits of summing through a mixer
altogether. The D/A A/D conversion process with poor electronics is far
more degrading than the benefits of analog summing.
Over time, I’ve created mix template sessions, which are sessions only
populated with various auxiliary tracks that contain plug-in chains on the
inserts. The use of template sessions is another concept that I picked up
and developed from my time with Jack. These not only save time but also
accelerate the mixing pace. These include dynamic manipulation: par-
allel compressors, gates for kicks, snares and bass; harmonic variance/
tone control: saturation to distortion; width: stereo image exaggeration; as
well as pitch, reverbs and delays. If starting the first song of a full record
or series of songs, I may import some auxiliary channels or these chains
directly from the template sessions to the tracks. When I import them, they
are already named and a bus is selected; thus, I may only have to add a
send to the channel. This allows me to simply unmute several possibilities
in a short amount of time and gain vast perspectives quickly, as a result.
This approach prevents me from having to change settings and go through
a bunch of pages of the plugin. Example compressor options in these tem-
plate sessions offer a variety of colors and ADSR shaping. The tones are
largely determined by their electronic design and range from various tubes,
transformers and transistors. These options will have various settings for
attack and release time along with ratios for the ability to increase attack
for punchiness or create a longer release for sustain, thus helping to bring
out the body of a snare. I’ll use high ratios to limiting (10–20–1000 to 1)
for putting a part front and center. In the reverb world, they vary in types
again based on personality and tone: from classic metallic and shimmering
plates and chambers to the odd irregularities of a spring. I continue these
options with delays (tape based to digital), harmonic effects (tape and tube
emulation to guitar amp simulators) and stereo image effects. I really love
the versatility of the Soundtoys EchoBoy.
Another time-saving strategy when moving between songs is an
approach I learned during my days working in a largely analog world.
The goal is to not to zero out the console, but to leave the outboard gear
patched in. Generally, a band or artist records an album in chunks or in
one fell swoop of tracking with a similar setup of instruments and mics,
thus creating the possibility for a very consistent sound for the record.
Chances are that you can leave all of the outboard gear patched in and,
similarly with DAWs, route or import all the of same plugin chains on top
of the corresponding tracks of the next song. Then you can simply toggle
between the insert points being engaged and out to hear EQ or dynamic
changes. These are presets in a sense, ones that you’ve created, and more
than a single effect but chains of processing. Now, you may have to do
some slight adjustments but you are not beginning from scratch, which
can save half an hour or so because it eliminates the need to redo all the
128 Dean Nelson
routing and plug-in selection. In the analog world, this patching can eat up
hours, along with valuable time in DAWs as well. In Pro Tools, you can do
this by setting up a mixing template or using import session data to copy
inserts from your previous session to the current session’s inserts of the
respective channels. Chances are the EQ and dynamic adjustment should
be close to what you attained on the previous song. When you import from
a template session, you bring in a few choices per key insert points on the
main instrument tracks. The result is easily toggling the bypass in and out
to offer a variety of sonic possibilities. Add to this a couple back buses
scenarios and you can go from zero to sixty in a few minutes. This allows
you to save time and to progress creatively in the mix exponentially. For
every record I mix, after I’m done with the mix I typically will make a
template session or export key plugin chains to these mix template ses-
sions. To clarify, I’ll have several back bused (parallel) chains for main
elements: kick, snare, overheads, toms, bass, guitars, vocals, etc. Then,
there a variety of reverbs and delays. This works in line with the concept
of using presets of plugins as a starting point. All of these chains usually
have some tonal commonalities: for example, an EMI compressor to an
EMI EQ. As I stated at the beginning of this section, the way I have these
organized falls basically into two categories: colorful to transparent. This
is usually determined by whether they have an origin ranging from tube to
transistor. Other specs considered are discrete, transformer coupled or not,
and the implementation on an integrated circuit. Possibly, the more metal
involved equals more color. Knowing these specs can help to eliminate the
guesswork when trying to capture the desired tonal shape and color. With
mixing, you are trying to capture a moment in time, and that requires the
ability to work with efficiency and effectiveness. One should work to know
which pencil, crayon or knife to grab while it does take time to exper-
iment. This will help in your speed and the ability to execute the sonic
landscape that you or the artist is searching for in the mix. Along with the
color aspect of the plug-in, I also consider how it allows me to detail and
control the ADSR of a signal. Broadly speaking, I decide whether I need
to retain the attack, add attack, truncate the sustain or exaggerate the body.
I may also consider whether I want to knock down and smooth out the
transients or add some punch. Knowing if the plugin is based on a tube
or transistor will offer insight into how fast the compressor will respond.
Generally speaking, the beauty of the transistors over time are their speed
of capturing transients, thus allowing for more attack. The downside to
them is the apparent loss of color. Let’s check out a couple of examples.
A typical chain for the kick could be a couple of different mics: in, out and
sub. I may give each a touch of individual EQ’ing and then sum down to
an auxiliary track for global control. For the most part, the strategy for this
setup is as follows: the inside kick captures the attack/beater for some nice
punch (the 120 Hz area) or sub depending on the mics used, the goal of
the outside mic on the resonant head is to complement the inside mic with
sub or punch, and the sub-mic like the converted NS-10 or the Yamaha
sub mic is just that sub, 20 Hz–50 Hz. What you have in this scenario
of the low-end world is the kick in segments with some overlaps, along
Between the Speakers 129
with what I call nesting qualities like 20 Hz–40 Hz from the sub mic, 40
Hz–100 Hz and then 100 Hz and above. This is very close to how the kick
was treated on Buck’s ‘Danger and Play’. This is a broad breakdown, but
provides one with many options for a kick sound. What to expect from
the mics is an important consideration. The mics are funneled down this
aux, which allows you to carve further to taste with respect to the song.
In regards to the kick chain, I may use some kick back busing options of
compressors broken down in this way: one for punch, one for sub and one
for attack. That’s it. I essentially imagine the ADSR chain for each signal,
which enables me to break it down and separate it. Basically, this is what
you can do with multiband compressors, but for my simple mind I like to
break it down into these planes or snapshots of time. Wrapping up using
templates, a pitfall may be that it pushes you to familiar territory, so if the
band, artist or song is requiring something fresh, don’t go that route.
Now that we’ve talked a bit about the prepping process, let’s move on to
the beginning stages of the mix.
As I mentioned earlier, my goal as a mixer is to serve the song. So,
where does one start after the prep stage? My strategy is to break down the
mix into two broad parts: production and engineering. Some things I’m
listening for from a production standpoint are the arrangement, the lead
and ancillary parts, instrumentation, fills and transitions (handshakes), the
execution of the performances in the areas of timing and tuning, phrasing
and the dramatic arc of song: the use of tension and release, how it ebbs
and flows, where the climax may be, the melodic and harmonic structure and
overall feel and mood. From a technical standpoint, I’m listening for clar-
ity, definition, the overall balance from low to high and possible holes
or frequencies that are missing or the buildup of frequencies. I consider
whether the overall balance is muddy or too bright. Other sonic consider-
ations are tone and timbre of the instruments: is there an overall warm and
fuzzy feel, or does it feel clinical and sterile? Also, I determine if there has
been any processing in the area of compression—too much, too little? In
the area of ambience, I consider if the mix is overall too dry, or is it more
of an established ambient setting? Is it too wet or too dry? Are there too
many effects, etc.? I’ll listen to the whole song from top to bottom in these
respective areas along with soloing certain elements to see how they sound
on their own to see if there are any anomalies or artifacts. Some things
can sound awful soloed, but in the mix is what matters with the relative
relationships. As you listen, take notes, not just mentally, but actually write
them down and make a checklist. Have a keen ear to the lyrics and note
key words or phrases that should have automation pushes (rides), as well
as effects, delays or reverb tails. What I hope to obtain by this approach
is a sense of any major concerns, pros and cons, whether I need to cover
up a poorly played part that cannot be fixed with editing or try to provide
clarity for the vocals. If the song sounds like it’s in a good spot regarding
production and engineering, I typically just take the approach of buffing,
highlighting all the beautiful moments and building off of what’s already
there. Mixing is this process of relationships and relativity. If you adjust
one element, you should consider the effect on the surrounding parts, not
130 Dean Nelson
just what is happening in the immediate vicinity (section), but down the
road in the following sections of the song. Analysis to me is what’s work-
ing within the mix and what seems unrealized or what’s not working to
its fullest potential. It all goes back to the song. As a mixer, I believe you
toggle the hat of a producer and engineer. On the production side, I ask,
“What does the song need?” Sometimes I believe that songs, while techni-
cally complete, may feel like something is missing, and that is something
you can provide during mixing. You may have to fly a part around, mute
something that is arranging, or rely on some bells and whistles of ‘tricks’,
weird delays, reverbs or effects. It can be like a sleight-of-hand card trick.
Mixing is guiding the listener through an aural landscape of a sonic story.
Some artists seem surprised or caught off-guard by this approach when
they realize that a mix is not just simply a process in sonics. It very much
has to do with a holistic understanding of the song and why it might then
be necessary to mute a part or fly something around. I can operate from
a strictly sonic/engineering angle and work with EQ, compression, pan-
ning, etc., but to me it’s the full picture from a production and engineering
standpoint.
There are two places to start the mix: (1) where the band/artist left off or
(2) completely from scratch with the faders set to zero and without any
plugins. I believe that picking up where the artist left off, which could
possibly mean leaving all the plugins inserted and routing intact that was
used during the tracking and recording process, allows one to hear the mix
from the exact place that the tracking process was left off. Mixing from
this point of view allows for continuity in tone and shape. In this approach,
bypass plug-ins to see how the treatment is working. Does the EQ’ing
feel accurate? Do the compression times and ratios feel good? How about
muting effect returns? Ask yourself if this treatment is adding to or taking
away from the song. Continuing in this way, you can have the client print
a rough mix at the end of the tracking process before they pass it on to be
mixed. Then, begin by zeroing out the faders and throwing all the plugins
into bypass. Next, build the mix back up based on the rough mix and rein-
state the plugins as you go. The result may be that you ultimately tweak
parameters to your liking and substitute your own preferred plugins. In a
sense, this can be a hybrid of how the mix came in with your own magic
wand waving on top. Beginning the mix from scratch is another possibility
for starting a mix. In this process, you would not refer to any rough mix
and rebalance from zeroed-out faders. This allows for the most creative
freedom from a mixing standpoint. After deciding which of the above
ways to take, there are then a couple of places to start with the song itself,
instrumentally speaking. The two main ways I begin a mix are either with
the foundational—low-end world of drums and bass—to the lead melodic
element, which in most cases are the vocals. See Phil Harding’s chapter in
this book (chapter 4) for an expanded discussion on this.
Between the Speakers 131
Since I was a kid, one of the consistent draws to record making has been
my obsession with tone. Of course, I did not know that then, but as much as
the song stuck with me there were also the tones and textures. A large part
of my tonal education came from my time spent at Ocean Way with JJP
and the chief tech Bruce Marien. I knew the instrument choices played a
large role in the sound, but it wasn’t until I got into recording that I learned
that the other part of the equation were the mics, pre-amps, compressors,
tape machines and the studios themselves. Words I use to describe tone are
warm, woolly, fuzzy, dull, dusty, woody, vibrant, shrill, metallic, glassy
and crystal clear, just to name a few. From an educational and financial
standpoint, an awesome thing about working with virtual plugins is the
ability to a get a gist of the classic pieces. The cons are, however, the access
to so many immediately. I often struggle with narrowing down the max
choices. It can be difficult to match the piece of equipment to that specific
sound in your head. In a very broad way, I think of tone as ranging from
really transparent (very true to how an instrument sounds before going
through the signal chain) to extremely colorful and vibrant. In order to
capture the spectrum of equalizers and compressors, a very general way is
to narrow down tone control choices into the fields of tubes and transistors,
along with the inclusion of transformers or op-amps to the current use
of ICs. Again, this is a very general division in order to begin narrowing
down gear/plugin choices. What you are getting from these different types
of electronics is the ability to shape the ADSR plus apply a certain amount
of color or not. Transistors provide better transient response, as well as less
color with the removal of transformers. For example, with Neve 1073 you
get transformers with a healthy amount of iron on the input and output,
resulting in some nice warmth and thickness and with a transformerless
GML a very true and clear sound. On mixes, I occasionally try to chal-
lenge myself to sticking to one model for EQ and compression. Depending
upon the sound and style of the band, with this approach you get great
tonal consistency. An analogy I like to make uses the medium of visual
arts. In drawing, let’s say you choose to stick with only pencils, charcoal
Between the Speakers 133
or pastels. You may find that you can still do all the dimensional detail, but
because there is only one medium being used by default, there is tonal con-
tinuity. All the records that have come out of historic studios like Olympic
with Helios, Abbey Road with EMI, Motown with Quad 8/Electrodyne
and Trident with Trident have a particular sound, a distinctive sonic fin-
gerprint. I’m guessing this is because there were specific and unique desks
in each of those rooms. In these classic studios, they designed and built
their own consoles and gear. And if they did buy stock equipment, chances
are that the studio techs modified it in some way that aligned with their
aesthetic. Hot rodding of sorts. One great example would be the Altec
436 mod EMI, which resulted in the RS124. (By the way, there is no other
compressor that sounds and responds like the RS124.) Beck had two Altec
436s with a version of the EMI mod, and those can be heard on the over-
head for ‘Ramona’, ‘Threshold’ and ‘Summertime’ of the Scott Pilgrim
soundtrack. All these studios, for better or worse, have sonic fingerprints.
Having worked at Ocean Way, I can hear the fingerprint of the chambers
on records from Frank Sinatra to Beck.
On Buck 65’s last record, the songs I worked on were done in Studio C
at Revolution Recording in Toronto. The desk at that time was a custom
Ward Beck. Sonically, it was clean and punchy, and while it had a superb
balance top to bottom EQ-wise, it did not have the ultra-round bottom of
a Neve, leaving out the ‘darkness’ on top. One very appealing factor of
working in that room and on that desk is the fact that it’s one of a kind. No
one’s going to go to the software inserts and load that plug-in. It only exists
at that studio (or it used to—I’m elsewhere now). Maybe I’m selfish, but
for me this is really cool and ties to my values of record making. Please see
the Buck 65 examples.
A concept based on this method of tonal continuity is to limit my spe-
cific model choices to the same pre/line amp, EQs or compressors on all
the needed elements of the record. For example, the Waves NLS (Non-
Linear Summer) has three models of consoles, a Neve, EMI and SSL.
When starting a mix, I may put a Waves NLS on every channel. Or, for
a bit of a more modern angle, I may have all the drums routed through
the Neve setting and the guitars routed through EMI and across the ste-
reo bus the SSL. A little mixture of flavors distributed to the instruments,
which tend to be flattered by the inherent electronics design of each model.
A classic combo is that of tracking through a Neve and mixing on an SSL.
Or, as with the Record Club Skip Spence, INXS and Yanni, those were
tracked on the custom API at Sunset Sound and mixed back at Beck’s
studio on the Neve 5315. One can start by tracking all of the warm and
woolly color of the Neve and then try to whip out the punchiness and
aggression by mixing through an SSL desk or, at the very least, using
an SSL compressor strapped across the stereo bus. A ballpark setup for
bus compression for me is a low ratio of 2:1 or 4:1 with about 3 dB–4
dB of gain reduction. Attack and release are dependent upon the tempo.
A helpful starting point is to start off with the slowest attack and fastest
release. Start increasing the attack and once the transients of the snare, for
example, disappear, back off. I will insert the stereo bus compressor fairly
134 Dean Nelson
early on in the mix and get some rough settings. Adjust the settings on the
loudest section of the song. If you set the stereo compressor on a quiet
section, like a verse, when a louder section, like a chorus, comes in, the
compressor will jump on it and suffocate the mix. The gain reduction will
be too much. Keep in mind throughout the mix to frequently check the bus
compressor to see that it’s not working too hard, killing needed transients
or pumping and breathing erratically. All the examples listed have an SSL
comp on the 2 mix.
I haven’t really hit on tube pieces yet—they’re all about the harmonics!
In the world of classic tubes, I think of the Pultec EQP 1A, LA-2A or the
Fairchild. These are pieces I’ll insert onto just about anything, certainly
any element that I feel is lacking what I would call ‘vibe’. For example,
acoustic guitar, vocal, piano and overheads really shine going through one
of the above. Richness is a common word to describe the sound. I’m not
even looking to do much in the area of compression, but just to add some
harmonic color. The Pultec has an amazing roundness to the low end and
a super smooth top end. The unique feature of the Pultec is the use of a
resonant shelf, which has the ability to cut and boost the same frequency.
I used the PuigTec on the Kick for Buck’s ‘Superhero in My Heart’. I like
to use it on the stereo bus with a little bump on the top and bottom. I was
lucky enough to be working for Jack when he did the PuigTec and Puig-
Child plug-ins with Waves. Jack sent his Fairchild to the Waves in Tel-Aviv
and once they did all the electronic measurements and coded up the first
beta, they sent us a copy. When we first opened it up, it responded like
his 670, but it was missing that ‘x’ factor. There is a beautiful open and
silky sound in the high end that comes from just inserting the unit into the
signal path without even pushing it to compression. So they sent us a ver-
sion, essentially with the hood open with a variety of parameters to tweak.
Eventually the ‘shimmer’/x-factor sound got dialed into Jack’s liking—a
fascinating experience. I guess when you have fourteen transformers and
twenty tubes, something epic should occur. In the end, I think the plugs
have the essence of the analog counterpart.
So far I’ve been referring to the use of tone control from the standard
use point of view, meaning ‘the technical and user manual correct way’.
What happens when these units were used in a slightly ‘wrong’ way is
what’s fascinating to me. I grew up in the 1980s and ‘90s on hip hop, punk
rock and skateboarding, and none of those things was about following the
user manual. Mixing for me gets really interesting when you start pushing
gear past the comfort zone, like hitting the input hotter and bringing down
the output. This is when you can really start to hear the true personal-
ity of some gear. Pushing the sonic comfort zone goes hand-in-hand with
my ethos of finding the ceiling. Working with Beck on a mix once, he
said, “sometimes you just have to throw orange paint on it”. I believe that
was his way of saying the mix was too safe, boring or lame and it needed
something extreme and beyond textbook logic. One route to explore in this
way is the ability to overdrive various stages of gear from the aforemen-
tioned tubes, transistors and transformers. Thinking from opposite parts
of the spectrum from mild to extreme, a very mild palette change is tape
Between the Speakers 135
you can just reel it back. Keep in mind, however, that sometimes you only
need to push one element to the ceiling. If you push every element, you
lose the use of juxtaposition. It all comes back to the point of arrangement:
aim for a glaring effect of reverb or delay or a crunch of distortion to get
the attention of the listener and continue guiding the ear through the mix.
My approach to developing moods that have been established in the pro-
duction shall be discussed here. For example, if the song is leaning towards
the aggressive side, I might add distortion and extreme compression on
the drums, snare, bass or vocals. If there is something mysterious or unre-
solved, add some small pitch fluctuations in a reverb or delay. I try to play
off the moods of the song, whether they are happy, sad, ironic, pensive,
melancholic, dark or light. These are basically the same adjectives one uses
to describe the moods established by keys and chords of the song. As I have
already mentioned possible options for tonal development, another aspect
to consider is depth of space. In my head, I imagine something like the
textures and the painting technique of impasto, similar to impressionistic
paintings where you can see the ridges, valleys and blurred edges of the
brushstrokes. A way to develop depth and size is with the use of high and
low pass filters along with delays and reverbs. I start by trying to establish
the ambient setting of the piece, hopefully with any natural ambience from
the recording. For example, I’ll bring up the room mics, and use compres-
sion on the drums or any other element that I can extract some natural
ambience out of. I start by considering close and far, foreground or long
shots in composing the dimensional depth aspect. I like to see how close
I can pull elements forward in the mix. One way to do this is by adding top
end with a shelf, somewhere around 6 kHz and above along with a high
pass filter. To push elements towards the background, I’ll use a low pass fil-
ter and roll off the top end. Some basic settings I use to divide up my delays
are 10 ms–30 ms for building up size, 30 ms–250 ms for trailing highlights
of slap, 250 ms and above for the clear intention of delay. My general view
of reverbs are to develop the size and timbre of the ambient space. I’ll
set up a near (less than 1 s, nonlinear reverbs like the AMS-RMX), mid
(1 ms–4 ms) and far (4 ms and above) to divide up the distance from fore-
ground to background. I’m a big fan of plates and chambers. I find myself
using the Rverb by Waves and McDSP’s Revolver quite a bit, along with
Dverb by Avid. It has it has a nice grittiness to it; it works great on tambou-
rines. A mix doesn’t have to be this nice linear dissension into the horizon,
though. For dynamics and boldness you could have a couple of elements
in the foreground and just throw a glockenspiel, pads or tambourine way
in the back drenched in reverb. Treating vocals with a type of backlighting
effect is an interesting approach to developing mood. Think of the contrast
of a tree with the sun setting or rising behind it. The overall light balance is
dusk or dawn. For me, this fits into the mood of the song based on the lyrics
or the key: minor/major or something bittersweet with the use of sevenths
or sus chords. The tree from the front perspective is dark. It’s mostly a sil-
houette, while the sky could have some hues of pink, red, orange and blue.
By brightening the reverb up with a shelf, around 7 kHz, allows the mix to
open and shimmer and illuminates the vocals. To further contrast the bright
Between the Speakers 137
reverb, the vocals could have a bit of the top end rolled off. Another thought
is to take the reverb signal and add a high shelf.
One last way to avoid predictability and to establish ambiguity or unre-
solved type of moods in songs is with asymmetrical panning. When listen-
ing to music, it’s obviously happening over time, so we can thus include it
as a dimension that we are mixing in. For example, in a verse, the right side
could be totally empty, or possibly include some stereo overheads, toms or
hat. But the song only has one lead/rhythm guitar on the left, while in the
chorus the doubled guitar comes in on the right or a keyboard part. That
empty space creates some tension, and the resolve is the balance in left and
right attained as the doubled part comes in at the chorus. For the most, part
of what is going to musically happen in a chorus is to provide some sort of
harmonic resolve. I separate panning into two categories: left, center and
right (LCR) and internal. The idea of only using LCR has a bit to do with
the old Delcon Console that was in Studio B at Ocean Way. It did not have
variable panning, so you were forced to only use hard left, right or both.
I typically start off by panning LCR, so elements like piano and overheads
recorded in stereo will be panned hard left and right. The actual width is
determined by the stereo miking configuration. XYs will have a better cen-
ter image and spaced pairs will have better width. Sometimes I’ll pan the
piano hard left and center, as a piano would be stage left in a mental image.
One distinction you can make for the image or landscape you are trying
to create is to set it up as a band would be on stage at a performance. Or,
make it more free in the sense that things are floating in this ethereal ether
of space. For example, if you have a singer/songwriter with very simple
instrumentation of vocals, acoustic guitar and percussion, you could pan
the acoustic guitar hard right, separating it from the vocals, thus creating
distance and space. This distance could, possibly, be interpreted by the lis-
tener as a loss of intimacy. On the flip side, in line with how you would see
a live performance, the guitar should be panned down the center. For me,
separation between elements creates more of a surreal, bold, dream-like
feeling, and keeping things panned closer together can be more associated
with seeing a live performance. In keeping with the LCR concept, each
portion of left, center and right can be separated into foreground (near),
midground and background. I like to reserve the center part of the spec-
trum for very strong mono elements, such as vocals, bass, kick and snare.
On the left and right, things like a rhythm guitar playing an eighth-note
part can sit with a Rhodes playing chords. One thing to consider when pan-
ning is the rhythm being played. Be cautious of panning similar rhythmic
parts on top of each other, because if they are not dead nuts in the pocket
together it can sound messy. If there is a shaker playing eighth notes and
the hi-hat is doing the same, I would pan them off each other, situating
one on the left and one on the right. Internal panning for me is the space
between center and hard left or right. I’m hesitant to just use every empty
space, because I think it’s a bit like cheating in the sense of not EQ’ing
properly. Also, things can get messy with every spot filled. Watch the clut-
ter. Sometimes when you give the ear connection points from center all
the way out to the outer edges of the stereo spectrum, I believe you lose
138 Dean Nelson
some of the width illusion, because the ear has this straight line to connect.
When there’s information in the center and the next bit exists on the perim-
eter of the spectrum, you can have a better illusion of expansiveness and
width. Keep in mind that there will be information in that space if there
were stereo mic setups used. I try and keep that space for any part that
needs to jump or needs to have a little of the spotlight. For example, toms,
hats or a little lead keyboard part.
In the end, listen, listen, listen (if you are spending most of your time
twisting knobs you probably should take some time to listen and walk
around the mix/console). Sometimes I’ll just pace around the desk for
thirty to forty-five minutes just listening between the speakers in the next
room. Leave room for chance and error, use limitations to drive creativity.
Have your staples that you trust and know to move efficiently throughout
the mix process and grab the listener’s ears by finding the ceiling colored
‘orange’!
mix approach was to take what came in from the tracking side and further
that vision. For this, I pushed the midrange of the guitars fairly extreme,
10 dB+ in the 2–3kHz range. It’s a delicate balance because you don’t
want the mix to be fatiguing while still remaining both forceful and edgy.
I recall Beck pushing me further and further with the tones and thinking
“really, I can EQ the guitars that far?!”
In regards to Buck 65, we met just before I moved to Canada from LA.
Once I got to Canada, we worked on a track called ‘Dolores’, and one
thing I remember doing that seemed adventurous was recording a marimba
through a Leslie cabinet. It was really spooky sounding and surely caught
Buck’s ear, as he next asked me to produce/record/mix some of the tunes
off the next record. One tune that comes to mind in particular is ‘Baby
Blanket’, a really dark and sad piece. This is one that was developed from
the recording process. At Revolution Studios in Toronto, they have a really
cool pump organ that we used; one thing we did was to make a loop of
the mechanical noise of the organ, which provides this really creaky and
dark vibe. In mixing this one, we transferred the multitracks to an 8-track
2-inch Studer. In this process, I summed the multitracks down to those 8
tracks and then printed them back into Pro Tools for some final mixing.
Going to tape helped further the warmth, depth and vibe of the final track.
At the of the day, having the opportunity to spend time growing and
evolving with such exceptional artists—and to have their trust with cre-
ative exploration—is both a huge honor and education. Throughout my
career, including my time with Jack Joseph, there are moments of extreme
tutelage and wood shopping. With Jack, I spent months simply listening
and watching, and slowly the trust formed and I moved into engineering
with more responsibilities. But it wasn’t until my time with Beck that I was
placed directly into the hands of the artist and left on my own to mix. You
must become the hands to the artist’s vision yet be guided by your own
intuition and solid foundation.
Discography
Buck 65 Neverlove 2–548156
Beck and Bat For Lashes “Let’s Get Lost” / Eclipse Original Motion Picture Soundtrack
523836-2
Scott Pilgrim 0343–2
Beck: Record Club
Skip Spence ‘Grey/Afro’ https://vimeo.com/9638358
INXS ‘Guns in the Sky’ https://vimeo.com/10245433
‘Never Tear Us Apart’ https://vimeo.com/11991409
‘New Sensation’ https://vimeo.com/10634950
‘Calling All Nations’ https://vimeo.com/12662870
Yanni ‘Santorini’ https://vimeo.com/12996440
‘Keys to Imagination’ https://vimeo.com/13184967
9
140
Mixing for Markets 141
and they should be auditioned where mentioned in the text. In doing this,
I have done my very best to let the artist speak directly to an academic
readership about his craft. It is my sincere hope that what follows—which
I would call a methodological hybrid of ethnographic interview work and
straightforward musicological analysis—models a viable method for aca-
demics to draw successful artists directly into academic discussions about
their work.
Mixing is balancing all elements in a track, and giving each track its own
place. At the same time, it’s guiding listeners through a song, dictating
what they should hear, precisely when you want them to hear it, to help
bring out the emotion of the song.
Mixing is different in that it is a lot more technical than many other tasks.
When you mix, you are trying to tease out the emotions in a track. The pro-
ducer makes decisions like, “Let’s play this chord here”, and “Let’s arrange
the song this way”. The editors edit it, and the mastering engineer masters
it, all to ensure the production sounds good as a final product. It is the
mixer, though, who takes all of the elements the producers and engineers
and musicians throw into the stew and makes it musically and technically
coherent and pleasing. It’s the mix engineer’s job to make a track ‘sound
good’, both from a technical and aesthetic standpoint. And you can’t make
a record without mixing, of course. So I’d say it’s one of the most essential
tasks in record production.
Engineers learn to mix by ‘playing’. You hear something, and you think,
“How can I change that? How can I make it better? How can I make it dif-
ferent?” You say, “What can I do with these sounds so they make sense as
a mix and as a song? And what tools do I need to make that happen?” Then
you just ‘play’ with the tracks until you get them to sound more or less
how you want. That’s the only way to learn, in my opinion—by playing,
and making a lot of mistakes, and doing a lot of things wrong, and learning
from it all. At the same time, you should be watching people who are good
at mixing and learning from what they do. When you watch people work,
you might hear some things you like and some things you don’t like. So
you’re watching and you think, “I kind of like this, and I don’t like that”,
and then you can adapt what you see to your own way of mixing things. In
the end, if I’m doing what I should be while I’m watching the pros work,
142 Alex Krotz with Jay Hodgson
which is ‘playing around’ with some mixes on my own, I’m going to learn
to mix tracks in my own unique way, even if I’m basing much of what I do
on the things I see and hear the pros do. And that’s what makes every mixer
different. It’s how every mix engineer eventually finds their own voice.
I don’t know if there’s any particular talents that might indicate someone
would be good at mixing. That said, you have to want to do it. You have to
want to mix to get good at it. And that’s not as common as you might think.
Mixing can be super hard, and it’s often long and drawn out as a process.
Unnecessarily so, in fact, given the wrong client! People might initially
think they’d like to be a mix engineer, but after their first few sessions, or
after they deal with their first difficult client, they’re not so interested in
becoming mix engineers anymore! You have to want to be a mix engineer
despite how grueling it is.
It’s a bit of both, to be honest. The way that I work, and the way of working
I’ve been around most in the last few years, is ‘mixing as you go’. It’s a
very popular way of working because you can keep throwing things at the
canvas—you can retain an experimental approach to arrangement, without
having to fully commit to anything—knowing that, if you change your
mind later, you can always just delete whatever you added to the arrange-
ment the day before. That said, you still have to make the production sound
good, you have to make it sound like something that gets the musicians
inspired to do more to it, so you mix as you go.
In a lot of ways, you actually have no choice but to mix as you go. If
you record a guitar too high a level, for example, and you don’t mix it into
the track a bit before moving on to the next track, then you won’t hear
anything but that blaring electric guitar when you listen. And so much of
production is listening back and making decisions. So you’ll have to mix
that guitar a bit, make it fit into the production, before you can move on.
Then the way you mix it may inspire you to go back and record the part
a little differently, or do different things here and there. But that’s part of
the production process. Things are a lot more open now, even if you know
precisely where you want to get to going into a project.
Another thing I’d add is that part of production in general is to figure out
where there are holes in the frequency spectrum. You have to ‘mix as you
go’ to hear that sort of thing.
Of course, this all said, once you’re done with the engineering phase,
and all that ‘mixing as you go’, then there is still the proper, standard
Mixing for Markets 143
‘mixing’ phase. In this case, you know you’re not adding any more ele-
ments. So you’re more interested in making all the elements you’ve got
fit together and make emotional sense in relation to the song. Sometimes
that’s done with the producer in the room, dictating what the mix engineer
should do. Other times, the producer just hands off the session file to a mix
engineer, with maybe a few words about where they’d like the production
to go, and leave everything else up to the mix engineer. In that case, the
mix engineer is expected to do a bit of subtractive arranging. They might
mute some tracks, move other tracks around, maybe add some delays and
phase effects here and there. They could say, “it would be cool with delays
here”, and they get creative with it on their end as another stage of the
production. At that point, it’s another phase, though. Generally the mixer
is provided with a ‘rough mix’ that the producer is happy with so he knows
the overall sound of the track, such as distorted vocals in the verses then
big clean reverb vocals on the chorus. This is guide for the mixer, but
doesn’t mean he has to or will do the same things. He can take some liber-
ties but at least has a starting point.
The revision process is funny. Sometimes you get one note for revision,
and sometimes you get a hundred. It always varies. It just depends on who’s
on the project, and how good you are at capturing whatever it is they said
they wanted when they handed off the session file, and how much your
creative tastes match up with what they like.
Generally, there’s no set person, or agency in the process, who gives
mix engineers their notes. It really depends on the project. Sometimes the
producer has a bunch of notes. Sometimes it’s the band who has the notes.
And sometimes it’s the management team that’s got a million notes for
you. So many people hear the mixes you submit, and often a lot of different
people will get a say in approving mixes, so there’s lots of opportunity for
people to disagree with the decisions you make while you’re mixing. But
usually, over time, as you get to work with people on a more regular basis,
the revision process becomes less and less drawn out—your clients will
learn to trust you a bit more with every mix they approve. And if you’re
willing to listen to their notes, and really honestly try to deliver the mix
exactly as they want it, and you don’t force your vision of the song on
them, they will trust you even more in the end!
In my case, I usually mix a track and then the producer listens. They’ll
have some notes before anyone else involved in the project ever hears the
mix. When I get their notes, I say, “Great!” It’s their project, after all, so
what do I care if they think, say, the snare should be a dB up or down,
and maybe the drums could be a bit louder in the chorus? I just do the
revisions and send the track back to them. They usually approve, or maybe
144 Alex Krotz with Jay Hodgson
have another note or two, but then the mix goes off to the band for their
approval. The band will have some notes, usually, which are generally not
as useful as the producer’s notes. Usually band members just want to hear
more of themselves—that’s the honest truth! You might get a note from the
guitar player, saying something like, “I want my guitar louder.” Then the
keyboard player chimes in, saying, “I want my keyboard louder.” And then
the drummer says “I can’t hear my kick!” The revision process becomes a
bit of balancing act, at that point. Part of being a mixer is dealing with cli-
ents in these situations; there’s a client management aspect to the process.
In fact, here is where the mix engineer might actually ‘overrule’ a mix
note. But you have to be really careful how you do it. You have to over-
rule musicians, often, because it’s just not practical to raise the volume or
profile of tracks any more than you’ve already done, but they just don’t
understand that. Sometimes, and this is being honest, the way to ‘overrule’
a mix note is to tell a little lie, saying, “Yep! I did that. Great ears!” And
you literally did nothing to the mix, but all of a sudden they like it better.
This makes me wonder if, sometimes, mix notes aren’t just a way for musi-
cians to feel like they’re staying involved in the process, because a lot of
musicians can feel like the project is ‘out of their hands’ once it leaves the
tracking stage.
Also, if there are just too many revisions to deal with, you have to ask,
“Was I the right guy to mix this? Maybe I shouldn’t be the guy to mix.
Maybe you should try another engineer.” That does happen quite a bit,
actually. Engineers have to say, “Maybe I’m not the right person for this”,
a fair bit. It doesn’t happen on lower budget projects as much, because
most projects blow their budget on the bare minimum. But it happens on
projects with larger budgets as well. It’s not always the mixer who says
that either. On other occasions, the producer or band realize after a few
rounds of mix revisions that “this guy just isn’t getting our vision”, and
they kindly say thank you and move on to another mixer.
That really depends on the production. A project I just did, for example,
was recorded in one day. And we did a ‘rough mix’ as we were recording
it. At the end of the day, we had a rough mix, and everything was edited.
Then we did a ‘mix session’ on a later day. The band then had a couple
notes—they wanted to try harmonies in a couple different spots, mute the
guitar here and there, and that was it—and then the track was off for mas-
tering. But that’s an exceptionally quick turnaround. And it’s rare that a
project gets done like that without having some corners cut as a result of
everything being so rushed.
On the flip side, I’m working on a project right now where I’ve submit-
ted four different mixes, each of which was supposed to be a final mix.
Two other engineers mixed it, before I got the gig, plus there were ‘rough
mixes’ made along the way before it even reached mix stage. And now it’s
Mixing for Markets 145
being sent off to another mix engineer, who will produce separate mixes
for the single releases. Once we get those mixes back, I’ll have to remix
the rest of the album to match balances with them! These tracks have been
mixed so often, now, that you just know there has to be a very sizeable
budget involved! And if you can afford to do it, why wouldn’t you? Why
not labor over your tracks until everyone involved feels like they sound as
good as possible? Once a track’s been released, especially when a band’s
operating at the very upper echelon of the industry, you can’t just turn
around and say, “Thanks for buying our album but we’ve decided we don’t
like the mixes anymore, so we’re going to have the album remixed and
then we’ll re-release it!” Once it’s released, you have to live with it. And
your reputation depends on what you release. So I understand why produc-
ers and bands are very, very careful about what they put out there, and why
the mix process is sometimes very, very drawn out as a result.
Let’s say it’s a jazz session, though. Are you still going to
start the process the same way? Are you going to start with
a kick and get your balance?
Not necessarily. Sometimes, if it’s a jazz record or in a similar genre, I’ll
still work on the drums first and work outwards from there. However, in
that case, I’m not worried so much about where the kick drum is sitting
dynamically. In fact, on a jazz record, you often barely hear the kick drum;
you hear the snare and the cymbals more than you hear the kick, because
the overheads and room mics are doing most of the work. In that way, mix-
ing is a bit ‘looser’. With rock, you have to be so careful about levels and
headroom because you know, at the end of the day, the mix is going to be
squeezed for every last bit of dynamic space the mastering engineer can
manage. So with jazz you know you have room to maybe sacrifice some
level in favor of a deeper tone or feel. But you still need to balance things,
and the kick is a solid place to start.
In fact, with genres like jazz, mixing entails a lot more trial and error
for me. That’s nothing inherent in the genre, though. I just don’t mix a lot
of jazz records, so I don’t really have any common practice techniques that
I rely on. I’m sure if I mixed jazz a lot, I would have a fairly rigid method
worked out for it. Given how important I just said the ride cymbals are in
jazz, I might start there.
In general, regardless of what genre I’m mixing for, I want to figure
out what’s important in the mix. If it’s a solo instrument that is going to
be really loud, then I hang all the mix elements around it, both in terms of
their level and tone.
For a pop mix, on the other hand, there’s a bit of a formula at work.
Certain obvious things need to be where they need to be. In this case, I’ll
start with a big bass-heavy kick drum. The bass is really important for me,
so I tend to start mixing there, whatever the music I’m working on. This
is partially preference, but it’s also a technical thing. Mixes get messy and
overwhelming, really fast, in the bass region. Bass pushes a lot of energy,
and often you don’t even consciously hear it doing so. Bass frequencies are
really big waveforms, and take a lot of energy to propagate, and take up
Mixing for Markets 147
a huge part of the speaker load. If you take a massive kick drum and then
add a bass guitar on top of that, and then a bass synth, or any other bass
element you might add, your bottom end will quickly blow up and get out
of hand. A mix where the engineer hasn’t filtered the bass can easily make
your 2-Bus compression really go insane, and that’s going to compromise
some level and your mix will sound all the quieter and muffled because of
it. So that’s why I like to start with the bass, regardless of genre. Almost
on reflex, I’ll start by rolling some bottom end off a bass, because I want
the kick drum to push the absolute bottom frequencies of a mix, and then
I will fit the bass guitar around that. It depends on the mix of course, but
that’s where I like to start. It is all about organizing the frequency spectrum
and giving each element a place.
a bit, and then it sticks out of the mix all of a sudden at that point. Then
the vocals are back in, so you bring the guitar part back down in the mix.
and shouting it can often seem like it’s loud. It’s present, of course, and you
can hear it in the mix, but it’s not the same level of loudness as in a pop mix,
for instance. The kick drum in metal is not as big as a rock kick drum, but
it’s very ‘ticky’. Half the time the metal performers are playing very fast and
intricate kick drum parts, so the drum has to be ‘ticky’ or else you aren’t
going to hear those intricacies. You can’t have a big kick drum going on when
the arrangement calls for thirty-second notes—it just isn’t going to translate.
Then you go to jazz or blues, and mixing decisions become more about
balancing all the parts in the track together, as a whole. There’s no big kick
drum as a center point, and there’s no big vocal as a center point, in those
genres. Instead you have to feature everything, in a way, and bring individ-
ual tracks forward only during solo sections.
these days. There just aren’t that many tracking engineers making head-
lines lately. They exist, don’t get me wrong. But for the most part, mix
engineers are more of a going concern in the industry.
Take a look at the back of a record in your collection. It will probably say,
‘Produced by so-and-so’ and ‘Mixed by so-and-so’. But who knows who
engineered it? A record’s cover doesn’t usually advertise who engineered
it. The fact that it often advertises the person, or persons, who mixed it tells
you all you need to know about the status of mix engineers in today’s market.
So, yeah, I think mixing is going to survive for a long while still, even
as money gets tighter and tighter. Sure, a lot of people record in their bed-
rooms nowadays. But they still want what they record to sound good. And
they’ll pay to have their records mixed by someone to make sure that it
does! That’s where ‘bedroom producers’ seem to spend their money now.
Then they try to get it mastered by whomever they can find, or they’ll pay
the mix engineer they hired a little extra money to have them slap a few
things on the stereo bus and ‘master’ it. With mixing, almost everyone
understands that it’s something they have to get done, and that it can make
a major difference in the way their record sounds. So ‘bedroom producers’
these days usually won’t pay money to rent a studio—and the cost of rent-
ing a studio for a day, even, can easily, and quickly, quadruple what it costs
to hire a mix engineer to see a track through to completion!—but they will
put their money into mixing. And that seems to be consistent.
Anybody can buy Logic or Pro Tools, and sequence and track some
sounds. But they probably don’t know how to make those sounds sit well
together, sound optimal as a group. And they almost certainly won’t have
the experience to know how to make the mix decisions they make best
serve the emotional content of a song. They’re usually too fixated on just
ensuring that their mix is commercially viable to worry about aesthetics,
since so much expertise goes into producing even just a basic balance.
They need a mix engineer, and they know it. And because of that, I think
the current model of mixing, where you have a separate mix engineer and
a separate phase of the production process dedicated to mixing, will con-
tinue on strong for at least the next little while.
The same thing will happen with how you treat your vocals, your guitars,
your bass. It’s really just a question of developing as an artist. You can’t
just deliver the same balances over and over again, or pretty soon your
mixes sound like they were in a different era. But it’s a slow process. And
engineers tend not to be particularly conscious of the changes they’re mak-
ing to their craft as they’re happening. They just know that they want to
keep growing as artists, so they continually fine-tune their craft. And this
happens at an industry level. So the sound of mixes is constantly evolving
as individual mix engineers evolve the sound of the mixes they produce. It
goes hand in hand with artists developing their sound as well.
Then there’s the technical aspect of mixing. Things have changed tech-
nically, almost to a point where the way you mix now is entirely different
from how you mixed even just five years ago. Or it can be, depending on
which technology you decide to use. Plug-ins have gotten better at doing
the things they’re supposed to do, for example. That’s a big one in the last
ten years. In fact, some of the emulation plug-ins have gotten so good
that even the most stubbornly out-of-the-box engineers have gone almost
entirely in the box in the last few years. That’s definitely a game changer.
Note
1 In this chapter, the word ‘market’ could be interchanged with the word ‘genre’.
10
153
154 Andy Devine and Jay Hodgson
Adam Marshall I’d define it in two ways, depending on the context. First,
(Graze/New Kanada): mixing in a production sense, I’d define it in the tradi-
tional way: of arranging and recording certain settings
during production or final mix down. Second, I’d also see
mixing in a DJ-specific sense, where it’s a live perfor-
mance of in-the-moment matching, cutting, fading and
riding two (or more) records during a DJ performance.
The main difference between the two would be that mix-
ing in traditional production practice is usually a tightly
controlled and planned operation, with a specific desired
outcome, whereas mixing in a DJ sense is more of an
in-the-moment experience where a lot of the magic (or
chaos) happens (or occurs) when taking chances.
Pierre Belliveau Mixing is finding the middle ground for all tracks to
(Gone Deville): live together without stepping on each other too much.
Mixing is about giving each layer its own space within
the spectrum.
Phil France Mixing is the stage of the process that you come to
(Phil France/ when you’ve got a track sounding as good as you can,
Cinematic Orchestra): before mastering.
Ryan Chynces (Rion C): Mixing is the art of combining distinct layers of sound,
or sound recordings, to produce a unified whole.
TJ Train (Room 303 I’d say it’s blending different instruments and sounds
/Night Visions): together to allow the space for each instrument to be
heard both independently and together.
Mixing In/and Modern Electronic Music 155
Andy Cole (LuvJam): Mixing is balancing audio elements to create the per-
fect output, where everything is clear—an audible
balance of sounds created purposely for your desired
audience or media output.
Ryan Chynces: I came to music production from being a DJ, and there-
fore my mixing sensibility was informed by combining
records in a live setting. I taught myself to mix in soft-
ware such as Ableton and Adobe Audition, with the help
of books on the topic, and lots of trial and error.
Andy Cole: From an early age, I have always been aware of the
balance of sound. Actually, my first experiences with
mixing developed through learning to DJ mix—sim-
ple fades, panning and balancing bass/treble/high
layers. I then further developed my mixing through
my television work, where I would layer sound-FX
onto my motion graphics. I learned not to ‘peak’ any
sounds, thereby creating a good, clear balance. Then
I have developed my skills further by using Pro Tools,
Logic Pro and Adobe After FX.
Craig Bratley: I learned to mix through a lot of trial and error. I also
read a lot of books on the subject, which led me to
study for a BSc in Music Technology. I still read a lot
on mixing today, as there is always something to learn.
There are also some great tutorials online, such as Pen-
sado’s Place ‘Into The Lair’.
Phil France: When I’m happy that a track is finished, I’ve generally
got a pretty good idea of how I want it to sound, and I will
do all the stem bounces, and make sure all my levels are
fine, and balanced, before I go into the studio for the
mixing session. Music that an artist(s) is working on can
really benefit from having a fresh pair of ears working on
it for the mix, as the artist(s) is generally pretty close to
it and may miss better ways of doing things, or ways of
making it sound better. At the moment, largely due to the
fact that my major financial investment is renting a studio
space, but also because I’m a bit of a control freak, I do
as much work as I can myself. I have a clear idea of how
I want it to sound, usually, but also understand the impor-
tance of having another pair of professional ears listen, to
potentially improve things and also to do a double-check.
If there isn’t the budget or time for working with another
producer or engineer for the mix, I’ll get it sounding
as good as I can on my own, check the levels and then
run it through some plug-ins I have. Basically, I want to
achieve a full, warm, defined and fat sound across all the
frequency ranges in the track. I’ll run a track through a
(stereo-bus) compressor, which helps to glue all the track
together. I’ll also try to remove any frequencies from the
individual tracks that I don’t like through equalization,
158 Andy Devine and Jay Hodgson
Noah Pred: My mixing technique has changed a lot over the years
and will surely continue to evolve and adapt, along with
technology and my knowledge of it. Currently, most
individual tracks in my projects have a UAD Neve 88RS
emulation on them for gentle compression, gating and
EQ, along with optional dx-160 and saturation units,
which I can activate as needed. Importantly, all of my
track volume faders are attenuated to −7 dB by default,
so there’s always plenty of headroom, and I try to be vig-
ilant about maintaining ‘green levels’ throughout every
gain stage. Once an arrangement has taken shape, I’ll
group my tracks into sensible instrument or frequency-
spectrum buses, at which point I’ll apply further EQ,
bus compression, and vintage excitation (often with a
combination of UAD plug-ins and iZotope’s Alloy 2).
I’ve also got all of my effect returns collected onto their
own group bus, with some basic EQ and excitation to
keep my reverbs and delays from getting muddy.
Ryan Chynces Each time I mix a track, I pick one of the multitracks
(Rion C): to be dominant. I set it to zero, and then adjust the levels
of all the other tracks relative to it. I keep the volume
of that dominant track constant throughout the entire
mixing process, but am constantly adjusting the volume
(and track delay) of the other supporting multitracks in
relation to it. I’m open about what multitrack will be the
dominant (the standard of course is the primary ‘four on
the floor’ kick beat), but if I have a great harmony or mel-
ody track, sometimes I’ll build everything around that.
Andy Cole: I tend to listen over and over to each layer of sound in
a mix, to be sure it’s working individually over time,
before I listen with all of my layers combined. I have
learned by mistakes not to master my own files, not to
use stereo-bus compression tools, and just to let my
mastering engineers work their magic there! However,
referring to your ‘open’ production process, I quite
often throw in a selection of sounds, and just see if
they sit well. Some do, some don’t—I’m quite exper-
imental in that sense. Over time, I have developed
further programming skills that enable me to almost
weave certain sounds in and out with fine attention to
detail to the sound levels of the individual layers (i.e.,
when a sound should be added or removed, can be
added or taken out, when a sound might sound better
alone, to enhance the impact, and so on).
Mixing In/and Modern Electronic Music 159
Craig Bratley: I’m very open about the way I mix, as electronic music
lends itself to trying different things. There are one or
two things I always do, like having a low cut filter on
every track except the bass and kick, but I never follow
100% the same approach.
Rick Bull: I feel like I have so much to learn. My ears tend toward
a specific sonic palette (and an interesting aversion to
certain frequencies), but my grasp over some tools can
also be a little heavy-handed, perhaps. I find it easy to
get attached to a certain way of doing things, which may
occasionally blind-side me to solutions to sonic prob-
lems which might be obvious to those less attached. For
example, I remember (about ten years ago) working with
a vocal in a track which was around a quarter-tone out of
tune. After struggling to disguise this fact and making a
mess of the vocal, I settled on a combination of some-
times apologetically auto-tuning the vocal and at other
times letting it sound in its own de-tuned glory. I feel
that within the mixing environment, there is no single
‘best’ way of approaching a sonic challenge, and that
listeners will generally re-cast an errant element with
repeated listens to become part of the desired fabric of a
tune. Trying too hard to ‘sterilize’ or ‘perfect’ an imper-
fect mix can often lead to a very two-dimensional listen-
ing experience and a profound sense of distrust from the
listener. Writers like David Toop and Brian Eno address
this very astutely and point out how much mixing
conventions also change over time. To use an extreme
example, many of the mixes of Contemporary Chris-
tian music in the ‘80s sought to remove the sibilance,
hum and physical ‘detritus’ in a recording to create one
illusion of sonic ‘transcendence’ and ‘perfection’. These
mix conventions within that tradition are very different
now, particularly as we have come to associate tropes
like sonic ‘authenticity’ with certain heavily nostalgic
processing techniques like tape saturation and compres-
sion. We hear through a haze of nostalgia, and we take
comfort in the memory of certain technology. Creativity
involves acknowledging these sonic assumptions and
knowing when they might be worth subverting.
Pierre Belliveau: I guess that, in some ways, I do prefer to remain open
about the mixing process. In some ways, I guess.
Boundaries and limitations don’t always have to be a
bad thing, especially when creating—it’s good to not
always draw lines around where you can and can’t go.
160 Andy Devine and Jay Hodgson
Phil France: Being open means that you may find something that
sounds better, but I’ve found that it helps to have a clear
overall focus about what you want in mind as well. If
something doesn’t work quickly, I’ll disregard it and
then move on to the next idea. One of the things I’ve
learned is that it takes me time to listen to something,
so I’ll review and double-check it on different speakers
away from the studio during all stages of producing the
work. I’ve learned that if I’m involved in a session with
other professionals, and I’m in charge, it helps to have
a clear idea about what I want and to be prepared and
decisive. For me, in this context, the decision is always
about you and your work, and how you feel and think
about it. I’ll always address the things that come to mind
when I’m mixing, then articulate them and figure out
how to try the change them, so I can ‘A/B’ the results.
All my pieces have to ‘settle’, so there are no niggles or
regrets about the mix in my mind.
Adam Marshall: Yes, I think so. With electronic dance music, it’s
assumed that the track will be played in a certain envi-
ronment (a club system) as the pinnacle, so a lot of the
mixing and production decisions are geared towards
this certainty. For more ‘listening’ types of music,
I think things need to sound good in a much wider
variety of environments, and this must be taken into
account when recording, mixing and mastering.
162 Andy Devine and Jay Hodgson
Pierre Belliveau: I think mixing in electronic dance has the same con-
cepts as mixing for other markets, but the priorities
are not in the same areas. For example, in electronic
dance music, a kick will take a much more central
place in the mix than indie music, and indie may not be
as comparatively ‘hard’ in the low end. At the end of
the day, it’s still the same science, no matter what. But
genres also encompass different mixing and mastering
techniques.
Ryan Chynces It certainly is. For one thing, most of the multitracks
(Rion C): that are used will be MIDI based, as opposed to audio
based, and working with MIDI has its own unique
challenges and opportunities. Another difference
is that electronic producers aren’t concerned with
reproducing a live aesthetic. A rock band, for exam-
ple, might want to mix their song so that it’s more
of a faithful reproduction of a live performance,
whereas a dance music song is produced to be the
live performance.
Andy Cole: Yes, possibly, but good sound is good sound! I have
mixed folk elements with electronic styles, and the ethic
is the same. I have not really mixed rock as such, but I can
imagine it could possibly be more difficult due to its ‘live’
nature. That said, the same principles apply.
Phil France: I would think that the basic principles are still the
same—sound source, signal chain, levels, phase, equal-
ization, compression, balance, etc. I think it boils down
to whether you are recording a machine or a human
being playing an instrument. I noticed from my time
in the Cinematic Orchestra that electronic instruments,
compared to traditional instruments, have a hell of a lot
more level. I’d also suggest that certain bits of gear are
more readily associated with particular instruments,
genres and sounds.
Craig Bratley: It can vary. Mixing and writing are often part of the
same process with regards to electronic music. It’s quite
common for an artist to mix their material as they are
writing. Sometimes you need to have the right drum
sound, or get the bottom end working together, before
you can make a start. Even the sounds you choose can
have an important impact on the final balance—select-
ing the right sounds in the writing process saves you
work trying to find a space for everything when you’re
mixing. Other times, you might just write a track really
quickly, and mix it later. And there are a lot of people
who write tracks and get a mix engineer to polish them
up later, as they might not have the production chops to
get the track sounding exactly how they want it to.
Andy Cole: I can only speak for myself. I develop ‘layers’, then go
back to add more ‘layers’. Perhaps, over time, I may
compose another layer, or add a vocal that comes up
after listening to the piece. However, perhaps at the start
of my production processes, I wouldn’t ‘compose’ as
such, but simply ‘try’, ‘experiment’ and ‘see what hap-
pens’, much like a visual artist would (i.e., put something
onto paper, add to it, take things out and so on). I guess
you may start with an initial idea in your head, but the
outcome may be completely different from where you
started. I guess that is the beauty of the electronic/dig-
ital mixing processes—you can cut, copy, paste, save,
delete and so on. In terms of mastering, though, I prefer
for a mastering engineer to create final masters. If my
mix elements are clear and balanced, then it lets a more
experienced person complete the project. That’s not to
say that I always work with a mastering engineer. I have
created my own internal masters, for personal use, and
166 Andy Devine and Jay Hodgson
Phil France: Yes, I work on a piece right until the end of the mix
session. You might notice something during the mix that
shouldn’t be there (for example, an audible click on an
audio edit), for example, which needs fixing. If you’re
happy with the mix, you also need to take it away from
the coalface, so to speak, and listen on different speak-
ers until you are entirely happy that it’s good enough.
Noah Pred: I do all my mixing in Ableton Live 9.5, with a com-
bination of Ableton’s audio devices, UAD-powered
plug-ins, and plug-ins from Eventide, iZotope, Wave
Mixing In/and Modern Electronic Music 167
Andy Cole: Logic Pro for production. Allen & Heath Xone 62/92
mixers for DJing.
Craig Bratley: I use the same tools as everyone else. I have a hybrid
studio so I use the best of both worlds—a bit of EQ,
compression, reverbs, delays and so on. Nothing eso-
teric, really. I also use a summing mixer.
Which (if any) of the tools you use for mixing would
you say are indispensable to your process? Why?
Craig Bratley: EQ, as it allows me to make space for the individual
elements.
The guys at TDE [Kendrick Lamar, Ab Soul and Jay Rock] do things on
records sometimes because they know I’m going to come behind them and do
something that’s gonna make it better. Over the years I’ve been so experimental
with my mixes that they’d come in and try new things with their voice just to
see what I could do with it. They use their voice as an instrument that can be
added to.
(Ahmed, 2014)
As far as how my process works, sometimes what I’ll do, for example when
I’m working with Saigon, is we’ll start in the B room then we’ll start to mix
completely in the box. Once we get it to a point where it sounds good, or it
sounds good to a certain point but we want to run it through the analog gear,
we’ll make stems of everything. Then we can bring those stems to the A room
and finish the mix from there. That’s pretty much what we did with the Jay-Z
record. We started in the B room, all in Pro Tools, and did as much of the mix
as we could in there because we weren’t sure which record was going to be the
single.
(quoted in iZotope, 2016: https://www.iZotope.com/)
170
Groove and the Grid 171
“I think hip hop for so long has been producer-driven—not that other
music isn’t—but in hip hop the producers are stars” (quoted in iZotope,
2016: https://www.iZotope.com/). Indeed, the producer is a central figure
within hip hop music, but so much of what constitutes production and
pre-production within hip hop requires fluency in audio mixing and its
constituent techniques. As such, this article will consider several tech-
niques occasionally associated with the arrangement and production of
recorded music within the purview of mix practice. This attitude—as
demonstrated through the above anecdotes and examples throughout this
chapter—reflects the current culture surrounding hip hop production.
Generic Overview
being a student of hip-hop in general, you take technical aspects from places.
You may take a rhyme pattern or flow from Big Daddy Kane or Kool G Rap.
But then you go to Tupac, and he made songs . . . Biggie told stories. I wanted
to do all that shit.
(Rolling Stone, 2013)
Appreciation for the sonic history of hip hop extends beyond lyrical
homage, however, as practitioners often pay tribute to both past hip hop
instrumentals and other foundational elements of the genre. The hip hop
producer’s primary means of engaging with previously recorded material
this way is through the act of sampling. Sampling dates back to the for-
mation of dub music in Jamaica in the 1960s. The music of Osbourne
‘King Tubby’ Ruddock, Lee ‘Scratch’ Perry and Errol Thompson, for
example, would sample so-called instrumental breaks from B-sides of 45
RPM records and compose around them by further emphasizing drums,
adding signal processing (especially reverb) and sound effects, and other
172 Matt Shelvock
instruments. By the 1970s in America, DJs such as Kool Herc, Grand Wiz-
ard Theodore, Grandmaster Flash and Jazzy Jay had adopted the practice
of sampling breaks. These DJs formed the backbone of hip hop music
through this quasi-compositional method, and expanded upon the tech-
nique of sampling breaks by adding vinyl scratching to the available sonic
repertory (Hager, 1984; Fricke and Ahearn, 2002; Chang, 2005).
Another distinctive characteristic of hip hop, in addition to the presence
of a rapper and sampled audio, is the general emphasis placed on drums
and low-end frequencies within the culture of listening. The kick drum
and bass tend to sound extremely prominent in hip hop mixes. Clarity of
low-end frequency content is of the utmost importance to hip hop fans,
who famously, for example, popularized the use of subwoofers within car
speaker systems. Additionally, in 2012, Dre Beats held a market share of
64% for headphones valued over $100, and was valued at $1 billion as
a company (Dorris, 2013; Neate, 2013). I clarify this fact because mix
engineers routinely consider the best possible sonic representation for the
audio they are working on. In order for this best fit approach to work, the
playback systems of music fans must be considered.
As stated, creative workflow within hip hop often blurs lines between pro-
duction and mixing activities. This primarily owes to the fact that con-
temporary hip hop is built through the digital audio workstation (DAW).
Even producers who originally worked primarily within the analog domain
or on hardware, such as DJ Premier, currently rely on a DAW from pre-
production through post-production. Premier explains his admiration for
the DAW by stating:
I think Pro Tools is a great gift for all of us who have dealt with tape. It can
do a lot of things that I couldn’t do in analogue; it allows you to mess up, and
re-mess up and redo and undo, and so on. I am just in a whole different world
and frame of mind these days and Pro Tools just enhances me as a producer
and a person.
(Tingen, 2007)
Here, the ability of the DAW to quickly perform edits, retake passages and
mix on the fly clearly aids Premier’s creative process. DAWs, when used in
this way, can act both as a sketch pad for rough ideas and as a program for
finalizing production, mix and mastering decisions.
This section discusses mixing hip hop with the assumption that DAWs
remain an integral part of the beatmaking, production and mixing processes
for contemporary hip hop music. ‘Crafting the Instrumental’ describes
methods for mixing hip hop instrumentals by first discussing a number of
standard generic components including samples, percussion and bass. This
information is accompanied by suggestions for implementation within a
mix composite. After discussing a process for mixing an instrumental track,
the proceeding section (‘Mixing Vocals’) will cover strategies for mixing
Groove and the Grid 173
Sampling
To sample refers to the act of incorporating pre-recorded material into a
new musical instantiation of some type. Producers can sample from any
number of musical elements including (but not limited) to basslines, drum
breaks, fills, background vocals, strings, gospel choirs, entire songs and
any other conceivable audio source. Production duo Christian Rich, for
example, sampled a yelling sound on Earl Sweatshirt’s ‘Chum’ at 00:16
(2013). The pair recount the following story regarding their search for this
sample in an interview with LRG clothing company:
We were fascinated with this song from the 90s—this Wutang hip hop song. It
took us 10 years to find the song [they sampled originally], and when we finally
found the song we liked this scream in there. We were like, “Yo, we gotta find
that album, or that song, and get that scream.” We found the scream, and that’s
the scream in the beginning.
(quoted in LRG Clothing, 2013)
I use em sort of as a tool I guess, you know? I’m a fan of publishing, so I try not
to sample too much, but it’s a tool in your arsenal. There is something to be said
about not sampling, and if you say “I’m not going to take that from someone
else,” I say “great, congratulations—I’m happy for you.” But, you can sample,
and what if a sample makes the song better? What I try to do is grab something,
take a loop out of it, then flip it, reverse it, distort it, or chop it up. Then I ask
myself “Did I create something?”
(quoted in Pensado’s Place, 2014)
Noise Removal
Sampling originally employed the use of vinyl discs, and although the
practice of so-called crate digging for inspiring recordings is alive and
well today, producers and artists may also choose to sample from tape, CD
or mp3. While digital media tends to sound full and clean, other types of
media may be bandwidth-limited or contain audible hiss and crackle. Hip
hop tracks may sample from either of these sources and can thus suffer
from a multitude of noise issues. Some hiss and crackle can be consid-
ered acceptable, and may even be masked by the other instrumentation in
a full mix. Kendrick Lamar’s ‘Rigamortis’ (2011), for example, features
this type of vinyl crackle throughout the song. The presence of this type
of noise is not appropriate, however, for every situation, and occasion-
ally noise-removal operations will have to be performed on the sample in
question.
For hiss and crackle, simply filtering out high-end frequency content can
fix issues at times. This solution, however, will also cause the sample to
sound dull in comparison with the original source. As such, simple filtering
may also be a poor choice for removing undesirable high-frequency noise.
In cases where filtering causes the sample to sound overly dull, instead opt
for multiband dynamics processing. Users should target the offending fre-
quency region only (e.g., above 3 kHz–5 kHz) with a fast-acting expansion
tool. Ratio and threshold can be tweaked to reduce any offending noise
contained within the region, but should only be increased insofar as the
source material requires. The advantage to using multiband expansion in
this way is that the end result should be much more transparent sounding
than simple filtering.
Another type of noise that occurs during the sampling process is the
accumulation of ground hum at 50/60 Hz. Noise removal via comb filter-
ing works well here, and can be performed through Tone Boosters’ TB_
HumRemover, for instance. This approach necessitates caution, however,
if musical pitches match the hum frequency (50/60 Hz). In this case, musi-
cal pitches will be attenuated along with the undesirable noise.
is visible on both tracks, and proceed to nudge the tracks forward or back-
ward on the time grid until the transients are aligned. Once alignment is
achieved, it may be desirable to emphasize the transient of either instru-
ment. To do this, simply use a sample delay to cause the desired transient
to occur first. If one wants to mask the snare drum present in a given
sample, then an additional snare instrument could be added and given a
negative delay setting (i.e., −3 m/s). This would cause the additional snare
to sound first, thus masking the transient of the snare within the sampled
source material. While spectral treatment may still be necessary, this type
of masking presents one starting point.
Hip hop features a strong focus on rhythmic events and groove. Groove
can be altered—as in real-world drumming—by causing instruments to
occur behind the beat by small (but varying) time intervals. This routinely
happens in hip hop via manipulating swing on Akai MPC style samplers.
The swing parameter determines how far behind a given sonic even occurs
from its original rhythmic value. MPC designer Roger Linn describes his
approach to the swing algorithm by stating:
The swing algorithm described by Linn has been employed within hip hop
beatmaking since the 1980s, and continues to be relevant within the DAW.
Ableton, for example, allows users to apply humanization from the so-called
groove pool to both audio and MIDI tracks. Additionally, the software offers
a large collection of MPC style grooves, which are intended to emulate
the MPC swing algorithm. If groove quantization does not provide enough
humanization, or perhaps where a different type of rhythmic approach is
desired, beatmakers can simply key in percussive events manually.
The above patch will produce a classic 808-style kick from a simple sine
wave oscillation.4 The resultant sound, however, despite sounding clean,
lacks harmonics and may be difficult to hear through a dense mix.
Sine-based kick drums, like those produced in the above patch, can
be treated to sound more prominent through boosting or adding addi-
tional overtone content. A number of strategies may be used to enhance
the overtone content of the 808. Saturation, bit-crushing and harmonic
enhancers constitute three methods for achieving this goal. The most
common of these, perhaps, is saturation. Subtle saturation can provide
slight compression and enhance the harmonic content of the 808 kick
in order to cause it to appear perceptually forward within the mix. In
178 Matt Shelvock
Other Kicks
808-style kicks are often blended with other synthesized or sampled
drums. 909 and 808 sounds are often layered, for example, and can be seen
within the free sample library released by hip hop producer Just Blaze in
2015. His sample library contains, for instance, two kick drums used on
Jay Z’s The Blueprint (2001), each consisting of a blend of synthesized
and sampled layers. In addition to synthesized 808 and 909 drum sounds,
kick drums can be sampled from other recordings or taken from sample
libraries. Just Blaze’s sample library, for example, contains such sampled
material. The ‘KICK_EASY.wav’ sample contains an audible vocal sound
left over from the source material (Camp, 2015).
Bass
Bass, while a prominent feature of both classic and contemporary hip
hop, most typically remains secondary to the kick drum. A number of
bass-producing methods may be used in hip hop production such as sam-
pling, live tracking and synthesis. In cases where samples feature prom-
inent bass lines, producers and beatmakers occasionally opt to use the
recycled bass line in the new arrangement. Often, as a consequence of
splicing and re-ordering samples, a new bass line forms from the edits
made to the original source material.
For songs where the sample does not already contain a bass line, or
where a song does not already contain a foundational sample as discussed
in section 2.1, a sampled bass line may be used. In this case, a bass line
from a sample library or previously recorded song is implemented within
Groove and the Grid 179
a given session and often receives treatment in the form of editing and
repositioning in order to establish a more original feel.
Perhaps the most common method for implementing bass within hip
hop, however, is synthesis. Sine-based bass instruments are quite com-
mon and—as with the kick drum—evolved from widespread TR 808
usage within the genre’s earlier stages. A simple sine bass can be created
in Native Instruments’ Massive, for example, by inputting the following
settings:
The above settings will produce a deep sine wave bass texture that is suit-
able for sub bass, with added distortion to generate harmonics.5 Addition-
ally, any of the mixing strategies discussed in section ‘808s and Heartbreak’
for enhancing the overtone content of the 808 kick drum may be replicated
on 808 bass tracks. Mixers and producers should be careful, however, that
the two powerful sounds do not compete for sonic territory.
Export/Bounce
The previous sections have focused on methods for mixing and selecting
samples, drums and bass. Other peripheral instrumentation of any type can
be added as well, but these sources are diverse and often play a more triv-
ial part in hip hop mixes than the sonic aspects discussed above. Once an
instrumental is completed, it is sent to an emcee for the addition of vocal
tracks. Before the file is exported, engineers may consider applying light
compression or limiting and EQ to the stereo bus to aid in delivering an
energetic sounding background track to the artist. This step is not neces-
sary, but may be appreciated by the vocalist.
Mixing Vocals
Quintessentially, hip hop is all about the relationship between the vocals and
the drums. The number one contestant with the voice is the snare. Finding a
way to make both the vocals and the snare prominent without stepping on each
other will make the rest of the mix fall nicely into place.
(Weiss, 2011)
Even when Jay’s working with another producer who has his own go-to engi-
neer, Jay takes me along to engineer his vocals. This is to do with my knowl-
edge of his way of working and the comfort and trust factor between us. We do
more than just recording: he also bounces off ideas with me, so there’s a kind
of synergy about the records we create together.
(quoted in Tingen, 2009)
This anecdote demonstrates that emcees may even employ the same engi-
neer repeatedly once an artist develops a level of trust in such an individ-
ual. Given the importance of vocal sound to both mix engineers and the
overall aesthetic of hip hop music, the following section will discuss strat-
egies for maintaining vocal intelligibility within hip hop mixes.
Groove and the Grid 181
Lead Vocal
When mixing lead vocals, there are few (if any) universally constant meth-
ods used by engineers. A variety of approaches for mixing vocals exist for
reasons such as physiological differences in human voices, differences in
tracking methods and differences in aesthetic goals from project to proj-
ect. Difficulty arises when one attempts to prescribe a best fit approach
for mixing vocals, particularly where specific numeric descriptions are
involved. There are, however, a number of general aesthetic tendencies of
hip hop vocals that can be described. The following sections will proceed
182 Matt Shelvock
Consistency
A key factor in maintaining vocal intelligibility throughout a given vocal
track is the overall consistency of amplitude. While skilled vocalists may
be conscientious of dynamic range throughout the tracking process, addi-
tional processing is often required to achieve a more consistent amplitude
level for the duration of a vocal track. Engineers can automate volume
manually or use a plugin such as Waves’ Vocal Rider in order to tame over-
all amplitude, or macrodynamic, issues within the vocal. Where slightly
more invasive peak taming is required, compression may also present a
good choice. A compressor with medium attack and release times, and a
light ratio and threshold, should address such peaks without providing an
abundance of timbral coloration.
In addition to macrodynamic shaping, a compressor can be used with
a microdynamic strategy in mind. Where the macrodynamics of a track
refers to its overall amplitude scheme throughout its duration, micrody-
namics refers to the dynamic characteristics of individual events within
the track. For instance, hard consonants at the beginning of words can
establish small amplitude spikes within individual words or phrases, and
in so doing create a microdynamic imbalance between the initial spike and
any proceeding lyrics. If problems such as this arise, a potential solution is
to apply compression with a light ratio and fast attack-and-release settings.
and reverb may be used. While reverb is the so-called ‘kiss of death’ for
rap lead vocals according to Jaycen Joshua, subtle amounts of reverb on
background vocals can be acceptable and will cause them to sound more
distant (Tingen, 2010). Both low- and high-pass filtering may also be used:
high-pass filtering will excise redundant low-end frequencies that already
exist in the lead vocal, and low-pass filtering will remove some excitement
and air from the background vocals. By removing high-end frequency
content and adding gentle reverb, engineers can effectively dull the back-
ground vocals, and will thus allow the lead vocal to maintain prominence.
Additionally, panning is used to remove these peripheral vocals away from
the center of the stereo image. Upon listening to ‘Still Dre’, for instance,
listeners will hear that emphasis dubs tend to be panned towards the left
perhaps 25% to 40%.
Conclusion
Notes
1 Demonstrations of this process can be seen on many episodes of Mass Appeal’s Rhythm
Roulette—a YouTube series that features celebrated producers and beatmakers such as 9th
Wonder, El-P and Kirk Knight. The show can be found at https://www.youtube.com/user/
massappeal (Accessed January 2016).
2 Different strategies may be used depending on both personal taste and the envelope of a
given sound. Beatmakers may deem it more appropriate to edit samples at the zero cross-
ing, or closer to the peak of the transient.
186 Matt Shelvock
3 Such irregularities are too precise to represent with standard notation. For example, occa-
sionally drummers may offset a snare drum intended to occur on beats two and four by a
few milliseconds. The drum in this case is still notated to occur on beats two and four, even
though the resultant groove has been effectively altered. This type of drumming can be
heard on Stax records and is a key feature of music played by Booker T and the M.G.’s, for
instance.
4 This patch can be replicated on other synthesizers. To aid with this process, an explanation
of a few Massive-specific options may be helpful. The “sin-square” wavetable option pro-
vides both sine and square wave input to oscillator one, and the “WT Position” set to 0%
ensures that only the sine wave can be heard. Additionally, Massive features four envelope
filters. Envelope filters 1–3 can be assigned to modulate other parameters within Massive
(but remain otherwise inactive), and Envelope 4 is a universal filter that all oscillators are
connected to by default. The 808 patch provided contains a discrete fundamental pitch that
can be altered by keying in different notes. If users want to hear more of the fundamental,
simply turn up the decay on envelope filter 4.
5 Where more punch is required, a duplicate version of this track can be added. This dupli-
cate should alter incoming MIDI data to sound an octave higher than the original, and
should also be filtered to eliminate spectral conflicts or masking with the initial bass track.
6 From A Tinkling Piano in the Next Apartment (1965) by Merv Griffin.
Bibliography
Ahmed, I. (2014). How to Make It in America: TDE’s MixedByAli on Becoming an Audio
Engineer. Complex CA. [online]. Available at: http://ca.complex.com/music/2014/05/
mixedbyali-tde-how-to-be-studio-engineer/ [Accessed: 18 January 2016].
Camp, Z. (2015). Just Blaze Releases Early ’00s Drum Samples, Making Good on Meow
the Jewels Promise. Pitchfork. [online]. Available at: http://pitchfork.com/news/62149-
just-blaze-releases-early-00s-drum-samples-making-good-on-meow-the-jewels-promise/
[Accessed: 18 January 2016].
Chang, J. (2005). Can’t Stop, Won’t Stop. New York: St. Martin’s Press.
Dorris, J. (2013). How Beats by Dre knocked Out Better Headphones. The Age. [online].
Available at: http://www.theage.com.au/digital-life/digital-life-news/how-beats-by-dre-
knocked-out-better-headphones-20130913–2tola.html [Accessed: 18 January 2016].
Fricke, J. and Ahearn, C. (2002). Yes Y’all. Cambridge, MA: Da Capo Press.
Hager, S. (1984). Hip Hop. New York: St. Martin’s Press.
Hodgson, J. (2011). Lateral Dynamics Processing in Experimental Hip Hop: Flying Lotus,
Madlib, Oh No, J-Dilla and Prefuse 73. Journal on the Art of Record Production 5. [online].
Available at: http://arpjournal.com/lateral-dynamics-processing-in-experimental-hip-hop-
flying-lotus-madlib-oh-no-j-dilla-and-prefuse-73/ [Accessed: 18 January 2016].
iZotope.com. (2016). iZotope Interview: Just Blaze. [online]. Available at: https://www.iZo
tope.com/en/community/artists/engineers/just-blaze [Accessed: 18 January 2016].
LRG Clothing. (2013). Off the Record: Christian Rich. [online]. Available at: https://www.
youtube.com/watch?v=c4nyHF23gnI [Accessed: 18 January 2016].
Neate, R. (2013). Dr Dre Beats Valued at More Than $1bn Following Carlyle Deal. The
Guardian. [online]. Available at: http://www.theguardian.com/music/2013/sep/27/dr-dre-
beats-1bn-carlyle-sale [Accessed: 18 January 2016].
Pensado, D. (2014). Noah ’40’ Shebib—Pensado’s Place #151. YouTube. [online]. Available at:
https://www.youtube.com/watch?v=ESUHhXgIaos&index=19&list=PLh0d_FQ4KrAV
F5LM2sCC_nncdjk7jCK7G [Accessed: 18 January 2016].
Rolling Stone. (2013). Eminem Reborn: Inside the New Issue of Rolling Stone. [online].
Available at: http://www.rollingstone.com/music/news/eminem-reborn-inside-the-new-
issue-of-rolling-stone-20131120 [Accessed: 18 January 2016].
Scarth, G. (2013). Roger Linn on Swing, Groove & the Magic of the MPC’s Timing—Attack
Magazine. Attack Magazine. [online]. Available at: https://www.attackmagazine.com/fea
tures/interview/roger-linn-swing-groove-magic-mpc-timing/ [Accessed: 18 January 2016].
Groove and the Grid 187
Sencio, B. (2013). “Student of Hip-Hop” A$AP Rocky Talks His Unique Musical Style.
Hot937.cbslocal.com. [online] Available at: http://hot937.cbslocal.com/2013/03/18/stu
dent-of-hip-hop-asap-rocky-talks-his-unique-musical-style/ [Accessed: 18 January 2016].
Senior, M. (2010). Mix Rescue: Preslin Davis. Soundonsound.com. [online]. Available at:
http://www.soundonsound.com/sos/feb10/articles/mixrescue_0210.htm [Accessed: 18
January 2016].
Senior, M. (2011). How to Build Tracks around Sampled Tunes. Soundonsound.com. [online].
Available at: https://www.soundonsound.com/sos/dec11/articles/steal-the-feel.htm [Accessed:
18 January 2016].
Tingen, P. (2007). DJ Premier. Soundonsound.com. [online]. Available at: http://www.sound
onsound.com/sos/jul07/articles/djpremier.htm [Accessed: 18 January 2016].
Tingen, P. (2009). Secrets of the Mix Engineers: Young Guru. Soundonsound.com. [online]. Avail-
able at: http://www.soundonsound.com/people/secrets-mix-engineers-young-guru [Accessed:
18 January 2016].
Tingen, P. (2010). Secrets of the Mix Engineers: Jaycen Joshua. Soundonsound.com. [online].
Available at: http://www.soundonsound.com/sos/aug10/articles/it-0810.htm [Accessed: 18
January 2016].
Titze, I. (1994). Principles of Voice Production. Englewood Cliffs, NJ: Prentice Hall.
Weiss, M. (2011). Tips for Mixing Rap Vocals. The Pro Audio Files. [online]. Available at:
http://theproaudiofiles.com/mixing-rap-vocals/ [Accessed 18 January 2016].
12
Upon presentation of the mix, the one-is-not is made apparent. In our con-
tinuing exploration, as producers and academics, we will expand our dis-
course of the conceptual group The And and their singular composition
‘The Song of a Thousand Songs’. We will take a new line of flight in
discussion of the mix, which suffers similar ontological and existential
concerns, but we will also keep our feet on the ground and explore a song
that we have been involved in at a gritty level. Our aim, if we have one, is
to show how different perspectives on the mix as multiple reveal differ-
ent authenticities, different values and how these are constructed. Most of
all, we just want to play between one and many, between idea and audio,
between heaven and earth.
As our starting points we use Alain Badiou’s exploration of ‘the one-is-
not’ as it is discussed in the opening meditations of Being and Event ([1988]
2007, hereafter abbreviated BE), although we will leave aside the mathe-
matical set theory used to establish his arguments and borrow instead some
approaches to considering the multiple that will be of use to us. Second, we
make reference to the multiplicity, the one and the many, as it is thought
conceptually in Gilles Deleuze and Felix Guattari’s A Thousand Plateaus:
Capitalism and Schizophrenia ([1980] 1987, hereafter abbreviated ATP).
We can, if it helps, take a step back and simplify our concern for a moment
in order to make our starting point clear, while asking that the simplification
not be taken as a fitting reduction that expresses all our discussion.
The simplification is this: the is singular, mix is multiple, the mix then is
singular multiple. The definite and totalizing ‘the’ performs the act of sin-
gularizing (the making of the singularity) upon a concept that is multiple
(mix), for mix designates the former presence, the constitution, of more
than one. The mix refers to a past, a history, a former state (its multiplicity,
its state of being multiple) when the mixing was being done. At some point
the verb thickened, set and became the noun (did we allow the verb to set
or did we just run out of time and let it congeal?).
188
The Mix Is. The Mix Is Not. 189
Mix differs to one. Mix already concedes to a whole, a sum of its parts,
which is not the same as one. Its not-oneness is a predicate as how can it
be a mix of just one thing? To be pedantic, we would simply refer to that
operation as being ‘more of one thing’ rather than a mix. We might stop
there and say that our investigation is over; the one might be fought over
with regards to the existence of a constitution, but the mix never claimed
to be anything other than a result of plurality of some type, but we wish to
consider the mix through its coexistent one-and-many multiplicity and its
apparently paradoxical singular-multiple. The Deleuzian multiplicity, the
rhizome, is one and many, the binary one or many is not in operation here
and we have plumped (for function and as well out of pretentiousness)
for the word ‘themix’ to capture this. For Badiou, that one is multiple is
a starting point taken from Plato’s Parmenides (1931, hereafter abbrevi-
ated P) that includes a dialogue on ‘the one is not’ but then this one-is-
not is explored as different types of multiple, the inconsistent multiple
(pure multiple) and the consistent multiple (the composition of ones). On
discussion of Plato’s Parmenides, Badiou writes in the second meditation
“that in the absence of any being of the one, the multiple in-consists in the
presentation of a multiple of multiples without any foundational stopping
point” (BE 33). While the consistent multiple is the “composition of ones”
(BE 35), “the multiplicity of composition which is that of number and the
effect of structure” (BE 25). With these concepts in mind, we will speak
of these modes:
1. themix as the-one-and-the-many,
2. inconsistent mix as pure mix (the multiple without ones, pure multiple),
3. consistent mix as mix that is composed of ones.
produce the event, and these various points of recognition and observation
allow that analysis to take place to reveal the authenticities and inauthen-
ticities that dwell within mixes. No further apologies then; we have adopted
and adapted these ideas for functional reasons regardless of how far they
may have been shifted from their philosophical homes. Besides, we’re not
claiming to be in total control of what we are trying out here, but try we
will. We are about to dive from the stage unsure of whether we’ll be heroes
carried aloft by the crowd or zeros with broken legs as we hit the ground.
So far we have spoken only in general terms with regards to the mix
and to singular-multiple, and much of this might also hold true for the
crowd, the grass, the band (The And), the many, the “pigs, stars, gods . . .”
(BE 30), etc. In order to offer something more than a simple hypothetical
substitution of one with mix, we will engage at ground level among the
“mud, hair, dirt” (P 49) with the musician in the studio and their recorded
product and the relinquishing of their produced thing to the control of the
musician who will mix it (with other things) so that we keep on connecting
“and . . . and . . . and” (ATP 25). Producer number 1 (she who produces
the sound—the musician in the studio) to producer number 2 (she that
will mix) to producer number 3 (the listener that turns the treble up on the
radio). These entities are as real and as tangible as we may be able to draw
upon in order to tie the abstraction to the world of the music producer, and
we have to escape the philosophical abstractions at times and come down to
earth where abstracts can sometimes look rather awkward, like a fish out of
water. We cannot take the unveiling of the onion skin layers of the one (that
is not) down to an un-further-able Planck length, we cannot zoom in all the
way to a quark that is not a mix of anything but a true one (if indeed it is, we
don’t know, we’re not quantum physicists). But we can speak of connecting
the velocity of striking an instrument to the distance to the microphone to
the line between breakpoints on the volume automation on the computer
screen as contenders for ones or multiples that may constitute the count-as-
ones of the mix. We can speak of being lost in the music that gives inertia
to presentation which delays the count-as-ones. We can speak of the collab-
oration of the player, the mix engineer, the listener. We can speak of mixing
that thickens, sets and becomes the mix (noun = verb stasis).
Besides, the Socrates of Plato’s Parmenides is concerned to ask if the
Idea is one or many, the discussion of the one and the many of real things
is tossed aside as being so obvious as to be hardly worth considering:
If a person wanted to prove of me that I was one and many . . . he would say
that I have a right and a left side, and a front and a back, and an upper and a
lower half, for I cannot deny that I partake of the multitude. When, on the other
hand, he wants to prove that I am one, he will say, that we who are assembled
here are seven, and that I am one, and partake of the one. In both instances he
proves his case.
(P 48)
But we will indulge ourselves at ground level and talk about real things as
well as playing around with ideas up in metaphysical hyperspace, and we
The Mix Is. The Mix Is Not. 191
will consider a mix we have been involved in, and if we “cannot go beyond
the trivial statement that differences exist” (BE xii) then so be it, we are
content to dwell in the differences and see what they may reveal.
A Conceit
It is perhaps the right thing to do at this point to put forward our agenda,
our clash of interests, our conceit, in this discussion of the mix. We pro-
pose our version of the discourse of the one and the many in the form
of the conceptual song ‘The Song of a Thousand Songs’. We see all of
popular music, all songs, as being one song, we have removed all the (non-
musical) borders (composers, song titles, etc.) that force songs to separate
so that we can enjoy the ongoing symphony of the thematic development
that is popular music. All contributors to this song are part of one band, The
And, any internal differentiations such as The Rolling Stones or The The
are merely localized calibrations within the multiplicity. We are only retell-
ing an old story, one that simply agrees that rock n roll will never die. But
it had not occurred to us to consider the mix of ‘The Song of a Thousand
Songs’ or indeed the mixing engineers or the mixing of it as it is not yet,
and may never, set, although it has congealed in some places and has rather
overhardened in others. It might best be described as ‘lumpy’. It is always
becoming, it never becomes. It is always mixing; it is never the mix. The
mix is not.
We will consider our conceit later but for now we will explore our
approaches to the mix and offer observations of studio practice though a
song that the world might recognize as a song rather than the disputable
singularity of ‘The Song of a Thousand Songs’.
Just as Plato makes plain the obviousness of a person partaking in the one
and the many and quickly discards it, so we will throw away some of the
multiples that could be considered here and focus on particular arguments
to the exclusion of others. ‘The Bins’1 (Johnson, 2016) starts with a sin-
gle ukulele playing chords, a repeating pattern that makes up the phrase.
We can break this down to the chords, then each chord to the notes of the
chord, then each note to its fundamental and harmonics and its timbres,
durations, etc., that are parts of its count-as-one, but we will focus on par-
ticular events as we see fit and discard others so that we can at least shed
light on something rather than keep digging and become lost in the hole.
The ukulele in ‘The Bins’ was indeed one ukulele, and its sound to us
as listeners is of one instrument being strummed to produce the chord.
The means of producing this oneness, though, is different. When strum-
ming and recording the chords in the studio, it was evident to us that the
tuning was outside what was acceptable to us with regards to signifying
a solitary ukulele being played somewhere on a beach (‘The Bins’ is a
192 Robert Wilsmore and Christopher Johnson
song about the summer). In this case, a slight out-of-tune-ness was accept-
able, even desirable, in signifying some authenticity around the boy-on-
the-beach normality of the event, but this was not captured in our first
recordings (one ukulele strumming the chords); it was too out of tune for
our liking. In the end, we recorded each individual line on each string on
separate channels. Each channel had a pitch correction auxiliary attached
with a relatively slow response setting to allow the individual quality of
the ukulele’s character to come out before being moved to a standard pitch
set. That is, the notes have a charming out-of-tune character on the attack
but then soon become ‘in tune’. The four channels (one string recorded
on each) were then mixed together to produce the one strummed ukulele.
A number of these mixes were done with the intention of sounding as two
ukuleles being played and presenting one in the left and one in the right of
the audio spectrum. In the studio, we decided against having two ukuleles
and plumped for one in the final version.
So what of this entity, of its ontology (if it has being), of its multiplicity?
We are reduced to earthly perception at this point, to the phenomenology
of the event. Upon perceiving the event, the chord, we can say ‘there is
one’ (we can hear one chord, one ukulele), we perceive (or simultaneously
perceive) that this is multiple ‘there are pitches’ (plural). It carries the one
and the many of themix. Upon presentation, the chord is pure multiple,
nothing but mix to infinity, inconsistent mix. Upon presentation within the
count-as-one of the chord, the chord presents the notes; the count-as-one
has ones (consistent mix). What perhaps makes the mix of interest is that it
presents both its history and its present (at least with regard to its tuning, if
not to the performance of the chord which was played one note at a time).
We have written of the construction of the chord and its tuning and
have implicated ourselves in the acts of authenticity and inauthenticity
in the doing of it, so let us expose the semiotic of this further. Within the
consideration of the consistent mix of a chord, the chord is made of indi-
vidual notes (composed of ones). In this mode, we reveal the tempered
authenticity of the mix. Its authentic attack has the notes at the pitches
as they are played at the moment of striking the strings, but we are not
keen to maintain those pitches, so our pitch correction auxiliary quickly
moves them to a familiar tempered scale, which is clearly inauthentic with
regards to it no longer being true to the pitch of the instrument. Even then,
the notes are short lived; ukulele notes do not sustain for long anyway, so
the now-tempered remainder of the note has a very short existence, but it
is long enough to make a difference that can be registered by the listener.
This is the very thing of Western cultural appropriation—we have taken a
‘native’ instrument (as we might describe it in colonialist terms), the uku-
lele, full of quirk and charm, and allowed it at its surface, its instant pre-
sentation, to exhibit this beauty. But of course it is but a split second before
we drag each pitch into the straightjacket of Western pitch constraints (just
as supermarkets quickly dispose of any vegetables that are too crooked
and simply not straight enough to be sold to a public that apparently can’t
cope with wonky veg). This is, however, ‘The Bins’, a summer character-
ized by melted Tarmac, fly-ridden bins and de-icer drunk drunks, its social
The Mix Is. The Mix Is Not. 193
take the performer as the instrument and the producer as the player (she
that produces the sound from the instrument), then the criticism cannot
stand with regard to performance. We still have to deal with the erasing of
the histories, though. With regards to themix, this is problematic, because
we have disconnected from the map to produce the trace, and this trace,
the final mix of the line, asks us to accept it as a count-as-one devoid of
its ‘true’ composition (the multiplicity from which it is drawn) and thus to
recognize this presentation as pure multiple without history (inconsistent)
and as being composed by count-as-ones (consistent), but that this count is
made of ‘what is there’ and that ‘what is not there’ (i.e., the takes that are
left on the virtual cutting room floor) never existed. In real-world terms,
the line can be analyzed with regards to its notes, its phrases, it syntax, etc.,
without taking into account the map from which the trace is made. Again
this seems inauthentic, the idea that we are hiding the truth, the operation
of the presentation relies on the muting (which is the erasure) of the parts
of the takes that are not used. What then if we try to undo this monster?
In real terms, we could leave all takes to sound together, at the moment of
greatest contingency (available choices) we could reveal all lines of flight
by playing all takes together. The result would be an audible map, though
we may not be able to comprehend it as such if our brains cannot cope
with that amount of material. This result might be pleasing to the modern-
ist (and it is no surprise that A Thousand Plateaus begins with a musical
quote from a complex score by Sylvano Bussoti. Let us not be mistaken
into thinking that the postmodernist Deleuze was anything other than a
modernist when it came to music), but it makes no sense in the world of
pop in which we are producing. It would immediately be inauthentic pop
music because it does not hold true to our beliefs (the rules of our game).
The Beatles might have been proud to show off their engagement with
modernism and Stockhausen in particular at times, but that is OK, the pop
world includes the avant-garde in small doses. We might do the same, for a
short time and for effect, but it seems that our paradigm is not happy with
revealing the full extent of the map and relies on the trace for its authentic-
ity (although we will explain later with regard to ‘The Song of a Thousand
Songs’ how pop puts its traces back on to the map).
This trace, at presentation, has no history, and so we should not call it
a trace at all but perhaps give it its more usual designation of the line or
the track (in comping terms, the master track, but without its history it is
no longer the master anymore but just a plain old track). The track as pure
multiple is innocent (with no available history to condemn it); when we
notice it as consistent with regard to its composition of ones, they are also
innocent (they also have no available incriminating history either), themix
is no longer available to us (themix is not) because if it is to be an authentic
themix “the tracing should always be put back on the map” (ATP 13), but
we have erased this so it cannot be viewed as trace. Themix then is guilty
at this point because we know themix, if it is one, must have an available
map to be put back on to. The bodies may be missing, but we can still
prosecute (the ‘no body, no murder’ rule no longer applies) on the reason-
able grounds that there must have been bodies in order to produce themix
196 Robert Wilsmore and Christopher Johnson
There would seem to be a cutoff point in the mix that enables easy sep-
aration. In our song process we have recorded and mixed, at which point
we cease messing around with the parts because (having run out of time,
money or patience) we have to send the track on to the mastering process,
where it is given the audio version of steroids and made to conform to stan-
dards, so that when it is played on radio or streamed, etc., it will fit nicely.
It will be fixed, normal and hence acceptable. We are not as cynical as we
might sound about this; we are simply aware of this as a protocol rather
than a deliberate removal of difference for any political reasons (although
we are aware that this approach is also present in our industry). This proto-
col allows us to be admitted into the public domain, to become accessible,
searchable, mixable, playable to and usable by the many—without it we
could but be accessed by the few, maybe by those who wish only to fish in
places where the protocol does not apply, a place where a particular notion
of creativity has not been squeezed into a straightjacket. But that is not our
philosophy—we have accepted the protocol so that we can be in the game
and become the game.
If songs are separated largely by non-musical signifiers (the composers,
the song titles, etc.) we can “nullify endings and beginnings” (ATP 25),
for these starts and finishes are nothing more than segmentation caused
by the effect of imposing non-audio signifiers onto audio. When we do
this, we cease to operate within a representational system. There are no
longer identifiable ones of songs—the removal of the artificial beginnings
and ends has shown that they are actually all joined together. In fact, it is
not correct to say that they are joined at all, once we have removed the
sticky labels marked ‘beginning’ and ‘end’ we see underneath that there
is nothing but continuity. ‘The Song of a Thousand Songs’ is the result of
an asignifying rupture that is “against the oversignifying breaks separating
structures or cutting across a single structure” (ATP 9). It is no longer pos-
sible to remark that ‘one song is like another song’ because in the being of
the multiplicity there is no other, it is all ‘The Song of a Thousand Songs’.
Ones are not, multiplicity is. However, the pure multiple breaks down
The Mix Is. The Mix Is Not. 197
when we consider the consistent mix and themix; here we note that the
multiplicity has localized calibrations. As we have said, the mix is lumpy;
it has not set evenly and is somewhat congealed in places (which is not
consistent with Plato’s one, even a circle cannot be one because “the round
is that of which all extreme points are equidistant from the centre” (P 59),
hence the identification of center means it has parts and, although it may
be a whole, it is not one. The lumps in themix similarly show it not to be
one but many).
We should then nullify the mastering as signifying an end to the mix, for
mixing goes on well beyond this point. The mastering is only the protocol
required to connect to the map and enter into the becoming of the rhizome
of pop music. Mastering of course has many other functions, but our con-
cern here is only with its operation as an instrument for standardization.
We need to explore the mix post-mastering, and some of these mixes have
become art forms in their own right. The DJ seamlessly segues from one
track to another by mixing the ‘end’ with the ‘start’ (by tearing off the
‘sticky labels’). In this sense, the DJ is making a trace from the map, iden-
tifying and fixing a particular line of flight; although this is fleeting, he will
do it again another night perhaps but it will follow a different line, again
a fleeting moment that once gone reopens contingency. These are local
and unstable segmentations, a highlighting of one possible way through
the pop music rhizome. The mash up, the mixing of songs together, some-
times simultaneously although more frequently a DJ-like segueing of
tracks, shows us another way of highlighting themix. Detractors might
state that the mash up demonstrates the possibility of a reduction, the mere
fact that songs can be played together proves they are shallow ornaments
of an underlying truth such as a simple chord progression (this is the very
stuff of Adorno’s pseudo-individualism). But it does not have to be seen
that way; to take the reductionist viewpoint is but a preference, not a proof.
Again this is due to the imposition of separations caused by non-musical
signifiers; the songs in themselves (if they are) hold within them many
lines that accord to, for example, harmonic rules (a simple Schenkerian
analysis may quickly reduce a whole song down to a I-V-I progression),
so given the songs in the mash up are not separate at all, then they are
no different to a melody and countermelody in a Beethoven symphony
or the subject and countersubject of a Bach fugue. We are of course all
DJs, all mixers, piecing together our playlists. These playlists were once
fairly fixed; we made mix-tapes for ourselves and for each other, though
we might record over them with a new set of tracks a few weeks later. Now
our playlists constantly connect and disconnect from the map, new lines of
flight taken, or old ones highlighted, bits kept, bits replaced. We have an
immense available pool from which to draw upon courtesy of the master-
ing protocol that allowed them to be available, to become part of the map.
We are all mixers of the music, regardless of how much time was spent on
setting up the distance of instrument to microphone, the sweeping through
of frequencies (and probably reducing things at 500 Hz), the panning of
voices, etc.; once free of the studio we turn up the bass on our headphones,
the treble on our car audio and the mix continues to be unfixed. We are well
198 Robert Wilsmore and Christopher Johnson
aware of this—we mix for little headphones, for big speakers, for cars, for
radio—the continuation of the mix beyond our studio control is something
we plan for, an attempt to be as usable as possible within the multiplicity
of listening.
We have taken the inconsistent mix, the pure multiple, if it can be other
than an idea, to be the state of un-analysis in the user, the inertia that delays
the splitting of the multiple, the we have turned up the bass and, head nod-
ding, we are singing along full volume. As odd as a form of a pure multiple
might seem, it also seems to be the aim of the mixer, the producer. Do we
want to make engines for engine analysts to analyze and then admire our
greatness, to even notice that there are parts? Not really, that sounds like
a game of hide-and-seek, just like a crossword puzzle, the answers known
and then hidden for us to uncover. Not to put down crossword compilers
and solvers, it is indeed an art form of its own, but our goal is not so that
we might be analyzed and hence the pure mix, as we use the term, is our
preferred state. With the consistent mix awareness, analysis, begins to seep
in and cracks open the pure multiple. Themix is a more pragmatic, earth-
bound observation, it is how pop music works: one fabulous, continuous
symphony. We realize that in our discussions we have crossed over the bor-
ders of pretentiousness and that we have played irresponsibly, footloose and
fancy-free, with postmodernism and metaphysics in our quest to find out
what is going on in the mix. So what, it’s only rock ‘n’ roll, but we like it.
Note
1 ‘The Bins’ was performed by The And Ensemble, a subgroup of The And and York St John
University’s leading Deleuzian pop ensemble. Our thanks to band members Abigail Hall,
Joe Collins and Angus Williams for their fine vocal performances on the recording and
their ebullient antics on the video.
Bibliography
Badiou, A. (2007). Being and Event, trans. Oliver Feltham. London and New York: Continuum.
Deleuze, G. and Guattari, F. (1987). A Thousand Plateaus: Capitalism and Schizophrenia,
trans. Brian Massumi. Minneapolis, MN: University of Minnesota Press.
Johnson, C. (2016). The Bins. York: Q22 Records.
Jowett, B. (trans.) (1931). The Dialogues of Plato, Volume IV, 3rd edition. London: Oxford
University Press.
13
Mixing Metaphors
Aesthetics, Mediation and the Rhetoric
of Sound Mixing
Mark Marrington
Introduction
199
200 Mark Marrington
It is also generally accepted that these kinds of tasks take place at a late
stage in the record production process, usually further to the realization and
recording of the music, with the implication that there is a certain degree of
separation between the sound mixer and the original musical conception.
Hepworth-Sawyer and Golding (2011: 178–193), for example, posit their
“Capture-Arrangement-Performance” (CAP) model to express what should
be done prior to the mixing stage. For these authors, the often maligned
situation of “fixing it in the mix” implies shortcomings in these earlier
activities and instead, “mixing should be something that flows naturally
from the intended arrangement as set out by the producer and artist” (2011:
230).5 Izhaki (2008: 28) reiterates this point in his comment that “mixing is
largely dependent upon both the arrangement and the recordings”. Sound
mixers themselves acknowledge the importance of their independence rel-
ative to the main production process. According to Owsinski (1999: 8),
the vast majority of mixers would “prefer to start the mix by themselves”,
while acknowledging that ultimately they cannot ignore the ambitions
of the producer and artists for the mix outcome. Lipson comments that
“it’s far more useful to have the artist hear something complete . . . than
for them to hear the whole process of me getting to that point” (Massey,
Mixing Metaphors 201
2009:197). Stavrou in more forceful terms suggests that the mixer, “turf
the producer and band out of the control room” (2003: 151). Objectivity
and neutrality are also valued by sound mixers when undertaking mixing
tasks. Senior, for example, refers to the mix as “data” about which one
needs to be able to “make objective decisions . . . irrespective of your own
subjective preferences” (2011: 2).6 To facilitate this objectivity, Stavrou, in
an extreme gesture in keeping with his unique theory of mixing, has advo-
cated playing the mix backwards because it “removes the language out of
the music . . . leaving you only with shape, and harmonics galore”, creating
“a whole new psycho-acoustical playground to dive into” and forcing “an
unbiased view of the sounds” (2003: 232). The idea that the mix constitutes
a form of relatively neutral sonic data has had a particular currency with
researchers in the area of automated mixing, involving the development
of intelligent systems to facilitate the application of mixing techniques in
relation to quantifiable aesthetic outcomes. De Man and Reiss (2013), who
are at the forefront of this field, even go so far as to suggest that “mix-
ing multi-channel audio comprises many expert but non-artistic tasks that,
once accurately described, can be implemented in software or hardware”.7
It is apparent that the authors’ notion of the mixing process in this sense is
in part a response to the position held in the ‘how to’ type of texts that the
practice can be broken down into a series of engineering specialties.8
Although mixing is commonly articulated in terms of its engineering
techniques, it is rarely the case that sound mixers will remain completely
detached from the musical material they are dealing with. Indeed, even
in the purely engineering context, sound mixers will be likely at some
point to engage with the sounds they are mixing in terms of their musical
connotations, as illustrated by the common practice of relating musical
pitch to frequency values when applying EQ to tracks. ‘How to’ type of
texts directed at mix engineers and other audio professionals frequently
advocate the usefulness of musical knowledge when decoding what is hap-
pening in the mix. Owsinski (1999: 12), for example, notes that the ability
to balance the various instrumental timbres of the mix as it progresses
is facilitated by “understanding the mechanics of a well written arrange-
ment”.9 Moylan’s comprehensive pedagogical treatise, The Art of Record-
ing, throughout stresses the need for audio professionals (or recordists)
to correlate their engineering knowledge to “sound’s artistic elements (or
the meanings or message of the sound)” (2007:4), in reference to pitch,
loudness, duration, timbre and space. Case goes further, suggesting that
the sound mixer develop a broad range of musicianship skills, advising the
pursuit of “music studies, a music degree, and proficiency on more than
one instrument”, and suggesting that being a “gigging performer makes
you an even better mixer” (2011: 226). For Case, the purpose is to enable
the sound mixer to gain clearer insights into the song and therefore make
mixing decisions that are “supportive of the music” (2011: 227). It is with
such comments that we begin to move from a concern with overtly techni-
cal matters to a consideration of mixing aesthetics, a term which I use here
to refer to the broader artistic objectives of the mixing process relative to
the musical material.
202 Mark Marrington
During the late sixties and early seventies I did not want to use the overdub-
bing features of multi-track recording. If I had a five-piece group I wanted five
guys to play simultaneously and if I had a ten-piece group I wanted ten to play
simultaneously . . . Everything was done on the fly.
(Buskin, 1999: 165)
Mixing in essence was both the start and the end-point of production,
making its mark at the moment of recording itself. The remit of the mix
engineer was also to a certain extent informed by the musical subject mat-
ter typically being recorded, which was either fully worked out prior to
the sessions (arrangements of light music or classical music, for example)
or dependent upon the live performance situation for its realization (for
example, improvised jazz or popular music honed in a live band context).11
The music thus possessed its own integrity and there was generally no
expectation that its substance would be modified further once it had been
recorded, other than to correct major errors. Certainly, the mix engineer
would not have presumed to radically manipulate the material beyond the
point of its live capture.12
What mixing tended to reflect in this early period was a particular ide-
ology of record production that Lee Brown (2000: 361) has referred to in
metaphorical terms as the “transparency perspective”. This describes a sit-
uation in which “sound recording is understood on the model of a transpar-
ent windowpane through which we can see things undistorted”, implying
both accurate representation of the music and minimal mediation.13 This
attitude still remains current in the thinking of mix engineers today and is
often expressed in terms of an aesthetic of ‘responsibility’ to the integrity
of the music. For example, Armin Steiner, an engineer who possesses a
background in pop, classical and film score recording and mixing, holds
the view that “We are not interpreters. We are servants of the music, and all
we’re doing is taking down what the composer did, and hopefully putting it
down in the perspective that he heard” (Droney, 2003: 127). Bruce Swed-
ien, queried as to his “philosophy about mixing”, has commented that
Mixing Metaphors 203
It comes from my early days in the studio with Duke Ellington and from there
to Quincy [Jones]. I think the key word in that philosophy is what I would
prefer to call responsibility . . . our first thought should be that our responsi-
bility to the musical statement that we’re going to make and to the individuals
involved.
(Owsinski, 1999: 198)
the power (what Max had called the aggression) was in the playing and sing-
ing. I tried to tone it down using reverbs, echo and EQ. The chant vocals were
placed back in the mix but I thought this just made it feel weak. Trying to stop
the mix taking its natural course was frustrating.
(Brown, 2010: 268)
204 Mark Marrington
For Brown this was ultimately an irreconcilable situation: “once the record
company gets a foot in the door on discussions about overall sound and
mixes, it’s all over” (2010: 269).
of the mix has been discussed and defined extensively in the literature,
from Moore’s “soundbox” (1993) to Moylan’s “perceived performance
environment” (2007: 263).19 What is usually being referred to is a virtual
three-dimensional space in which the elements of the mix are organized,
comprising depth (the sound’s apparent proximity to the listener), the hor-
izontal axis (placement of sounds in the stereo field) and the vertical axis
(individual frequency bands of sounds—sometimes expressed as height)
to construct the virtual ‘soundstage’ for the performance.20 A principal rhe-
torical use of spatial parameters is to mark significant musical elements of
the mix for listener consciousness, the choice of which will depend upon
the nature of the music being mixed and the sound mixer’s appreciation
of the relevant genre conventions (or codes).21 In commercial pop song
mixes, for example, EQ, panning and reverb may be used to foreground
the voice relative to the other instruments to draw attention to the pri-
mary artist, the lead melody and the lyric.22 The positioning of particular
musical elements in space can often be critical in conveying the essential
musical attributes of a given genre. In a house music track, for example,
the common practice of placing the kick and snare drums centrally has the
effect of maximizing the energy distribution of these sounds between the
two speakers, ensuring that the beat, a key rhetorical figure in dance music,
is communicated unambiguously.23 The sound mixer’s approach to space
can also help to contribute to the authenticity of the listener’s perception
of a given musical genre, in certain cases entailing the construction of a
spatial environment that did not exist during the recording process. In a
rock context, for example, where liveness and authentic musicianship are
valued, mix elements that were tracked separately might be staged in a way
that suggests they were performed together in the same space at the same
time, as well as treated sonically in a way that lends them immediacy.24 As
Moylan has observed, an important aspect of sound mixing is concerned
with creating “illusions of space” (2007: 52), and such illusions comprise
an important part of the rhetoric of recordings from the sound mixer’s
perspective. A final point to note is that genre-specific spatial configura-
tions of mixes tend to become conventionalized (or overcoded) with use,
as illustrated by Gibson’s extensive taxonomy of genre-based mix types
(2005) and Dockwray and Moore’s recent survey of ‘normative’ stereo
mixes created between 1965–72 (2010).25 While this can provide useful
consistency for the commercially inclined sound mixer, it also presents
opportunities to play upon such expectations and employ a spatial rhetoric
which is distinctive in its lack of observation of convention and becomes a
creative gesture in its own right.26
In addition to spatial concepts, sound mixers are also concerned with
the organization of the musical material in its own terms. This can entail
making adjustments to the both the structure and content of the mix to
improve the clarity and coherence of the musical ‘oratory’. Many sound
mixers, for example, consider that it is within their remit to contribute
opinions on the musical arrangement, and they may even take it upon
themselves to re-configure this during the mixing process.27 The kinds
of changes that a sound mixer might make to the recorded material can
206 Mark Marrington
several times I’ve been congratulated for “putting a new part” into a mix, when
all I’ve done is dredge up one of the client’s own buried sounds that had long
since faded from their memory. The beauty of this trick when mixing other
people’s work is that it’s low-risk, because anything in the original multitrack
files implicitly bears the client’s seal of approval, whereas there’s nothing like
that guarantee for any truly new parts you might add.
(Senior, 2011: 85)28
minor tasks that mix engineers consider are simply corrective, such as
making adjustments to the timing or tuning of audio material (‘perfect’
production is itself a rhetorical gesture), while with the use of staple tools
such as the EQ and the compressor, the mixer is potentially moving into
the territory of sound design.29 Stavrou’s position on mixing aesthetics (in
his book, Mixing with Your Mind) is notable for its emphasis on the active
role that studio devices can play in sculpting the musical vision. Indeed,
the creative use of tools arguably constitutes his definition of the art of
mixing. To this end, he devotes a great deal of discussion to re-thinking
the employment of EQ, compression and reverb in terms of their musical
possibilities. In regard to reverb, for example, Stavrou suggests that the
effect’s ability to alter time relationships in a track is more important than
what it indicates about the space the mix is in: “Reverb isn’t so much about
being true to an acoustic space, it’s more about the effect it has on the
groove”. He also suggests that in certain cases “the only purpose of some
instrument is to become a key input for an interesting reverb effect” (2003:
182).30 In other words, the mix engineer ought to be actively seeking out
material in the multitrack that would provide an opportunity to use a par-
ticular tool to embellish the music.
One final aspect of the rhetoric of mixing that I wish to consider, which
has already been touched upon in the preceding discussion, is the notion
of mixing idiolect. Idiolect is a term derived from semiotic theory, which
is usually employed to account for the specific identifying fingerprints
of individuals within areas of broadly consistent practice.31 In effect,
this implies the assertion of the sound mixer’s creative personality upon
the material of the mix and might thus be regarded as the antithesis of
the transparency perspective. Certainly in some quarters, this has been
regarded as crossing the line. Bill Bottrell, for example, has expressed a
particularly negative view of engineers who allow their own creativity to
take precedence during the mixing process:
It all started to go wrong in the late ’70s when engineers and producers started
being allowed to impose their frustrations as musicians on the records. And
that should never have happened. Because the artists know what to do, if they
are really an artist, and the producer should just set up a situation where the
artist feels free to do what they do. The engineer should just record it and get
out of the way.
(Droney, 2003: 17)
Bottrell even goes so far as to suggest that artists have been “brainwashed”
to think that it is an engineer’s prerogative to impose ideas, and that “there’s
a whole generation of artists who think that’s how records are made, and
they don’t question it” (2003: 17). It is clear that Bottrell’s view is not
widely shared, however, as evidenced by the many sound mixers who place
208 Mark Marrington
Hannett’s equalization cuts the brunt of Sumner’s fuller live sound down to an
echoing squeal. In search of vocal clarity and space for delay and reverb to ring
out, Hannett relegates the guitar to hard-panned stereo placement in later tracks
and thins the robust double-humbucker sound of Sumner’s Gibson SG.
(Ott, 2010: 64)
We played the album live loud and heavy. We felt that Martin toned it down,
especially the guitars. The production inflicted his dark, doomy mood over the
album; we’d drawn this picture in black and white, and Martin coloured it in
for us. We resented it, but Rob [Gretton] loved it, [Tony] Wilson loved it, the
press loved it, and the public loved it. We were just the poor, stupid musicians
who wrote it.
(quoted in Savage, 1994)
Tankel (1990), in reference to the practice of remixing, has used the word
“recoding” to describe what takes place when remixers re-imagine the
materials of an existing studio multitrack to discover new perspectives on
the original recording. ‘Recoding’ would also seem to be an apt word for
what took place in Hannett’s case, especially given the subsequent conflict
between the band’s sense of its ‘live’ identity and what was created in the
studio.36 Remixing, for the purposes of the present discussion, can usefully
be regarded as a subbranch of mixing practice in which the sound mixer is
given carte blanche to be creative. In other words, there is no question here
that idiolect should be sought. What the practice also highlights is the fact
that the mix is not a fixed entity, or a definitive statement, and that there
can be many legitimate opinions on what the best approach should be.
In this sense, it is a thought-provoking question as to whether the record
would have been as successful had the basic recorded material been mixed
in a way that simply clarified its essence (“in black and white”) rather than
elaborated it (“in colour”). A final point worth adding is that in the wider
context of record production, Hannett is often cited as an architect of the
1980s’ Manchester post-punk sound.37 This suggests that his idiolect was
potentially more far-reaching than his re-molding of Joy Division, serv-
ing to define (or re-define) a domain of practice within that scene more
generally.38
210 Mark Marrington
Conclusion
This chapter has considered some of the ways in which sound mixers
can be implicated in the communication of the message of the record-
ings they mix. I have suggested that although on one level mixing can
be regarded as a technically oriented engineering activity whose pur-
pose is to put the best sonic face on a given musical recording, it cannot
avoid having a bearing upon the articulation of the artistic elements
therein. This is because sound mixers are caught up in particular ideol-
ogies of record production that define their mixing aesthetics and deter-
mine how they mediate the material. Sound mixers must make a variety
of decisions in accordance with these aesthetics, whether in the service
of achieving a transparent rendering of the music, responding to a com-
mercial remit or searching for the best means of expressing a unique
artistic vision. The sound mixer also has a rhetorical function relative
to the recorded material, because they essentially speak on behalf of it.
This requires an ability to interpret the material in reference to the spe-
cific rhetorical devices expected by the audience in a particular musical
context.39 As Gibson has commented, “the mix should be appropriate.
Appropriate for the style of music, appropriate for the song and all of
its details, and appropriate for the people involved” (2005: 48). To this
end, conversance with a wide range of record production approaches is
advocated by sound mixers, as Case (2011: 226) comments, “We need
to know our history—enjoy and study the most important recordings
in our chosen styles of music . . . avidly seek out and analyze the most
popular contemporary recordings”. Indeed, it is this conversance with a
broad range of record production strategies that potentially determines
the scope of what a sound mixer’s creative contribution might be to a
mix. Thus, those sound mixers who adhere closely to specific genre
contexts might produce conservative but nonetheless commercially
viable mixes, while those who are open to drawing intertextually from
across different genres are more likely to be innovative.40 Ultimately,
it is an eclectic approach that enables the ‘art’ of mixing to progress,
because it admits of elements that are likely to re-configure the param-
eters of a given mixing domain.41 This latter point I have illustrated
in terms of idiolect, whereby the sound mixer may place a stamp of
individual identity on the musical material that has little to do with
the artist’s intentions or commercial imperative. Finally, it must not be
forgotten that the intended target of this rhetorical activity is the audi-
ence, whose interpretation of the message is not necessarily predictable.
Thus, sound mixers must also take pains to anticipate the general effect
of their work here, as illustrated by the common practice of gauging
the impact of completed mixes through playback via typical consumer
audio systems, such as car stereos, mobile phones and laptop speakers.
If the sound mixer can secure the audience’s accurate perception of the
mix in acoustic terms, they may stand a good chance of ensuring the
accurate reception of its message in artistic terms.
Mixing Metaphors 211
Notes
1 I use the term ‘sound mixer’, derived from Kealy (1979), throughout this discussion
to refer to an individual who mixes the multitrack at a point post the initial production
stage. This is not to discount the fact that mixing can also take place during the pro-
duction process itself. However, as will become apparent, my discussion is founded on
the idea that sound mixing takes place in reference to an already defined vision of the
production.
2 In addition to the general discussions of rhetoric as persuasive speech that can be found
in classic modern treatises, such as Corbett (1998), my thinking on rhetoric in relation to
record production and popular music studies is also informed by discussions in Brackett
(2000), Zak (2001) and Toft (2010).
3 My use of the words ‘code’, ‘idiolect’ and ‘message’ also implies semiotic theory, of
which rhetoric is considered to be an adjunct. See, for example, Eco (1976: 276–288) and
Tarasti (2012: 271–300).
4 Approaches to discussing the subject in the practitioner literature are also reflected in the
columns of the trade periodicals (Sound on Sound, Computer Music, Tape Op, Mix, etc.).
Supplementing the practitioner guides are also a number of interview compilations—
Schwartz and Stone (1992), Buskin (1999/2012), Droney (2003), Massey Vols. 1 and 2
(2000/2009)—which frequently draw out references to mixing practice, as well as some use-
ful autobiographical accounts by mixing engineers and producers including Phil Ramone’s
Making Records (2007) and Phill Brown’s Are We Still Rolling? (2010). The other major
source of contribution to the discussion has come from academics working in the “musi-
cology of production” field, including Moore (1993/2001), Zak (2001), Doyle (2005) and
Zagorski-Thomas (2014). Typically, writing in this area has been concerned with notions
of ‘staging’ and the classification of mixing strategies relative to era and particular musical
genre.
5 See the discussion of fixing it in the mix in Hepworth-Sawyer and Golding (2011: 227–
28). ‘Fixing it in the mix’ is an ideological position that is becoming more acceptable,
particularly in the context of the DAW where compositional and arrangement decisions
can be postponed well into the mixing stage.
6 Indeed, “objectivity” is a word which recurs frequently during Senior’s discussions of
mixing.
7 Italics are my emphasis.
8 De Man and Reiss’s provocative conjecture that mixing involves “non-artistic” tasks is
particularly debatable given that even the most minor adjustments to the sound may have
the potential to impact upon the music’s message.
9 Gibson (1997: 4) notes that “The recording engineer is quite commonly the most knowl-
edgeable person in the studio when it comes to being aware of all the types of musical
instruments and sounds available.”
10 Indeed, many mix engineers still refer to mixing in these terms. The subheading of Stav-
rou’s book, Mixing with Your Mind (2003), for example, states that the author is offering
“Closely guarded secrets of sound balance engineering”. See Hodgson (2010: 149–157)
and Owsinski (1999: 2–6) for further discussion of the historical evolution of mixing and
the mix engineer.
11 For a detailed discussion of the ontology of recorded music, see Gracyk (1996).
12 This is not to imply that the recorded material was not re-edited extensively during
the production process in this era, as occurred even in the case of classical music. The
purpose, in most instances, was nonetheless to achieve a close correlation to the live
rendition.
13 See also Moorefield’s (2010) notion of the “illusion of reality”.
14 See Kealy’s classic essay ‘From Craft to Art’ (1979) for a full assessment of the devel-
opments that occurred this period. Moorefield’s The Producer as Composer (2010) also
contains relevant discussion.
15 It is worth noting that much of Levine’s contribution to the mixing process occurred
during the recording stage (i.e., in the live capture mode).
212 Mark Marrington
16 Indeed, it is from this point that production and mixing tend to become more synonymous
as producers explored mixing, and sound mixers explored production. This perspective
is accommodated in mixing theory. Moylan’s use of the word “recordist”, for example,
describes the conflation of recording producer and engineer—in other words, an individ-
ual who mixes music that they have also played a key role in shaping during the recording
process. Kealy (1979) has used the term “artist-mixer”.
17 Zak (2001: 141), for example, states that the object of mixing is to create “a composite
image of an apparently unitary musical performance”. Moylan (2007: 319) suggests that
“the actual process of executing the mix is very similar to performing”. Golding and
Hepworth-Sawyer (2012: 182) have notably compared the role of the modern mix engi-
neer to that of an orchestral conductor, presumably because sound mixers, like conduc-
tors, are essentially dealing with a type of blueprint for a potential musical performance.
18 Musicologists have used words such as “montage” (Zak, 2001: 141) and “sonic narrative”
(Liu-Rosenbaum, 2012) to account for the ways in which sound mixers may be thinking
when considering the mix in terms of its temporal evolution.
19 Albin Zak (2001:141–160) has also outlined a spatial theory of mixing based upon
George Massenburg’s “four dimensional space model”.
20 I borrow this term from Moylan (2007: 328).
21 It is apt, where rhetorical ideas are concerned, that Hodgson (2010) refers to mixing as
“the space of communications”.
22 There is an interesting discussion in Senior (2011: 122) concerning what is sacrificed or
reduced in the service of foregrounding particular elements that suit the mix style.
23 For a thoughtful account of the rhetoric of dance music, see Hawkins (2003). Snoman
(2014) also provides many useful insights into dance music mixing.
24 As Moylan (2007: 195) notes, “the characteristics of this envisioned performance envi-
ronment will greatly influence the conceptual setting for the artistic message of the work.”
25 For a particularly interesting discussion of the spatial strategies of a particular era of
record production, see also Doyle (2005).
26 For suggestions as to what this flouting of convention might entail, see Dockwray and
Moore (2010) and Liu-Rosenbaum (2012).
27 Gibson (1997: 1) notes that “even when there is a producer, he or she will rely heavily on
the values and critiques of the engineer. In fact, groups often go to major studios solely
because of the production assistance they get from professional recording engineers.”
28 Again, the transparency perspective is implicit in the sense that mixing here is about
locating some essence of the artistic vision that needs to be accurately perceived and
faithfully transmitted. Stavrou (2003: 175–176) also echoes Senior’s idea when he recom-
mends “searching the tracks for the most inspiring player”.
29 An EQ, for example, is potentially acting like a synthesizer filter and can certainly remove
a sound a considerable distance from its natural state, while the compressor can be used to
substantially alter a sound’s envelope and rhythmic characteristics. For further discussion
of sound shaping aspects of recording practice, see Moylan (2007: 46–47).
30 Stavrou’s discussion of reverb (2003: 181–189) contains many interesting observations
regarding its creative use.
31 For further contextualization of “idiolect” and examples of its use in analysis see Eco
(1976), Middleton (1990: 174) and Ibrahim and Moore (2004: 139–158).
32 This includes their seminal album, Unknown Pleasures (1979) and its follow-up, Closer
(1980), as well as their early work under the name of New Order on the album Movement
(1981).
33 Discussion of Hannett’s background can be found in Sharp (2007).
34 While Hannett was assisted by Chris Nagle (engineer at Strawberry Studios, Stockport)
during a number of Joy Division sessions, it is clear that Hannett regarded himself as the
one doing the mixing.
35 For accounts of Hannett’s recording approach by members of Joy Division, see Reynolds
(2009: 229–243) and Hook (2013: 150–159).
36 To hear the difference, it is instructive to listen to the band’s early recordings under
the name of Warsaw, particularly their 1978 version of ‘Transmission,’ which Hannett
re-recorded and mixed in 1979.
Mixing Metaphors 213
37 See, for example, the sleeve notes for the retrospective single collection, Zero: A Martin
Hannett Story 1977–1991.
38 Reynolds (2005: 187) notes, with great insight, that “Hannett believed punk was sonically
conservative precisely because of its refusal to exploit the recording studio’s capacity to
create space.” For discussion of Hannett’s contribution in the general context of post-
punk, see also Witts (2009).
39 Hepworth-Sawyer and Golding’s concept of “Production Plotting” (2011: 177) is also of
interest in this regard.
40 Albin Zak’s discussion of “resonance” in The Poetics of Rock (2001: 184–197) provides
much food for thought here.
41 I use “domain” here in the Csikszentmihalyi sense of the term. Summarized broadly,
Csikszentmihalyi (1988: 325–339) considers creativity relative to the particular environ-
ment within which the individual operates. He uses the term “domain” to refer to an
existing context of practice from which one assimilates patterns of creative approach
(the rules of the game as it were) and “field” to refer to the social factors (namely people
and institutions) which determine those creative contributions that are most likely to be
accepted into the domain.
Bibliography
Bonds, M.E. (1991). Wordless Rhetoric: Musical Form and the Metaphor of the Oration.
Cambridge, MA: Harvard University Press.
Brackett, D. (2000). Interpreting Popular Music. Berkeley and Los Angeles, CA: University
of California Press.
Brown, L.B. (2000). Phonography, Rock Records, and the Ontology of Recorded Music. The
Journal of Aesthetics and Art Criticism 58 (4): 361–372.
Brown, P. (2010). Are We Still Rolling?: Studios, Drugs and Rock ‘n’ Roll—One Man’s Jour-
ney Recording Classic Albums. n.p.: Tape Op Books.
Buskin, R. (1999). Inside Tracks. New York: Avon.
Buskin, R. (2012). Classic Tracks: The Real Stories Behind 68 Seminal Recordings. London:
Sample Magic.
Case, A.U. (2011). Mix Smart: Pro Audio Tips for Your Multitrack Mix. Oxford: Focal Press.
Corbett, E.P.J. and Connors, R.J. (1998). Classical Rhetoric for the Modern Student. Oxford:
Oxford University Press.
Csikszentmihalyi, M. (1988). ‘Society, Culture, and Person: A Systems View of Creativity.’
In The Nature of Creativity, ed. R.J. Sternberg. New York: Cambridge University Press,
pp. 325–339.
Day, T. (2002). A Century of Recorded Music: Listening to Musical History. New Haven, CT:
Yale University Press.
De Man, B. and Reiss, J.D. (2013). A Semantic Approach to Autonomous Mixing. Journal
on the Art of Record Production 8. [Online]. Available at: http://arpjournal.com/a-seman
tic-approach-to-autonomous-mixing/ [Accessed: 5 January 2016].
Dockwray, R. and Moore, A.F. (2008). The Establishment of the Virtual Performance Space
in Rock. Twentieth-Century Music 5 (2): 219–241.
Dockwray, R. and Moore, A.F. (2010). Configuring the Sound-Box 1965–1972. Popular
Music 29 (2): 181–197.
Doyle, P. (2005). Echo and Reverb: Fabricating Space in Popular Music Recording, 1900–
1960. Middletown, CT: Wesleyan University Press.
Doyle, T. (2013). Peter Walsh and Scott Walker: Producing Bish Bosch. Sound on Sound.
[Online]. Available at: http://www.soundonsound.com/sos/jan13/articles/scott-walker.
htm.
Droney, M. (2003). Mix Masters: Platinum Engineers Reveal Their Secrets for Success. Bos-
ton, MA: Berklee Press.
Eco, E. (1976). A Theory of Semiotics. Bloomington, IN: Indiana University Press.
Eisenberg, E. (2005). The Recording Angel: Music, Records and Culture from Aristotle to
Zappa. New Haven, CT, and London: Yale University Press.
214 Mark Marrington
Gibson, D. (1997). The Art of Mixing: A Visual Guide to Recording, Engineering, and Produc-
tion. Vallejo, CA: MixBooks.
Gibson, D. (2005). The Art of Mixing: A Visual Guide to Recording, Engineering, and Produc-
tion. Boston, MA: Thomson Course Technology.
Gracyk, T. (1996). Rhythm and Noise: An Aesthetics of Rock. Durham, NC: Duke University
Press.
Hawkins, S. (2003). ‘Feel the Beat Come Down: House Music as Rhetoric.’ In Analyzing Pop-
ular Music, ed. A.F. Moore. Cambridge: Cambridge University Press, pp. 80–102.
Hepworth-Sawyer, R. and Golding, C. (2011). What Is Music Production? Professional Tech-
niques to Make a Good Recording Great. Oxford: Focal Press.
Hodgson, J. (2010). Understanding Records: A Field Guide to Recording Practice. New York:
Continuum Books.
Hook, P. (2013). Unknown Pleasures: Inside Joy Division. London: Simon and Schuster.
Ibrahim, A. and Moore, A.F. (2004). ‘Sounds Like Teen Spirit: Identifying Radiohead’s Idio-
lect.’ In The Music and Art of Radiohead, ed. J. Tate. Farnham: Ashgate, pp. 139–158.
Izhaki, R. (2008). Mixing Audio: Concepts, Practices and Tools. Oxford: Focal Press.
Jones, S. (1992). Rock Formation: Music, Technology, and Mass Communication. London:
Sage Publications.
Kealy, E.R. (1979). From Craft to Art: The Case of Sound Mixers and Popular Music. Work
and Occupations 6 (1): 3–29.
Liu-Rosenbaum, A. (2012). The Meaning in the Mix: Tracing a Sonic Narrative in ‘When the
Levee Breaks’. Journal on the Art of Record Production 7. [Online]. Available at: http://
arpjournal.com/the-meaning-in-the-mix-tracing-a-sonic-narrative-in-‘when-the-levee-
breaks’ [Accessed: 5 January 2016].
Massey, H. (2000). Behind the Glass Vol. 1: Top Record Producers Tell how They Craft the
Hits. San Francisco, CA: Backbeat Books.
Massey, H. (2009). Behind the Glass Vol. 2: Top Record Producers Tell how They Craft the
Hits. Milwaukee, WI: Backbeat Books.
Middleton, R. (1990). Studying Popular Music. Buckingham: Open University Press.
Moore, A.F. (1993). Rock: The Primary Text. Milton Keynes: Open University Press.
Moore, A.F. (2001). Rock: The Primary Text. Aldershot: Ashgate.
Moorefield, V. (2010). The Producer as Composer: Shaping the Sounds of Popular Music.
Cambridge, MA: MIT Press.
Moylan, W. (2007). Understanding and Crafting the Mix: The Art of Recording, Second edi-
tion. Oxford: Focal Press.
Ott, C. (2010). Joy Division’s Unknown Pleasures. New York: Continuum Books.
Owsinski, B. (1999). The Mixing Engineer’s Handbook. Vallejo, CA: Mix Books.
Ramone, P. (2007). Making Records: The Scenes Behind the Music. New York: Hyperion.
Reynolds, S. (2005). Rip it Up and Start Again: Postpunk 1978–1984. London: Faber & Faber.
Reynolds, S. (2009). Totally Wired: Postpunk Interviews and Overviews. London: Faber &
Faber.
Savage, J. (1992). An Interview with Martin Hannett, 29th May 1989. Touch-Vagabond.
[Online]. Available at: http://www.rocksbackpages.com/Library/Article/an-interview-
with-martin-hannett-29th-may-1989 [Accessed: 20 November 2015].
Savage, J. (1994). Joy Division: Someone Take These Dreams Away. Mojo. [Online]. Avail-
able from: http://www.rocksbackpages.com/Library/Article/joy-division-someone-take-
these-dreams-away [Accessed: 20 November 2015].
Schwartz, D. and Stone, T., eds. (1992). Music Producers: Conversations and Interviews with
Some of Today’s Top Record Makers. London: Omnibus Press.
Senior, M. (2011). Mixing Secrets for the Small Studio. Oxford: Focal Press.
Sharp, C. (2007). Who Killed Martin Hannett?: The Story of Factory Records’ Musical Magi-
cian. London: Aurum.
Snoman, R. (2014). Dance Music Manual: Tools, Toys, and Techniques. Burlington, MA:
Focal Press.
Stavrou, M.P. (2003). Mixing with Your Mind: Closely Guarded Secrets of Sound Balance
Engineering. Mosman, New South Wales: Flux Research.
Mixing Metaphors 215
Tankel, J.D. (1990). The Practice of Recording Music: Remixing as Recoding. Journal of
Communication 40 (3): 34–46.
Tarasti, E. (2012). Semiotics of Classical Music: How Mozart, Brahms and Wagner Talk to Us.
New York: Walter de Gruyter.
Toft, R. (2010). Hits and Misses: Crafting Top 40 Singles, 1963–1971. New York: Continuum
Books.
Witts, R. (2009). ‘Records and Recordings in Post-Punk England, 1978–80.’ In The Cam-
bridge Companion to Recorded Music, eds. E. Clarke, N. Cook, D. Leech-Wilkinson and
J. Rink. Cambridge: Cambridge University Press, pp. 80–83.
Zagorski-Thomas, S. (2014). The Musicology of Record Production. Cambridge: Cambridge
University Press.
Zak, A.J. (2001). The Poetics of Rock: Cutting Tracks, Making Records. Berkeley and Los
Angeles, CA: University of California Press.
Zak, A.J. (2005). ‘Edition-ing’ Rock. American Music 23 (1): 95–107.
Zak, A.J. (2009). ‘Getting Sounds: The Art of Sound Engineering.’ In The Cambridge Com-
panion to Recorded Music, eds. E. Clarke, N. Cook, D. Leech-Wilkinson and J. Rink.
Cambridge: Cambridge University Press, pp. 63–76.
14
216
Mix as Auditory Response 217
(1973/2011), the concluding track on Dark Side of the Moon, the protago-
nist has lost his mind completely.
Regardless of how one interprets the track—I’ve always heard a child-
birth from the child’s perspective, in fact—‘Speak To Me’ remains, in the
first instance, a finely detailed mix. As such, it provides us with a valu-
able tool for elucidating the aural parameters of a mix’s auditory narrative,
using soundbox analysis. I will thus turn my attention now to elucidat-
ing the track’s sonic details, knowing that it is only through these details
that the jarring narrative interpretations I mention above can in the first
instance emerge.
As already noted, the first component of a soundbox is what I call ‘The
Auditory Horizon’.10 In ‘Speak To Me’, the Auditory Horizon is estab-
lished during the first eleven seconds of the track, when the heartbeat fades
in and reaches its maximum volume. Behind that Horizon is silence. The
Auditory Horizon thus constitutes the total reach of a mix’s ‘earshot’. If
a track fades in—as with the heartbeat in ‘Speak To Me’, for instance—it
begins its trek towards the Auditory Horizon from behind it, which is to
say, from a place ‘too far away’ to be heard. Conversely, if a track fades
out—as with, for example, the heartbeat heard during ‘The Dark Side of
the Moon’ (1973/2011), the album’s final track—it ends its trek past the
Auditory Horizon, beyond ‘earshot’.
A mix’s ‘Horizontal Plane’, the next component of a soundbox, describes
where a sound is heard in relation to center, and we call the total horizontal
expanse of a mix its ‘Horizontal Span’. What I call the ‘pneumatic drill’
and ‘electronic drone’, which sound at fifty-five seconds into ‘Speak To
Me’, are panned variously throughout their brief twenty-one second exis-
tences. They oscillate between, rather than leap from, left to right positions
along the Horizontal Plane. The track that sounds farthest to the right,
however, is the loop of ticking clocks, first heard at thirty-one seconds in.
Sixteen seconds later, a male voice confesses that he’s “always been mad”
at the farthest position left along the Horizontal Plane. Together, these two
tracks create the Horizontal Span of the mix for ‘Speak To Me’.
Perhaps the most significant component of a soundbox is its Proximity
Plane. This component describes the position of sounds in a mix vis-à-vis
its Auditory Horizon. As such, the Proximity Plane represents a mix’s abil-
ity to hear in depth, with the Auditory Horizon comprising its far limit. In
‘Speak To Me’, as what I call ‘the pneumatic drill’ pans along the Horizon-
tal Plane, it is also faded to an ever higher volume and, thereby, moved ever
‘closer’ along the Proximity Plane. In this way, the track is made to sound
as though it is ever increasing in ‘proximity’ and, thus, ever encroaching
upon the listener.
Alongside an Auditory Horizon, Horizontal Plane, Horizontal Span and
Proximity Plane, every soundbox should also feature a Vertical Plane and
Vertical Span. Just as the Horizontal Plane and Horizontal Span together
describe a mix’s total horizontal earshot, the Vertical Plane and Vertical
Span describes its capacity to hear vertically. The loop of ticking clocks
heard at a hard-right position in ‘Speak To Me’, for instance, also sounds
‘over and above’ everything else in the mix, given that it features a plethora
Mix as Auditory Response 221
Psychoacoustic Profiles
Indeed, every sound on a record has been ‘heard before’, and from a
clearly specified vantage. As noted, the Auditory Horizon, Horizontal
Plane, Horizontal Span, Proximity Plane, Vertical Plane and Vertical Span
together describe an auditory response to particular sounds, not the partic-
ular sounds as such. And we know this because what I call ‘psychoacous-
tic profiles’ are fixed on record, even as they are ceaselessly dynamic in
concert. Crafting and fine-tuning these profiles, then, must comprise the
aesthetic craft of mixing in toto.
The concept of a psychoacoustic profile defies any simple or narrow
definition. For now, though, it will suffice to conceptualize a psychoacous-
tic profile as the sum of modifications exacted on a soundwave by acoustic
and psychoacoustic interferences. Even if recordists use randomizing algo-
rithms or modular analog processes, virtual synthesis and digital sequenc-
ers, to generate truly aleatoric (random) timbres, pitches and pauses, every
‘sound’ on the resulting record would nonetheless bear a fixed and pre-
cisely repeatable psychoacoustic profile, including completely dry (con-
spicuously nonreverberant) tracks. This means that sequenced synthesis
tracks without any signal processing applied bear a psychoacoustic profile
which is just as fixed as, say, Robert Plant’s vocal track on Led Zeppelin’s
‘No Quarter’ (1973), or Philip Glass’s synthesizer track on the Nonesuch
re-release of ‘Two Pages’ (1994), or any of the vocal tracks on Noah Pred’s
Third Culture (2013) LP. The conspicuous absence of reverberation is a
reverberation profile, after all!
To be clear, psychoacoustic profiles don’t modify the sounds a mix
hears. They are an integral property, a holistic part and parcel, of those
sounds. Listeners cannot disentangle a recorded snare drum hit from its
reverberations, for instance, whether those reverberations were captured
during a live tracking session, synthesized or applied during mixdown
using signal processing. We might move to the left or right of our speak-
ers, or even walk between and through them, but the snare drum remains
fixed wherever the mix engineer positioned it. Likewise, we might don a
222 Jay Hodgson
pair of headphones and run a city block, but the snare drum nevertheless
stays forever put wherever it is mixed to be. And we certainly can’t move
around the singer, to garner a less obstructed perspective on the snare, as
we might at a concert (security allowing).
Single-point perspectival painting provides an obvious analog here.
Every shade of darkness and light, every geometric distortion, in a perspec-
tival painting ultimately combines to form a broader visual perspective for
the painting at large. It is from this perspective, from this precisely spec-
ified spatiotemporal point of view, that every painted object is conveyed
to viewers, making it a kind of narrator for the painting. People don’t just
see a vase when they look at a single-point perspectival painting of flowers
on a table, for instance. Rather, they see a vase geometrically shaped and
shaded to triangulate with the painting’s vanishing point. Whatever can
be seen in perspectival paintings has already been seen, then, and from
a clearly specified place in time and space. And this is true whether the
vista is abstract (as in Cézanne) or irrational (as in Picasso) or hyperrealist
(as in Charles Bell). It is possible, even, to consider this point of view the
painting’s primary subject.
As with perspectival painting, the broader auditory perspective a
record construes—its mix—simply cannot be moved, modified or super-
seded. Listeners can move about the listening environment and alter their
perspective on a mix vis-à-vis room acoustics, to be sure, and they can
equalize the output however they deem fit. But they don’t change the psy-
choacoustic profiles they hear in so doing, nor do they in any manner alter
the broader mix those profiles combine to construe. What they do change
is their perspective on those profiles, their perspective on the auditory per-
spective a mix already construes. Returning to our case study of ‘Speak To
Me’ above, listeners might choose to hear the record from inside a shower
two rooms away, with the bass boosted to an obscene measure, and though
they would likely hear only frequencies under about 500 Hz as a conse-
quence, not a single psychoacoustic profile would change—the heartbeat,
pneumatic drill, electronic drone, maniacal laughter, confessions of mad-
ness, melodic screams, and every other sound the mix hears would all stay
forever put wherever in the soundbox they happen to be. And this can only
be the case if ‘Speak To Me’ comprises an idealized auditory response to
sounds, not sounds per se.
data, if used correctly, moves speakers and headphones at the precise rates,
and in roughly inverse directions and amplitudes, that human tympana
should move to register particular auditory phenomena. Mix engineers
thus ultimately paint on the canvas of human psychophysiology. They
orchestrate psychophysiological reactions to disturbances in air pressure,
each of which conjures a unique auditory delusion.
Tweaking the reverb and equalization on an acoustic guitar track, for
instance, mix engineers can orchestrate a psychophysiological reaction in
listeners that compels them to localize the now-dulled instrument behind a
less reverberant and brighter lead-vocal track. Even then, though, listeners
don’t actually hear a guitar or a voice when they play the record. What
they hear instead is a single acoustic phenomenon, designed by mix engi-
neers to provoke a particular auditory response in listeners, specifically,
to an acoustic guitar and singer performing in tandem. This process, that
is, programming particular auditory responses via speaker and headphone
technology, comprises the substantive basis of all mixing. As such, it com-
prises the substantive basis of all recorded musical communications.
Conclusion
So what does this all mean for recording practice in general? To answer
this question, I’ll need to quickly consider the role that sample rates play
in facilitating the auditory responses I note above. Indeed, few realize that
recording practice comprises a kind of stop-motion auditory animation.
Every technical phenotype of storage medium, from wax cylinders to the
Voice Memo app on your iPhone, stores discontiguous packets of data
called samples. Film provides a useful analogy here. Video cameras—
analog and digital phenotypes—generate discrete snapshots quickly
enough that when they are displayed in sequence and at the same rate
they were encoded, the images seem to animate, taking on an illusory life
beyond empirical two-dimensionality. The animated images don’t actually
move, of course, just as soundwaves don’t actually undulate on records.
Mixes, that is, recorded musical communications, are comprised of
audio samples which portray soundwaves in various states of propaga-
tion and decay. Once we record you plucking the second-lowest string of
your acoustic guitar, for instance, we can always scrub past the first three
seconds to hear the exact same guitar timbre in precisely the same state
of decay, and according to the same precisely fixed acoustic and psycho-
acoustic variables (we can always skip the first three seconds’ worth of
samples, in other words). We might even replace a few samples we don’t
like with other samples, using the ‘sample replace’ function on whichever
DAW we use, as is so often done by mix engineers working on drums
nowadays. We don’t hear the plucking of an acoustic guitar when we play
the record, after all, and we don’t hear an A below middle-C propagate
and decay. What we hear are discrete samples played at a very precise rate
(44.1 kHz, most often), each of which depicts a different moment of prop-
agation and decay. Thus does the map cover the terrain . . .
224 Jay Hodgson
Notes
1 For more on ‘veridic’ and ‘non-veridic’ sounds in modern recording practice, see John
Andrew Fisher’s ‘Rock and Recording: The Ontological Complexity of Rock Music’ in
Musical Worlds: New Directions in the Philosophy of Music, ed. P. Alperson (Pennsylva-
nia: University of Pennsylvania Press, 1998), 109–123. I complicate Fisher’s taxonomy in
my own book, Understanding Records: A Field Guide to Recording Practice (New York:
Bloomsbury, 2010), 71–72.
2 To be clear, this is not a criticism of so-called ‘interdisciplinary’ studies of modern
recording practices. I simply agree with Louise Meintjes’s observation that “sociology
and media studies have led the way in generating discussions about studio-based creativ-
ity”, in Louise Meintjes, Sound of Africa! Making Music Zulu in a South African Studio
(Washington, DC: Duke University Press, 2003), 27.
3 See Jay Hodgson with Steve MacLeod, Representing Sound: Notes on the Ontology of
Recorded Musical Communications (Kitchener/Waterloo: Wilfrid Laurier University
Press, 2013).
4 The mix I consider here appears as track five on the 2003 remaster of Nick Drake’s Pink
Moon (Island: B000025XKM, 1972).
5 The mix I consider here appears on The Beatles, ‘Can’t Buy Me Love (2015 Stereo Ver-
sion)’, track five on The Beatles, 1 (2015 Version) (Capitol Records: B01576X99U, 2015).
Mix as Auditory Response 225
6 For an edifying academic discussion of the ‘soundbox’, see Ruth Dockwray and Allan
Moore, ‘Configuring the Soundbox 1965–1972’ (2010) in Popular Music 29 (2): 181–
197. To see how it is typically used in audio engineering pedagogy, see David Gibson, The
Art of Mixing: A Visual Guide to Recording, Engineering and Production, Second Edition
(Boston, MA: Course Technology PTR, 2005).
7 Specifically, the 2011 remastered version of Pink Floyd’s Dark Side of the Moon (Capitol:
B004ZN9RWK, 1973/2011).
8 See Nicholas Schaffner, Saucerful of Secrets: The Pink Floyd Odyssey (Philadelphia, PA:
Delta Books, 1992).
9 See Herbert Marcuse, One Dimensional Man: Studies in the Ideology of Advanced Indus-
trial Society, Second Edition (Boston, MA: Beacon Press, 1991).
10 For more on the practical mixing tools and techniques recordists use to create these com-
ponents of a ‘soundbox’ see, for instance, Roey Izhaki, Mixing Audio: Concepts, Prac-
tices and Tools, Second Edition (Boston, MA: Focal Press, 2011); Mike Senior, Mixing
Secrets in the Small Studio (Boston, MA: Focal Press, 2011); David Gibson, The Art
of Mixing: A Visual Guide to Recording, Engineering and Production, Second Edition
(Boston, MA: Course Technology PTR, 2005); Alexander Case, Sound FX: Unlocking
the Creative Potential of Recording Studio Effects (Boston, MA: Focal Press, 2007); Wil-
liam Moylan, Understanding and Crafting the Mix: The Art of Recording, Second Edition
(Boston, MA: Focal Press, 2007); Bobby Owsinski, The Mixing Engineer’s Handbook,
Second Edition (New York: Cengage, 2006).
Bibliography
Case, Alexander (2007). Sound FX: Unlocking the Creative Potential of Recording Studio
Effects. Boston, MA: Focal Press.
Dockwray, Ruth and Moore, Allan (2010). Configuring the Soundbox 1965–1972. Popular
Music 29 (2): 181–197.
Fisher, John Andrew (1998). ‘Rock and Recording: The Ontological Complexity of Rock
Music.’ In Musical Worlds: New Directions in the Philosophy of Music, ed. P. Alperson.
Pittsburgh, PA: University of Pennsylvania Press, pp. 109–123.
Gibson, David (2005). The Art of Mixing: A Visual Guide to Recording, Engineering and
Production, Second edition. Boston, MA: Course Technology PTR.
Hodgson, Jay (2010). Understanding Records: A Field Guide to Recording Practice. New
York: Bloomsbury.
Hodgson, Jay with MacLeod, Steve (2013). Representing Sound: Notes on the Ontology of
Recorded Musical Communications. Kitchener/Waterloo: Wilfrid Laurier University Press.
Izhaki, Roey (2011). Mixing Audio: Concepts, Practices and Tools, Second edition. Boston,
MA: Focal Press.
Marcuse, Herbert (1991). One Dimensional Man: Studies in the Ideology of Advanced Indus-
trial Society, Second edition. Boston, MA: Beacon Press.
Meintjes, Louise (2003). Sound of Africa! Making Music Zulu in a South African Studio.
Washington, DC: Duke University Press.
Moylan, William (2007). Understanding and Crafting the Mix: The Art of Recording, Second
edition. Boston, MA: Focal Press.
Owsinski, Bobby (2006). The Mixing Engineer’s Handbook, Second edition. New York: Cengage.
Schaffner, Nicholas (1992). Saucerful of Secrets: The Pink Floyd Odyssey. Philadelphia, PA:
Delta Books.
Senior, Mike. Mixing Secrets in the Small Studio. Boston, MA: Focal Press, 2011.
Discography
The Beatles, 1 (2015 Version) (Capitol Records: B01576X99U, 2015).
Nick Drake, Pink Moon (Island: B000025XKM, 1972).
Pink Floyd, Dark Side of the Moon (Capitol: B004ZN9RWK, 1973/2011).
15
Introduction
226
An Intelligent Systems Approach 227
No
Evaluaon Processing
How does it sound? What effects and
Objective achieved? what settings to use?
for later use, they lack the ability to take intelligent decisions, such as
adapting to different acoustic environments or different set of inputs. Sec-
ond, most state-of-the-art audio signal processing techniques focus on
single-channel signals. Yet multichannel or multitrack signals are perva-
sive, and the interaction and dependency between channels plays a criti-
cal role in audio production quality. This issue has been addressed in the
context of audio source separation research, but the challenge in source
separation is generally dependent on how the sources were mixed, not on
the respective content of each source. New, multi-input multi-output audio
signal processing methods are required, which can analyze the content
of all sources in order to improve the quality of capturing, editing and
combining multitrack audio. Finally, advances in machine learning must
be tailored towards problems and practical applications in the domain of
audio production. This chapter presents an overview of recent advances in
this area.
Enabling concepts
Audio Input
Feature
Extraction
Constraints
(psychoacoustics,
Features
best practices,
machinelearning) Rules Signal
Feature
Processing Processing
Controls
Side-chain processing
Audio Output
Figure 15.2 Block diagram of an intelligent audio effect. Features are extracted by
analysis of the audio signal. These features are then processed based on a set of rules
intended to mimic the behavior of a trained engineer. A set of controls are produced
which are used to modify the audio signal
An Intelligent Systems Approach 229
Constraints
Feature … Feature
Feature
Extraction Extraction
Extraction
Cross-Adaptive
Feature Processing
where there are M input tracks and L channels in the output mix. K is the
length of the control vector c and x is the multitrack input. Thus, the resul-
tant mixed signal at time n is a sum over all input channels, of a control
vectors convolved with the input signal.
Any cross-adaptive digital audio effect that employs linear filters may
be described in this manner. For automatic faders and source enhance-
ment, the control vectors are simple scalars, and hence the convolution
operation becomes multiplication. For polarity correction, a binary valued
scalar, ±1, is used. For automatic panners, two mixes are created, where
panning is also determined with a scalar multiplication (typically, the
sine-cosine panning law). For delay correction, the control vectors become
a single delay operation. This applies even when different delay estimation
methods are used, or when there are multiple active sources. If multitrack
An Intelligent Systems Approach 231
And later values may either be given by a sliding window, which reduces to
2
xRMS (n + 1) = x 2 (n + 1) / M + xRMS
2
( n) − x 2 ( n + 1 − M ) / M , (4)
2
xRMS (n + 1) = β x 2 (n + 1) + (1 − β ) xRMS
2
(n) . (5)
α and β and represent time constants of IIR filters and allow for the control
vector and RMS estimation, respectively, to smoothly change with varying
conditions. Eq. (4) represents a form of dynamic real-time extraction of a
feature (in this case, RMS), and Eq. (5) represents an accumulative form.
Psychoacoustic Studies
Important questions arise concerning the psychoacoustics of mixing mul-
titrack content. For instance, little has been formally established concern-
ing user preference for relative amounts of dynamic range compression
used on each track. Admittedly, such choices are often artistic decisions,
but there are many technical tasks in the production process for which
listening tests have not yet been performed to even establish whether a
listener preference exists.
Listening tests must be performed to ascertain the extent to which lis-
teners can detect undesired artifacts that commonly occur in the audio
production process. Important work in this area has addressed issues such
as level balance preference (King et al., 2010, 2012), reverberation level
preference (Leonard et al., 2012, 2013), ‘punch’ (Fenton et al., 2015), per-
ceived loudness and dynamic range compression (Wilson et al., 2016), as
well as the design and interpretation of such listening tests.
Before they are ready for practical use, intelligent software tools need to
be evaluated by both amateurs and professional sound engineers to assess
their effectiveness and compare different approaches. In contrast to sep-
aration of sources in multitrack content, there has been little published
work on subjective evaluation of the intelligent tools for mixing multitrack
audio. Where possible, prototypes should also be tested with engineers
234 Joshua D. Reiss
Recent developments
Faders
The most common form of multitrack automatic mixing system is based
around simple level adjustments on each track. In almost all cases, it
begins with the assumption that each track is meant to be heard at roughly
equal loudness levels.
Mansbridge et al. (2012b) provided a real-time system, using ITU 1770
as the loudness model. The off-line system described in Ward et al. (2012)
attempted to control faders with auditory models of loudness and partial
loudness. In theory, this approach should be more aligned with perception
and take into account masking, at the expense of computational efficiency.
But Wichern et al. (2015) showed that the use of an auditory model offered
little improvement over simple single-band, energy-based approaches. Inter-
estingly, the evaluation in Mansbridge et al. (2012b) showed that autono-
mous faders could compete with manual approaches by professionals, and
test subjects gave the autonomous system highly consistent ratings, regard-
less of the song (and its genre and instrumentation) used for testing. This
suggests that the equal loudness rule is broadly applicable, whereas pref-
erence for decisions in manual mixes differs widely dependent on content.
Equalization
The rules and best practices for equalization typically fall into two catego-
ries: artifact correction, such as hum removal (Brandt et al., 2014) and the
equalization of salient frequencies (Bitzer et al., 2008), or creative equal-
ization (which may still follow rules and best practices), where equalizers
are applied in order to achieve a certain overall spectrum (Pestana et al.,
2013; Deruty et al., 2014).
Table 15.1 Classification of intelligent audio production tools since those described in Reiss (2011)
Stereo Positioning
The premise of Mansbridge et al. (2012a) is that one of the primary goals
of stereo panning is to ‘fill out’ the stereo field and reduce masking. It set
target criteria of source balancing (equal numbering and symmetric posi-
tioning of sources on either side of the stereo field), spatial balancing (uni-
form distribution of levels) and spectral balancing (uniform distribution
of content within each frequency band). It further assumes that the higher
the frequency content of a source, the more it will be panned, and that no
hard panning will be applied. Finally, it used a multitude of techniques to
position the sources; amplitude panning, timing differences and double
tracking.
Pestana et al. (2014a) took a different approach, where different fre-
quency bands of each multitrack are assigned different spatial positions in
the mix. This approach is unique among the intelligent multitrack mixing
tools since it does not emulate, even approximately, what might be per-
formed by a practitioner. That is, practitioners aim for a single position
(albeit sometimes diffuse) of each source. However, it captures the spirit
of many practical approaches since it greatly reduces masking and makes
effective use of the entire stereo field. In fact, Matz et al. (2015) showed
that dynamic spectral panning had a larger effect in the overall improve-
ment provided by automatic mixing than any of the other tools they con-
sidered (intelligent distortion, autonomous faders and multitrack EQ).
Reverb
Of all the standard audio effects found on a mixing console or as built-in
algorithms in a digital audio workstation, there has perhaps been the least
effort on intelligent systems design for reverberation. Chourdakis et al.
(2016a, 2016b) proposed an adaptive digital audio effect for artificial
238 Joshua D. Reiss
Mix Evaluation
One of the chief distinguishing characteristics between the early work
on intelligent mixing systems and those described herein is that very
few of the early systems had any form of subjective evaluation, whereas
now this is standard practice. A popular form of evaluation for such sys-
tems has become multistimulus rating, similar to that used in MUSHRA.
An Intelligent Systems Approach 241
Conclusions
of their craft, and help inexperienced users create high-quality mixes. The
other type of system would be a ‘black box’ for the musician that allows
decent live sound without an engineer. This would be most beneficial for
the small band or small venue that doesn’t have or can’t afford a sound
engineer, or for recording practice sessions where a sound engineer is not
typically available.
There are major concerns with such an approach. Much of what a sound
engineer does is creative and based on artistic decisions. It is doubtful that
such decisions could be effectively reproduced by a machine. But if the auto-
mation is successful, then machines may replace sound engineers. However,
it is important to note that these tools are not intended to remove the creativ-
ity from audio production. Nor do they require software to reproduce artistic
decisions, although this would be an interesting direction for future research.
Rather, the tools rely on the fact that many of the challenges are technical
engineering tasks, some of which are perceived as creative decisions because
there is a wide range of approaches without a clear understanding of listener
preferences. By automating those engineering aspects of record production,
it will allow the musicians to concentrate on the music and allow the audio
engineers to concentrate on the more interesting, creative challenges.
Bibliography
Aichinger, P., et al. (2011). ‘Describing the Transparency of Mixdowns: The Masked-to-
Unmasked Ratio.’ In 130th Audio Eng. Soc Convention.
Bittner, R., et al. (2014). ‘MedleyDB: A Multitrack Dataset for Annotation-Intensive MIR
Research.’ In 15th International Society for Music Information Retrieval Conference
(ISMIR).
Bitzer, J., et al. (2008). ‘Evaluating Perception of Salient Frequencies: Do Mixing Engineers
Hear the Same Thing?’ In 124th Audio Engineering Society Convention, Amsterdam.
Brandt, M., et al. (2014). Automatic Detection of Hum in Audio Signals. Journal of Audio
Engineering Society 62 (9): 584–595.
Cartwright, M., et al. (2013). ‘Social-EQ: Crowdsourcing an Equalization Descriptor Map.’
In 14th International Society for Music Information Retrieval, Curitiba, Brazil.
Cartwright, M., et al. (2014). ‘Mixploration: Rethinking the Audio Mixer Interface.’ In 19th
International Conference on Intelligent User Interfaces Proceedings (IUI14), Haifa,
Israel.
Chourdakis, E.T., et al. (2016a). ‘Automatic Control of a Digital Reverberation Effect using
Hybrid Models.’ In AES 60th International Conference Leuven, Belgium.
Chourdakis, E.T., et al. (2016b). A Machine Learning Approach to Design and Evaluation
of Intelligent Artificial Reverberation. Journal of Audio Engineering Society (to appear).
Clifford, A., et al. (2010). ‘Calculating Time Delays of Multiple Active Sources in Live
Sound.’ In 129th AES Convention, San Francisco.
Clifford, A., et al. (2011a). Reducing Comb Filtering on Different Musical Instruments Using
Time Delay Estimation. Journal on the Art of Record Production 5: 1–13.
Clifford, A., et al. (2011b). ‘Microphone Interference Reduction in Live Sound.’ In 14th Int.
Conference on Digital Audio Effects (DAFx-11), Paris.
Clifford, A., et al. (2013). Using Delay Estimation to Reduce Comb Filtering of Arbitrary
Musical Sources. Journal of the Audio Engineering Society 61 (11): 917–927.
De Man, B., et al. (2013a). ‘A Knowledge-Engineered Autonomous Mixing System.’ In 135th
AES Convention, New York.
De Man, B., et al. (2013b). A Semantic Approach to Autonomous Mixing. Journal on the Art
of Record Production (JARP) 8: 1–23.
An Intelligent Systems Approach 243
De Man, B., et al. (2014a). ‘The Open Multitrack Testbed.’ In 137th AES Convention, Los
Angeles.
De Man, B., et al. (2014b). ‘An Intelligent Multiband Distortion Effect.’ In AES 53rd Interna-
tional Conference on Semantic Audio, London, UK.
Deruty, E., et al. (2014). Human–Made Rock Mixes Feature Tight Relations between Spec-
trum and Loudness. Journal of Audio Engineering Society 62 (10): 643–653.
Dewey, C., et al. (2013). ‘Novel Designs for the Parametric Peaking EQ User Interface.’ In
134th AES Convention, Rome.
Durant, E.A., et al. (2004). Efficient Perceptual Tuning of Hearing Aids with Genetic Algo-
rithms. IEEE Transactions on Speech and Audio Processing 12 (2): 144–155.
Fenton, S., et al. (2015). ‘Towards a Perceptual Model of “Punch” in Musical Signals.’ In
139th AES Convention, New York.
Giannoulis, D., et al. (2013). Parameter Automation in a Dynamic Range Compressor. Jour-
nal of Audio Engineering Society 61 (10): 716–726.
Hafezi, S., et al. (2015). Autonomous Multitrack Equalisation Based on Masking Reduction.
Journal of the Audio Engineering Society 63 (5): 312–323.
Heise, S., et al. (2010). ‘A Computer-Aided Audio Effect Setup Procedure for Untrained
Users.’ In 128th Audio Engineering Society Convention, London.
Jillings, N., et al. (2013). ‘Performance Optimization of GCC-PHAT for Delay and Polarity
Correction under Real World Conditions.’ In 134th AES Convention, Rome.
King, R., et al. (2010). ‘Variance in Level Preference of Balance Engineers: A Study of Mix-
ing Preference and Variance Over Time.’ In 129th Audio Engineering Society Convention,
San Francisco.
King, R., et al. (2012). ‘Consistency of Balance Preferences in Three Musical Genres.’ In
133rd AES Convention.
Kirchberger, M., et al. (2016). Dynamic Range Across Music Genres and the Perception of
Dynamic Compression in Hearing-Impaired Listeners. Trends in Hearing 20: 1–16.
Kokkinis, E., et al. (2011). ‘Detection of ‘Solo Intervals’ in Multiple Microphone Multiple
Source Audio Applications.’ In 130th AES Convention.
Kuk, F.K., et al. (1992). The Reliability of a Modified Simplex Procedure in Hearing Aid
Frequency Response Selection. Journal of Speech and Hearing Research 35 (2): 418–429.
Leonard, B., et al. (2012). ‘The Effect of Acoustical Environment on Reverberation Level
Preference.’ In 133rd AES Convention, San Francisco.
Leonard, B., et al. (2013). ‘The Effect of Playback System on Reverberation Level Prefer-
ence.’ In 134th Audio Engineering Society Convention, Rome.
Lopez, S.V., et al. (2010). Quantifying Masking in Multi-Track Recordings. Sound and Music
Computing 1–8.
Loviscach, J. (2008). ‘Graphical Control of a Parametric Equalizer.’ In Audio Engineering
Society Convention 124.
Ma, Z., et al. (2013). ‘Implementation of an Intelligent Equalization Tool Using Yule-Walker
for Music Mixing and Mastering.’ In 134th AES Convention, Rome.
Ma, Z., et al. (2014). ‘Partial Loudness in Multitrack Mixing.’ In AES 53rd International
Conference on Semantic Audio, London.
Ma, Z., et al. (2015). Intelligent Multitrack Dynamic Range Compression. Journal of Audio
Engineering Society 63 (6): 412–426.
Maddams, J., et al. (2012). ‘An Autonomous Method for Multi-Track Dynamic Range Com-
pression.’ In Digital Audio Effects (DAFx), York, 1–8.
Mansbridge, S., et al. (2012a). ‘An Autonomous System for Multi-track Stereo Pan Position-
ing.’ In 133rd AES Convention, San Francisco.
Mansbridge, S., et al. (2012b). ‘Implementation and Evaluation of Autonomous Multi-Track
Fader Control.’ In 132nd Audio Engineering Society Convention, Budapest, 1–8.
Mason, A., et al. (2015). ‘Adaptive Audio Reproduction Using Personalised Compression.’ In
AES 57th International Conference, Hollywood, CA.
Matz, D., et al. (2015). ‘New Sonorities for Early Jazz Recordings Using Sound Source Sep-
aration and Automatic Mixing Tools.’ In ISMIR, Malage.
Mecklenburg, S., et al. (2006). ‘subjEQt: Controlling an Equalizer through Subjective Terms.’
In Computer-Human Interaction EA ’06 Extended Abstracts on Human Factors in Com-
puting, Montreal.
244 Joshua D. Reiss
Moorer, J.A. (2000). Audio in the New Millennium. Journal of the Audio Engineering Society
48 (5): 490–498.
Mycroft, J., et al. (2013). ‘The Influence of Graphical User Interface Design on Critical Lis-
tening Skills.’ In Sound and Music Computing (SMC), Stockholm.
Neuman, A.C., et al. (1987). An Evaluation of Three Adaptive Hearing Aid Selection Strate-
gies. The Journal of Acoustical Society of America 82 (6): 1967–1976.
Pardo, B. (2015). Data on SocialEQ. J.D. Reiss.
Pardo, B., et al. (2012a). ‘Building a Personalized Audio Equalizer Interface with Transfer
Learning and Active Learning.’ In 2nd International ACM Workshop on Music Information
Retrieval with User-Centered and Multimodal Strategies (MIRUM), Nara, Japan.
Pardo, B., et al. (2012b). ‘Towards Speeding Audio EQ Interface Building with Transfer
Learning.’ In New Interfaces for Musical Expression, Ann Arbor, MI.
Pestana, P. (2013). Automatic Mixing Systems Using Adaptive Audio Effects. PhD, Universi-
dade Catolica Portuguesa.
Pestana, P.D., et al. (2013). ‘Spectral Characteristics of Popular Commercial Recordings
1950–2010.’ In 135th AES Convention, New York.
Pestana, P.D., et al. (2014a). ‘A Cross-Adaptive Dynamic Spectral Panning Technique.’ In
17th Int. Conference on Digital Audio Effects (DAFx-14), Erlangen, Germany.
Pestana, P.D., et al. (2014b). ‘Intelligent Audio Production Strategies Informed by Best Prac-
tices.’ In AES 53rd International Conference on Semantic Audio, London, UK.
Rafii, Z., et al. (2009). ‘Learning to Control a Reverberator using Subjective Perceptual
Descriptors.’ In 10th Int. Conf. Music Inf. Retrieval (ISMIR), Kobe, Japan.
Reed, D. (2000). ‘A Perceptual Assistant to Do Sound Equalization.’ In 5th International
Conference on Intelligent User Interfaces, New Orleans.
Reiss, J.D. (2011). ‘Intelligent Systems for Mixing Multichannel Audio.’ In 17th Interna-
tional Conference on Digital Signal Processing (DSP2011), Corfu, Greece, 1–6.
Sabin, A., et al. (2008). ‘Rapid Learning of Subjective Preference in Equalization.’ In 125th
Audio Engineering Society Convention, San Francisco.
Sabin, A., et al. (2009a). ‘2DEQ: An Intuitive Audio Equalizer.’ In ACM Creativity and Cog-
nition, Berkeley, CA.
Sabin, A., et al. (2009b). ‘A Method for Rapid Personalization of Audio Equalization Param-
eters.’ In ACM Multimedia, Beijing, China.
Sabin, A.T., et al. (2011). Weighted-Function-Based Rapid Mapping of Descriptors to Audio
Processing Parameters. Journal of Audio Engineering Society 59 (6): 419–430.
Scott, J., et al. (2011). ‘Automatic Multi-Track Mixing Using Linear Dynamical Systems.’ In
8th Sound and Music Computing Conference (SMC), Padova.
Stasis, S., et al. (2015). ‘A Model for Adaptive Reduced-Dimensionality Equalisation (Best
Paper, 2nd Prize).’ In 18th Int. Conference on Digital Audio Effects (DAFx-15), Trond-
heim, Norway.
Stelmachowicz, P.G., et al. (1994). Preferred Hearing-Aid Frequency Responses in Simulated
Listening Environments. Journal of Speech and Hearing Research 37 (3): 712–719.
Verfaille, V., et al. (2006). Adaptive Digital Audio Effects (A-DAFx): A New Class of Sound
Transformations. IEEE Transactions on Audio, Speech and Language Processing 14 (5):
1817–1831.
Wakefield, G.H., et al. (2005). Genetic Algorithms for Adaptive Psychophysical Procedures:
Recipient-Directed Design of Speech-Processor MAPs. Ear Hear 26 (4): 57S–72S.
Wakefield, J., et al. (2015). ‘An Investigation into the Efficacy of Methods Commonly
Employed by Mix Engineers to Reduce Frequency Masking in the Mixing of Multitrack
Musical Recordings’. In 138th AES Convention.
Ward, D., et al. (2012). ‘Multi-Track Mixing Using a Model of Loudness and Partial Loud-
ness.’ In Audio Engineering Society (AES) 133rd Convention, San Francisco.
White, P. (2008). Automation for the People. Sound on Sound 23 (12).
Wichern, G., et al. (2015). ‘Comparison of Loudness Features for Automatic Level Adjust-
ment in Mixing.’ In 139th AES Convention, New York.
Wilson, T., et al. (2015). ‘A Semantically Motivated Gestural Interface for the Control of a
Dynamic Range Compressor.’ In 138th AES Convention.
Wilson, A., et al. (2016). Perception of Audio Quality in Productions of Popular Music. Jour-
nal of Audio Engineering Society 64 (1/2): 23–34.
16
Introduction
245
246 Gary Bromham
He followed his nose. He knew how to do stuff but didn’t know why. He was
determined not to allow himself to become constrained by knowledge. He
saw no barriers. He was as Tom Magliozzi aptly put it “Unencumbered by the
thought process!”
Apprenticeship
If that person is a bullshitter then you don’t really stand a chance. That system
can be very arbitrary. Academic practice is however more likely to be reliable.
In academia from a scientific point of view at least you’re more likely to find
someone who knows what they are talking about. In studios you might be more
likely to find out more about what constitutes creativity. It’s a bit of a lottery
in many ways.5
With the adoption of digital recording and the ubiquitous nature of hard-
disk-based recording systems, the DAW has become a metaphor for the
recording studio in its conventional form. There is a greater need than
248 Gary Bromham
People have forgotten the process and function of balance; engineers got that
right at the recording stage. Big differences existed between the live source and
the medium which it was being played back from, a lot of engineering in the
early days was overcoming the limitations of that medium. Tape definitely had
shortcomings!6
A great deal of this statement stems from necessity, and it wasn’t really until
the mid-1980s with the advent of 48-track sessions that the mix started to
become a post-production process. From the late 1980s onwards, and with
the emergence of digital recording technologies, the demographic would
change again! The digital revolution in a sense changed mixing forever.
No longer were there concerns about degradation of audio quality or about
restrictions imposed by having a limited track count. Instead, the possi-
bilities became endless. This changed the role of the mix engineer into an
interpreter who would have to imagine what the artists’ intentions were
and act as someone who was expected to have vision and creativity when
shaping the sound of a mix. Thus was the ‘superstar’ mix engineer con-
ceived. Probably no one personified this more than Bob Clearmountain, a
house engineer at the Power Station in New York, who started to be hired
for his creative skills as a mix engineer. The important thing is to acknowl-
edge his pedigree as a recording engineer in the first instance. Andrew
Bourbon says, “You need to see recording as part of mixing, they are not
a separate entity. Look at all the great mix engineers and they are all good
recording engineers”.
In a sense, the function of the studio has come full circle with the use
of the laptop as mobile recording and production medium. In his chapter
‘The end of the world as we know it’ in The Art of Record Production, Paul
Théberge (2012) refers to the studio as a mobile entity and talks about
how this has impacted record production and mix practice. The bedroom
studio is now ubiquitous, and the use of headphones for monitoring is
now commonplace. The typical home studio is entirely unsuitable for
mixing records, so there is a greater need than ever to grasp how acous-
tics will impact our environment and how to work around these inherent
shortcomings.
Listening Skills
critical listening being a key part of their curriculum. The fact that many
of us work in less-than-perfect recording spaces has meant that a grounded
knowledge in both technical listening skills and acoustics is not a lux-
ury anymore. It is not uncommon to see a mix engineer walking around
the control room in a studio when evaluating bass response, for example.
When used in combination with a basic technical understanding of room
modes, room shape, reflections and diffusion, for example, this knowledge
can help to inform creative and aesthetic decisions that impact the sonic
quality of the final mix. Positioning and choice of speakers are very much
determined by the sound of the room they are placed in. We cannot ignore
the fact that most popular music is now consumed via headphones, and it
is therefore remiss of the sound engineer to ignore these facts when mix-
ing a song. Some type of formal academic training in the area of listening
skills and physics is now almost certainly essential to making critical deci-
sions when mixing sound. In Jason Corey’s book Audio Production and
Critical Listening—Technical Ear Training (2010), “Technical ear training
is a type of perceptual learning focused on timbral, dynamic, and spa-
tial attributes of sound as they relate to audio recording and production”.
A concept of spatial evaluation is also necessary when learning to listen
to sound. An ability to assess the perceived width and depth of the sound
stage is important when creating a space to mix and record in. Nowhere
will this have more impact than in our choice of placement of speakers, as
has already been mentioned.
It is also important to make a distinction between critical and analytical
listening. In Understanding and Crafting the Mix (2006), William Moylan
talks about technical and scientific listening skills vs. artistic or creative
expression. He says that understanding sound is perceiving sound for its
inherent qualities. Francis Rumsey and Tim McCormick discuss a simi-
lar phenomenon in their book Sound and Recording (2009) when talking
about auditory perception. Though not essential to becoming an accom-
plished mixing engineer, a basic understanding of how psychoacoustics
work can be helpful. The word ‘perception’ is used several times in this
chapter, particularly in the context of using vintage technologies and their
inherent sonic imprints. Psychoacoustics is essentially the scientific study
of sound perception. More specifically, it is the measurement of psycho-
logical and physiological responses to sound. Quite often how we perceive
something to sound is invariably not how it actually sounds; it is influenced
by many factors outside of the auditory experience. The GUI (graphic
user interface) in a DAW, an audio plugin or an iOS app, depending on
their visual reference point, will influence the outcome of processing a
sound. This becomes an important factor for the engineer to consider, as
employing a certain degree of scientific study will tell them that they are
not always hearing what they think they are hearing. It is also particularly
relevant in an age when there are so many visual representations in the
studio and we are often told that we should listen to sound and not look at
it. This challenge is more pronounced today than it was thirty years ago
when engineers relied far less on visual feedback. Most music today is
composed and produced on a computer.
How Can Academics Inform Mix-Craft? 251
Since 2002, the software company Waves Audio has released a string of
collections of plugins endorsed by famous mix engineers and producers.
Known as its Signature Series and including such luminaries as Eddie
Kramer, Chris Lord-Alge and Jack Joseph Puig, the plugin bundles pro-
vide the user with the opportunity to use, albeit virtually, a software
re-creation of the analog hardware used by the aforementioned profes-
sionals in their studios. Lists of presets created by these same famous
mix engineers for kick drum, rhythm guitar and lead vocal, among oth-
ers, provide an insight and a glimpse into the engineers’ technical and
aesthetic worlds. Slate Digital and Universal Audio have created a simi-
lar business model feeding off the same retro obsession. They draw upon
the phenomena of technostalgia to sell us a tiny part of a golden age of
studio technology used to make classic recordings that only a privileged
few were lucky enough to have been acquainted with. Accompanying
this trend is a myth that you, the amateur, or professional amateur, can
mix records to sound like one of your idols. An interesting observation
might concern the way such practices impact creativity. Creating an orig-
inal sonic stamp or signature often comes from the misuse or misappro-
priation of technology. It could be argued that the use of ‘famous mix
engineer’ presets doesn’t encourage the use of digital plugin technology
in an enquiring way but more likely encourages its use in an unquestion-
ing manner.
The practice of taking some of the sonic signature from mix engineers
and producers without necessarily understanding why you might use them
promotes a sort of sonic tourism without a map. Without an ability to inter-
rogate these technologies and their accompanying workflows, the use of
endless lists of presets and templates to mix a song merely serves as a
shortcut without providing an understanding of the underlying tools used
to execute a mix. Paul Théberge has discussed the history behind the use
and purpose of presets in Any Sound You Can Imagine (Théberge, 1997).
The original concept was to provide a keyboard player with a means of
taking the sounds they had created in the studio into a live context, but
more and more this idea has become an excuse for manufacturers to fill
a new machine with hundreds of presets to market new technologies. The
role and function of presets for mixing has largely been ignored or swept
under the carpet; this is for good reason because of the negative stigma
attached to such practices. It could be argued that the use of presets can
also limit creativity. A paint-by-numbers approach to using presets in mix-
ing can easily produce similar generic results in people’s mixes. It has been
suggested that many tracks produced in current popular music sound the
same today and that they have similar sonic imprints. This might largely be
because the same plugins with the same presets are being used. A ‘one size
fits all’ approach to mixing without any understanding of the tools being
used will probably not yield a unique-sounding mix. Nowhere is this seen
more forcibly today than in mastering, where it is not uncommon to see a
preset chain used to correct a less than satisfactory mix.
252 Gary Bromham
Academic practice and study can help us understand some of the tech-
niques used by well-known studio practitioners. A student should aim to
reflect on the practices and aesthetics used by producers and engineers
and not merely try to emulate them. Academic practice encourages us to
interrogate and understand why some of these digital recreations are being
used. For many young producers and engineers, or those without a con-
ventional internship, the context of the original equipment could easily be
lost on them.
Another interesting aspect is the obvious marketing benefits gained by
selling products based upon the legacy of producers and engineers and the
nostalgic value attached to their equipment. There is an assumption with
limited validity that there might be a shortcut to you sounding just like
them. The new Stephen Slate, Chris Lord-Alge mixing course found at
www.audiolegends.com takes this concept to a new level. A full session in
a DAW of choice is sold and subsequently downloaded with all the plugins
needed to copy the mixing workflow of Chris Lord-Alge. It is in many
ways an ingenious marketing strategy! It might arguably be more import-
ant to ask ourselves what the impact of copying the techniques and tech-
nologies of others might be. Can it have a negative impact on our ability
to produce original work? Misappropriating technology can often produce
more interesting results, and using some of our favorite mix engineers’
techniques out of context might be of more interest when considering their
impact on creativity: “Sometimes, not knowing the theoretical operation
of a tool can result in more interesting results by thinking outside the box”
(Cascone, 2004).
The trouble begins with a design philosophy that equates ‘more options’ with
‘greater freedom’. Designers struggle endlessly with a problem that is almost
How Can Academics Inform Mix-Craft? 253
non-existent for users. How do we pack the maximum number of options into
the minimum space and price? In my experience, the instruments and tools that
endure (because they are loved by their users) have limited options.
(Eno, 1999)
Software emulations are not inherently nostalgic, though much of the market-
ing surrounding them capitalizes on the desire to harness the past. Since digital
audio processes are distinctly different from analog electronic and acoustic
ones, these products present a functionality that masks the actual technology
involved.
Perceptual experiments are another great way to help us evaluate the mood
created when listening to music, and the ability to measure similarities and
differences between mixes can assist in informing future technologies for
engineers. This can help us to identify attributes and patterns between mixes.
Research at Queen Mary University of London using technologies found at
www.isophonics.net, such as Sonic Visualiser, Soundbite and Mood Player,
can help to inform mix practices by analyzing sound attributes and assess-
ing their impact on workflow. A great deal can be learned from conducting
mood-based studies and analyzing their effects upon attitudes to a mix.
Conclusion
The word ‘academia’ can send mixed messages when mentioned in creative
circles. Some would argue, as Tony Platt has suggested in our discussion,
that academics have a need to ‘pigeonhole’ work processes or put every-
thing into neat boxes to explain what is taking place in the studio. He finds it
frustrating that academics think there is a golden key, which will somehow
unlock the door to knowledge. Francis Rumsey conversely suggests the pos-
sibility that maybe academic training is a substitute for the fact that record-
ing studios aren’t providing a conventional learning environment anymore.
Notes
1 Dr. Andrew Bourbon. Interview with author 2015.
2 Tony Platt. Interview with author 2015.
3 Dennis Weinreich. Interview with author 2015.
256 Gary Bromham
Bibliography
Bech, Søren and Zacharov, Nick (2006). Perceptual Audio Evaluation. Chichester: Wiley.
Cascone, Kim (2004). ‘Audio Culture: Readings in Modern Music.’ In The Aesthetics of Fail-
ure: “Post-Digital” Tendencies in Contemporary Computer Music. London: The Contin-
uum International Publishing Group Inc., pg. 392.
Corey, Jason (2010). Audio Production and Critical Listening: Technical Ear Training.
Oxford: Focal Press.
Czikszentmihalyi, Mihalyi (1996). The Psychology of Discovery and Invention. New York:
Harper Collins.
De Man, Brecht and Reiss, Joshua (2015). Analysis of Peer Review in Music Production.
Journal on the Art of Record Production (Vol. 10).
Emmerson, Simon (2000). Music, Electronic Media and Culture. Farnham: Ashgate Publish-
ing Ltd.
Eno, Brian (1999). The Revenge of the Intuitive. Wired Magazine (7).
Fazekas, G., Raimond, Y., Jacobson, K. and Sandler, M. (2010). An Overview of Web activ-
ities in the OMRAS2 Project. Journal of New Music Research on Music Informatics and
the OMRAS2 Project.
Lanier, Jaron (2011). You Are Not a Gadget. London: Penguin Books.
Moylan, William (2006). Understanding and Crafting the Mix: The Art of Recording, Second
edition. Oxford: Focal Press.
Rumsey, Francis and McCormick, Tim (2009). Sound and Recording, Sixth edition. Oxford:
Focal Press.
Stables, R., Enderby, S., De Man, B., Fazekas, G. and Reiss, J.D. (2014). ‘SAFE: A System for
the Extraction and Retrieval of Semantic Audio Descriptors.’ In 15th International Society
for Music Information Retrieval Conference (ISMIR 2014).
Taylor, Timothy (2011). Strange Sounds: Music, Technology and Culture. Abingdon: Rout-
ledge Publishing.
Théberge, Paul (1997). Any Sound You Can Imagine: Making Music/Consuming Technology.
Middletown, CT: Wesleyan University Press.
Théberge, Paul (2012). ‘The End of the World as We Know It: The Changing Role of the Stu-
dio in the Age of the Internet.’ In The Art of Record Production: An Introductory Reader
for a New Academic Field, eds. Simon Frith and Simon Zagorski-Thomas. Farnham, Sur-
rey: Ashgate.
Williams, Alan (2015). Technostalgia and the Cry of the Lonely Recordist. Journal on the Art
of Record Production (Vol. 9).
Wilmering, T., Fazekas, G. and Sandler, Mark (2012). ‘High-Level Semantic Metadata for
the Control of Multitrack Adaptive Digital Audio Effects.’ In 133rd AES Convention, San
Francisco.
Zappi, Victor and McPherson, Andrew (2015). ‘The D-Box: How to Rethink a Digital Musi-
cal Instrument.’ In ISEA 2015: Proceedings of the 21st International Symposium on Elec-
tronic Art, Vancouver, Canada.
17
Introduction
The final stage of mixing, and indeed the final responsibility of the mix engi-
neer, is usually the handover to mastering, which brings a number of creative
and technical considerations. In the traditional approach to music production,
mastering is conducted by a specialist audio engineer once the final mixes
have been consolidated to a stereo format. In the early days, mixes would be
recorded to a physical two-track analog tape that would then be shipped to
the mastering engineer. Nowadays it is more common for the stereo mixes
to be sent as lossless audio files through an Internet file transfer. The final
signing-off of the mix is an intrepid point in the process, requiring the art-
ist, music producer and mix engineer to agree that they have completed the
mixes, which can reveal any uncertainties or insecurities that they may bear
in relation to the project. The mix sign-off and handover to mastering is there-
fore seen as a critical and crucial point in the music production process.
Approaches to, and technologies for, mixing and mastering have evolved,
as have all aspects of music production. New methods and approaches
bring opportunities to simplify and reduce the cost of production, although
with the potential for practitioners to inadvertently cut corners and under-
perform in both creative and technical contexts. Modern processing tools
enable mix engineers to also master their own music, and there are a num-
ber of arguments for and against the use of mastering techniques at the
mixing stage. For example, it can be argued that mix engineers need to
take a greater responsibility towards technical attributes such as dynamics
and noise cancellation. Whereas, in contrast, the use of mix-bus limiting
when generating draft listening copies can confuse and falsify the sign-
off process. Furthermore, it may be seen that mastering engineers prefer
or are requested to work from mix stems (i.e., a number of consolidated
audio tracks that collectively make up the mix), but does that mean they
are effectively mixing as well as mastering the songs?
This chapter discusses the critical point of completing the mix and mov-
ing towards mastering, that is, it considers the crucial process of ‘signing
off’ a mix and reaching agreement between stakeholders that a song is
ready for mastering. The discussion draws on the experience and expertise
257
258 Rob Toulson
The final signing-off of the mix is a hugely difficult task, as it is the crit-
ical point in the music production process where the most significant cre-
ative and technical decisions are committed for the final time. Signing off
can expose any uncertainties or insecurities that the artist, producer or mix
engineer may bear in relation to the project. Historically, mixes would be
bounced down through an analog mixing desk to a physical two-track analog
mix-master tape. The tape would then be shipped to the mastering engineer,
who would then manipulate the audio where necessary and cut the master
disc that would be used to manufacture vinyl records. Therefore, the sign-off
process was final—it was far too time-consuming and costly to make any
changes after final mixes were committed to tape. In many ways, the modern
process makes things more flexible. Total recall of ‘in-the-box’ mixes means
that it is very easy to make small changes at any time, and the possibility to
quickly share files over the Internet means that, on the surface, there is little
additional cost or time constraint with making such changes. However, it
appears that the modern process, though much more flexible and reversible,
often makes it harder for artists and engineers to sign off a mix, which can
be considerably disruptive if the mastering has already started.
Mandy Parnell, of Black Saloon Studios in London, is a Grammy
Award–winning mastering engineer who has worked with the likes of
Bjork, Annie Lennox and Aphex Twin, among many others. Mandy
explains that artists and mix engineers now regularly find it hard to agree
on the final mix:
I’ve noticed over time that when clients can hear their music on professional
loudspeakers, they realize there are details they hadn’t heard before, so one of
the trends is that the number of remixes that are presented to mastering has
increased over the years.
Mandy explains that the time and cost implications of this ‘back and forth’
can be substantial:
People just want to tweak. It’s a big problem. We have to take notes on every-
thing; in mastering we are still using a lot of analog tools, so to recall a whole
album is a lot of work if the client has gone back in and tweaked the mix.
George Massenburg, whose career has covered many aspects of music pro-
duction and technology development, earning him Grammy Awards and
production credits with artists Earth, Wind & Fire, Billy Joel and Lyle
Lovett, agrees on the benefits of working with an analog mix setup:
Working on a desk means you are learning the song as it goes by. You mix live
as the song goes by and you’re making a commitment to the levels.
When you work with an analog tape machine, you press play and you listen to
and mix the whole song, and then when it’s done you rewind and press play
again. Nowadays people loop a 10-second section for half an hour, then they
loop the next 10 seconds for half an hour too. I insist that the students who
I teach listen to the whole song. I tell them not to loop, to play if from the
beginning, so when they make a change they know whether it works or not
for the whole song. The rewind time is important too, because you think and
reflect while the tape is turning back. I play from the beginning to the end and
I mix the song. If you hear things 20 times in a loop you get bored and you lose
judgement of its context with the whole song.
George Massenburg explains a wise approach for agreeing and signing off
aspects of the mix as it is in process:
In the very earliest steps of the mix, especially a big complicated mix, the thing
you have to get right is the vocal. To get that right you need the artist to sign off
on a vocal sound before you build the mix around it.
Part of the reason why people are putting off decision-making until the last
possible second is because they are working in environments that make it dif-
ficult to make those final decisions with any confidence. One of the downsides
to people working at home in an uncontrolled environment is that it’s very
difficult to have the same level of confidence as there was years ago when all
the processes needed to be conducted in a professional studio. I think insecu-
rity is part of the equation, it’s not just a lack of knowledge, but, because of
the circumstances in which people are working—and in many cases within the
budget which they are working—it’s very difficult for artists and mix engineers
to say “ok, yeah now I’m sure that it is done”.
Darcy adds that sometimes being too close, or too emotionally attached, to the
music can cloud the sign-off process, and means that moments of accidental
beauty might be discarded if someone has an unattainable vision of perfection.
It is usually therefore beneficial for artists and songwriters to work in teams
with specialist mix and mastering engineers, rather than alone:
Ronald Prent also emphasizes that the reversible processes that have been
allowed by digital audio workstations (DAWs) actually encourage practice
that is counterproductive with regards to decision-making:
I think in an education context we should ask Avid to make a Pro Tools version
that doesn’t have an undo button and only allows destructive recording. I’m
serious, because that’s how we learn to make decisions, but not only technical
decisions, very much musical decisions too. If you add something new and it
doesn’t work, then it’s because what you have recorded is good already and
there’s no space for hundreds of keyboard overdubs and percussion samples
and programming, because you’ve already determined musically what is right.
If you do destructive recording, once you hit record the other content is gone,
so you have to make musical decisions based on technical knowledge—the
undo button killed music as far as I’m concerned.
An important trait for artists to adopt is having courage, owning your own
music, having intent—why are you making it sound like this?
It is clear, therefore, that if artists and mix engineers are brave and focused,
and have reasons for their artistic choices, then this positivity can drive the
record production forward in a successful and productive manner. How-
ever, one further issue at the mix sign-off stage is with artists being unsure
how their mix will translate into the real world of music consumption,
with its unlimited number of playback systems and formats. Ronald Prent
explains:
Signing off the mix has changed in many ways. More and more often, the
artist is not there when you are mixing, and they want an mp3 to listen to after
the session, because that is how it will sound on iTunes or an iPhone. This is
the quality they are used to judging; if they are in the studio listening through
a pair of real speakers, it scares the shit out of them sometimes, because they
hear the detail that we are used to hearing, but they don’t know how to deal
with it.
I stream to my client when I mix live and then they can choose whichever
mobile device, earbuds or computer speakers they want to listen on. I stream
MP4 live the whole day and they can listen in whenever they want; if they are
well equipped I can stream 96/24 lossless too. Artists are starting to demand
that and it really works. The comments I get back are good comments—they
can Internet chat with me in shorthand and I get far better comments, they are
coherent and easy to translate.
In signing off a mix, artists and mix engineers are often drawn to compare
their work with that of other practitioners. This can be a dangerous element
to throw into the critical and delicate process of mix sign-off. George Mas-
senburg recalls:
There is an old aphorism that says “comparison is the route of all discontent”.
Darcy Proper also agrees that making music in comparison to some other
art is essentially a contradictory process:
It’s a funny message that we get from our clients; their music is completely
unique and impossible to characterize and they are very proud of that fact,
but they also want it to sound exactly like whatever is on the radio right now!
Artists should celebrate their music for the fact that it is unique and let every
part of the production process celebrate that uniqueness, rather than trying to
turn it into a song that sounds like something that is already there in the Top 40.
It’s good advice for the mix engineer to get reference material from the artist
at the very beginning of the process. Play me a CD or whatever, give me some
idea where you want to go with this and we can at least tell you if it’s not
possible! If necessary you can tell the artist “you can’t get there from here”,
because that’s not what you have on the recording. But identifying the artist’s
irrationalities early on is helpful!
If artists can give us direction that helps. Some of my clients describe things in
pictures and abstract art, they come in with laptops, show me pictures, play me
music, we talk a lot. I think there has to be sufficient communication; you can’t
just send the mix without the context or a reference point.
If you just mix a project, it’s two or three songs of total guesswork and some-
times you find out after mixing three songs you’ve just been enhancing what
they’ve recorded and carried on with that vision. But it’s totally not what they
want, they want a completely different record, like they’ve recorded acoustic
guitars and nice piano but they want an AC/DC sound. And sometimes I have
to sit there and say, sorry you didn’t record that.
One of the best reference points for me is Songs in the Key of Life by Stevie
Wonder, the song ‘Isn’t She Lovely’, it’s very low volume. It’s better than many
modern references; they often sound crappy and distorted.
The Dreaded Mix Sign-Off 263
Another great reference record is Joshua Judges Ruth by Lyle Lovett, which was
mixed by George Massenburg. That record changed my life. It has the right foot-
print sonically, dynamically, because it’s made with passion for music and audio
and you can put it on in any studio in any place and if you know how that record
is supposed to sound, there you have the best reference for mixing. I grew up
with these records and if you get anywhere close then you’re doing a great thing.
I find especially with young mix engineers and producers, they don’t have a
suitable reference point. The modern references they have are already mas-
tered, a lot of them to a ridiculously loud level, so they are chasing those pro-
ductions and mixing to compete with something that isn’t technically correct.
I’m trying to educate people to listen back to classic recordings. But even some
of the classic albums have been remastered recently, and they are not a great
reference point sonically, in comparison to the records I grew up with.
People are asking me to master from stems, but basically they are just failing
to make decisions or sign off.
Furthermore, the use of stems at the mastering stage means that the evalu-
ation of balance—the relative volumes of instruments—becomes the ulti-
mate responsibility of the mastering engineer, which Mandy points out
requires different listening skills:
Darcy Proper also emphasizes that a different mindset is required for evaluat-
ing balance versus the sonic attributes of the song as a whole once complete:
The reason I’m not a mix engineer is because mixing is a different mindset.
When you are rebalancing things, it’s a very different focus than the fine-tuning
of EQ and compression that’s involved in mastering.
I don’t do stems; that’s not my job as a mastering engineer, that’s mixing. I feel
strongly that if you can’t make the decision then you are not ready for master-
ing. It’s lazy; if you’re hired as a mixing engineer to do the mixes, then mix it,
don’t bring it to me to do your job!
With mastering, there are just two channels. You are asking “how does the
content present itself when it’s finished; dynamically; EQ-wise; fades; spaces;
order?” Mixing is balance. If I’m asked to take stems that means I’m not being
completely objective with the presentation of the music. If I can sit back and
listen to the presentation, then I can develop an opinion on how it sounds and
I can make an appropriate judgement and decide the actions to take from there.
When the mix engineer creates mix stems, an important aspect is to ensure
that the individual stems, when summed together, are identical to the ste-
reo output that was used to monitor and sign off the mix. Unfortunately,
this is very often not the case, as Ronald Prent explains:
If you give the stems for mastering—no matter how well you set up your rout-
ing on your console and how well you manage your gain structure—if you add
the stems up, it’s not the same mix, because it doesn’t use the processing that
you use at the end to make your mix.
Darcy Proper agrees; although she is willing to work with stems if neces-
sary, the best results come when a mix has been properly signed off:
The projects that have turned out the best in my career, they have all come
from stereo mixes, not stems. Of course stems have an important place in film
and game music, but, most of the time, the sum of the stems doesn’t equal ‘the
mix’—it’s never quite as simple as summing up the stems and then playing
those all off at zero.
The argument that stems allow different aspects of the mix to be mastered
separately is not one that simplifies the process, indeed it makes mastering
much more complicated, as Darcy explains:
If they haven’t submitted a complete mix, then what’s often needed is far more
complicated than just adjusting the individual stems by a certain level for the
whole piece. I may likely need to look across the whole song, raising vocals
during the chorus and dropping them maybe in the verse. It really is mixing and
it’s very different to what we should be doing at the mastering stage.
The Dreaded Mix Sign-Off 265
Does anybody like stems? I hate having to send stems for mastering. It’s not
as simple as people think. For example, if the A&R department gets back to
you and says “well, you got to bring up the bass”, it’s a lot more than that; you
have to add the bass and a little more kick and whatever else is in the low end,
the low end of the piano, the low end of the voice, everything changes. These
things do not exist independently, and the fiction of stems is that you can allow
the record company to make what is in their mind an improvement, and they
are just idiots, it’s not an improvement.
Working with stems is basically a means for A&R to get their hands on your
record and to control something about the record and, most importantly for
them, to own a part of it. It allows insecure A&R men to take ownership of
something.
We are asked to print stems just in case. It’s usually the client, artist, producer
or in some cases record companies that demand stems for remixing purposes.
But they don’t want to pay any more, or they want to give your stems to some-
one else to remix, which I find very underhanded.
There are cases where with the right person who knows what they are doing,
with the right perspective and intent can take both roles. But it is decidedly
two hats, because your objectives for mixing and mastering are completely
different. It’s not simple to do—putting something on the output bus of a mix
and saying that is mastering—that’s not mastering, that’s something completely
different, that’s finishing the mix.
266 Rob Toulson
This is art and art is expression. We have a left brain and a right brain, and I think
of it like a pendulum. The brain has a creative side and a technical side and if
you’re the artist and you’re going to try to record and mix yourself, you think
“I’ve got a great idea for a song, now I’m gonna open up a preset on a plug-in and
connect this thing in here, wait a minute, what was that lick again?” I feel like the
pendulum never swings nearly as far as it should on either side because it’s con-
stantly going back and forth and you’re not getting your full artistic expression,
nor your best technical level of capturing that. So consider hiring professions
throughout the chain to be able to do the best they can to represent that art.
Darcy Proper also emphasizes the difficulties with taking the responsibil-
ity for both mixing and mastering a record:
As the artist or mix engineer, by the time you get to the mastering process you
are in one of two states: Either you feel that it’s absolutely your best work and
you may be tempted to believe that it’s perfect and doesn’t need further master-
ing, while a few minor adjustments could really make it shine. Very often the
opposite is true too, in that it doesn’t excite you anymore, because you’re just
tired of it. Then you start fixing things that aren’t broken, and when you start
fixing things that aren’t broken you are actually doing harm and damage. That’s
one of the advantages of having another person involved who doesn’t bring any
baggage to the session.
I’ve seen people say “I’m gonna start a song and I’m gonna put a couple of EQs
and a compressor on the master bus and I’m gonna mix the song”. That to me
is like saying “I’m gonna start with a cup of salt and I’m gonna now make me
some soup!” Do either of those things make any sense to anybody?
When someone says to me they are going to master a track by adding some
EQ to the master bus I always go back to the question ‘why?’ And if they say
it’s because the guitars are not bright enough—well fix the guitars, work on
that. That’s not mastering, that’s just fixing a mix and for us to blur those lines
between mixing and mastering is dangerous for the art form. As a mastering
engineer I might not put any EQ on it, I might run it through a valve processor
and realize it doesn’t need any EQ at all.
George Massenburg and Mandy Parnell emphasize that there are certain
cases where the mix engineer might master their own music. George states:
There are occasions where there’s barely a budget for mixing, let alone master-
ing, so I’ll do some EQ and a little bit of dynamics on the bus. But if I run into
The Dreaded Mix Sign-Off 267
a wall I go back and change the bass level or I change the kick if it’s punching
holes, I’ll pull it back a tad. I think there is a case to be made for doing that with
sensitivity and objectivity sometimes.
It works maybe for some kinds of electronic music and some metal which is
very aggressive. I think there is a place for it, but most of the time I’d say it
should be avoided.
Darcy Proper has practical advice for producers working on a tight budget,
where the mixing and mastering must be conducted by the same engineer:
My advice is don’t mix and master in one step. Finish your mix, print your
mix, then take that mix and master it as a separate process, not within the same
DAW session. If you do have to master it yourself, take a break, take some time
between the mixing and the mastering. It should be two separate steps.
One unique and regularly prevalent issue is that review mixes are rarely at
the loudness of commercial music, so clients and artists struggle to evalu-
ate the mix, given that it will be much quieter than mastered songs that are
heard on the radio or in the charts. In this case, it is often necessary to sup-
ply ‘listener copies’ of the mix that have been raised in loudness with some
form of compression or limiting. Listener copies made in this way cause
much confusion for artists and engineers, so it is important to have a clear
and justified method for creating the listener copies. In addition to Darcy’s
previous advice about mixing and mastering as separate steps, she adds:
If I could say one thing to every student or young mixer out there, it would be
“don’t mix through any loudness software”. It is understandable that in order
to get client approval—as many lack the imagination to understand how a mix
might sound when it is mastered—you can give them a loudness-enhanced ver-
sion, but don’t mix through the software that makes it loud. Make a good solid
mix and then right at the end put the loudness tool across it and keep it as close to
the intention of your mix as possible. That way you can give the mastering engi-
neer both the reference version that you have sent to the client and you can also
provide a finished mix that doesn’t have that extra loudness maximization on it.
Mandy Parnell also discusses the issue of mixing with limiters on the mix bus:
There is a problem with this listening copy model, that mix engineers are being
pushed to use limiters because artists and record labels want to hear loud mixes
for review. The problem with mixing into limiters is that the final mix sent for
mastering then has no relation to the listening copy, which is distorted and
noisy. But that’s what everyone has signed off from, and that is our starting
point. Many times I have to go back and ask them to put a bit of the listener
limiting back on in order to get the sound they are after, but still it’s too loud,
and I’m starting from such a wrong place to master the record. I’d like to see us
getting away from that, but it’s about education and people have been working
like this for maybe ten years now.
268 Rob Toulson
Darcy Proper also adds that any mix bus processing can actually mask
issues that will still need to be resolved at the mastering stage, whereas
they could have been treated better during mixing:
The mastering engineer acts as a quality control expert for the final stage
of a project, but can also act as an invaluable mix consultant, too. George
Massenburg explains that he would regularly seek a mastering engineer’s
opinion throughout the mixing process:
Doug Sax had a unique knack for, through mastering, moving the irrational to
the practical. For many years he was our great advisor who we would talk to in
the middle of the project, take him rough mixes and he would have no problem
telling us if it sucked, or to do nothing.
Mastering engineers really need to advise producers and mixers more nowa-
days. I want to get a really great mix in, so it’s in my interest to get a good mix
first than to try to work miracles.
Something I’ve started doing in the last five years, if they have the capability of
going back and mixing in the box, I will send them back to remix rather than
try to polish it in mastering. I give them a lot of direction and listen to lots of
references with them and explain what they need to try and do with the mix,
where it’s going wrong.
We can be objective and have perspective and bring a fresh set of ears. When
I hear a song for the first time I’m hearing it as a listener would be hearing it
for the first time, and I know I’ve got it right when I go from not knowing the
song and not caring about it or having any feeling for it, to having dialed into
the emotional content to the point where I have the goosebumps and I’m emo-
tionally connected to it.
The Dreaded Mix Sign-Off 269
My first perspective when a client comes in is the most important thing that
they are paying for. My first listen, my first response, my first notes. The feed-
back they get from that point is the most important I can give, especially if the
mix is wrong.
I say always attend the mastering session and build a relationship. I tell young
producers to pick a mastering engineer and stick with them for a couple of
years or projects. Use them as a sounding board so you can develop your mix-
ing and build up an understanding.
Conclusions
Clearly, the boundary between mixing and mastering has become gradu-
ally porous in recent years, with digital technologies and the availability
of home production systems creating an interesting circumstance which
allows anyone to effectively take the role of mixing or mastering engineer,
or both, at any given time in the production process. These advancements
have in some way enabled us to cut corners with the production process
and reduce the number of expert ears that might be involved in the record-
ing, mixing and mastering. The opportunity to put off decision-making to
a later stage in the production is enabled by technology and the possibility
to master from stems, for example. However, this trend actually appears
to lead to greater indecision and insecurity in artists and engineers. Mas-
tering engineers are wise and experienced listeners who can assist with
mixing as well as mastering, and it is therefore beneficial to utilize them in
any professional project. Ultimately, all studio engineers are here to assist
the artists and to enable them to most accurately achieve their creative
vision for a record, and any professional recording, mix or mastering engi-
neer will be passionate about helping the artist realize that goal, as George
Massenburg humbly concludes:
At the end of the day, it’s their record. Their picture is on the front and we are
here to serve the artist and at the pleasure of the artist—and if we don’t get
it right we have every right to be fired. I’ve tried to do everything an artist
ever asked me to do and sometimes I learned something and sometimes they
learned something. We are here to chase down their vision.
18
Conclusion
Mixing as Part-History, Part-Present
and Part-Future
Russ Hepworth-Sawyer
In the introduction to this book, we discussed our intention that this book,
and indeed the Perspectives on Music Production series as a whole, should
serve as a multi-perspectival investigation into the history, present and
future of the production of music for auditory consumption.
Many of the articles in this book explore historical and present incarna-
tions of the mixing process. For this, the concluding chapter, we felt it best
to extract and consider the future implications of these contributions. We
also felt that to indulge me in a few of my own observations, and some of
my predictions, would be useful to conclude this introductory and explor-
atory volume.
History
270
Conclusion 271
level of knowledge might have only been reserved for those with the title
of engineer at a later stage in their careers. Their roles are, and continue
to be, to develop and change in tandem with technological developments.
In today’s contemporary music studio, the assistant needs to understand an
ever-widening range of equipment for those clients, such as me, who visit
Abbey Road for the odd day without the tacit knowledge or experience of
that particular room.
To further understand these changes in roles, it is worth us revisiting the
changes to the equipment over the generations. Our rich heritage in audio
engineering had its own pioneers who progressed our art forward with
each new technique and tool. As almost each new development emerged,
seized upon because of perceived benefits to the quality of the audio, new
roles would often be created. For example, the creation of magnetic tape
and its adoption within the studio environment meant that no longer were
engineers capturing the performance that would effectively serve as the
master for production (production here meaning manufacture of the listen-
er’s product). As soon as magnetic tape made an entrance, it was necessary
for a new stage, a new task, whether undertaken by the same person as the
recording, whereby the audio material needed to be transferred or ‘cut’ to
vinyl.
Similarly, as compact disc (CD) became the new format for everyone
to jump on to in the 1980s, new roles specializing in digitizing analog
audio appeared in all major studios to ensure they can translate their past
catalog masters to that medium. A new role emerged ensconced in Sony
PCM1610’s and U-Matic tape machines. The trend can be analyzed fur-
ther with the adoption of Sonic Solutions to enable the first widespread
non-linear system for CD compilation.
The continual metamorphosis of the audio engineer throughout the ages
has led us to a huge diversity of engineering duties, yet concurrently one
person could, once more as in the very beginning, fulfill all roles neces-
sary. Burgess (2014) observes that in the early days of recording the “mix
capability was limited to positioning the musicians, instructing them with
respect to dynamics and tone control, and guiding the vocalists in their
proximity to the collecting horn”. The notion of the mix being ‘complete’
at the time of the recording is paralleled in the classical conducting of
an orchestra. Terrell et al. (2014) note that an orchestral conductor has
the equivalent role of the mix engineer: “A critical control parameter is
acoustic intensity, which the conductor controls through instructions to
the musicians and the mixing engineer controls through adjustments to his
mixing interface”.
‘Mixing’ as a term in itself would not have been considered so before
the dawn of the discrete stage. ‘Balance’ would have been the more appro-
priate term. Recordists prior to any form of multichannel audio would
have considered the blend of the instrumental performance and moved the
musicians closer to the transducer to ensure the required outcome. The
producer’s role could be arguably a reinterpretation or extension of the
word ‘arranger’ as the instruments were arranged around the room. This
272 Russ Hepworth-Sawyer
was a strategy that Phil Spector continued at times even as multitrack tech-
niques were prevalent to great success.
The rise of the engineer as personality perhaps began with Joe Meek,
who made a name for himself within IBC studios during the 1950s. Meek
was perhaps one of the first engineers to have what Gary Bromham1 notes
as a ‘sonic signature’—a sound desired by artists unique to that engineer.
Bromham also coins the phrase ‘sonic trend’ to describe the marketability
or ‘brand’ of an engineer. Meek became a highly respected and sought
after engineer for achieving chart success and a unique sound that made his
records stand out. Meek quickly and inquisitively assessed and understood
the qualities that made a good song translate well over vinyl in jukeboxes
and on radio. Meek in those days would have been expected, and he him-
self relished, the need to work on each aspect of the project from recording
to the ‘mix’. Meek’s propensity to secrecy for fear of espionage of his
techniques (from fellow engineers) made him less than collegiate within
IBC. The sonic ‘brand’ of Meek, of course, even within IBC, became well
known with his own built equipment, such as the spring reverb unit made
from a fan heater. Meek also made pivotal limiting devices to improve the
sound of the audio. Despite the multitrack dream, much larger track counts
of eight or even sixteen was to be over a decade away. Joe Meek shone the
way forward with his enormous and sustaining influence on sound produc-
tion per se, shaping many of the mixing engineer’s ‘given’ techniques that
persist to this day. In many ways, Meek was one of the first superstar mix
engineers.2
Meek was also a pioneer within the UK studio scene by being one of the
first engineers to leave a major studio business to go it alone. Meek, whilst
not technically freelance as the term would later be known, did work in
his own studio in his flat in Holloway Road. It would be, again, a decade
away before engineers could start to be independent of studio employment
in the mainstream.
As multitrack recording started to become the norm for most profes-
sional recording studios in the 1960s, again new skill sets were emerging
because of the technology. Engineers, certainly in my native England at
that time, were most often employees of one of the major studios of the
day, whether that be IBC, Abbey Road, Apple Corps or later Trident, for
example. Engineering roles in those early days were more a rite of passage
that saw their training start in the cutting or transfer room. Engineers (or
even producers later in their careers) may all have started at this point. The
belief was that recording engineers ought to understand the limitation of
the medium to which they would record. It was common to move between
studios and projects within the building as an employee also. So to record
or mix any given project would not have necessarily been of concern. Art-
ists and producers would not necessarily question the ability of any engi-
neer in mixing or recording employed by the studio. This, in the most part,
was accepted practice. So to assert that there was any consistent delinea-
tion between a recording and mix engineer to any great extent, certainly in
the UK, would have been less than common.
Conclusion 273
The Present
The mix has always been borne out of the need to achieve balance: to
arrange order from chaos. Mix engineers today have, in many cases, a
more complex role than they once had. Bob Clearmountain in the 1980s
in many cases may have just received an Ampex 456 two-inch track tape
with 24 tracks on which he’d need to mix. He’d ensure (where budget and
deadline provided) that he’d have ample time mixing through the installed
console in front of him and the outboard processing around him, mindful
in the knowledge that the mix needed to be committed there and then. To
recall the mix (either by memory or by the primitive total recall systems
of the day) was many hours’ work. Similarly, due to the relatively limited
274 Russ Hepworth-Sawyer
update”. Nelson also shares the frustration felt when he too falls into the
trap of the distracting plug-in search. Engineers of a previous generation
used the equipment available to them on the whole: the EQ within the
desk, the few compressors in the rack and the plate reverb or echo cham-
ber. Records were made and achieved. In many ways innovation, and tech-
nological advantage, has slowed the music creation process.
The outcome or expectation of standard stereo mixing remains rel-
atively unchanged to that of few decades before, despite key techno-
logical changes. Formats, however, have come and gone. Digital Audio
Tape (DAT) once used to deliver two-track masters after open reel tape
became suddenly considered ‘old school’. CD-R later became an option
for many as the players and media became more reliable and widespread.
Today, mixes are delivered by audio file most often over the Internet.
Two major areas of change have been that of surround sound mixing and
high-resolution audio. Surround sound mixing has been on the agenda
ever since Quadraphonic in the 1970s. Despite enjoying huge acceptance
within the film community, encouraging music consumers to adopt sur-
round sound has been difficult.4 This is due to a whole raft of issues,
but mostly with the playback. Consumers’ living rooms were not always
of an ample size and optimum shape to perfectly situate five full-range
monitors based on ITU standards. Then, of course, there’s the expense
of this. Surround sound for music was not easily possible at the desired
resolution until Super Audio Compact Disc (SACD), which was adopted
slowly due to cost of both the machine and the discs.
Since the launch of the SACD into the mainstream, there has been a
slow progression towards listening to music on headphones, due to the
prevalence of the mp3 player and most significantly the iPod. This has
meant that mixing has never been so far removed from the consumer main-
stream possibility of surround sound, as headphones could not deliver true
5.1 surround.
As the iPod made significant changes, alongside the iTunes plat-
form, engineers had to ensure their mixes and masters would translate
both to the still-popular CD and also to the highly data-compressed
mp3. Convenience had won over quality. While the humble CD has
stuck in there, we’re now faced with many format options when con-
sidering buying a new release. The CD may remain (although this for-
mat is often being missed off the checklist with current releases in
my studio), mp3 will be a potential upload to an aggregator, but more
commonly some form of high-definition file, most likely Mastered for
iTunes (MFiT), will be uploaded at 96 kHz/24 bit, which is discussed
more a little later.
However, the format to ensure is mentioned at this juncture is the hum-
ble LP, often released even when CD is not. Vinyl has made a significant
comeback in the past few years and whilst not of immediate interest to
the mix engineer, there are a number of reasons that the future of mixing
might permit different working methods. The present has become some-
what uncertain as to what the future will hold.
276 Russ Hepworth-Sawyer
Part-Future
In the recent past, one might be able to argue that there has been a techno-
logical slowdown when it comes to music production development commer-
cially. Computing power has become ample for large-scale, high-resolution,
multitrack projects right from a notebook computer. On the face of it, few
game-changing developments are commercially available for mixing.
Recent developments, however, do show signs of the future. Touch-
screen technology, currently being championed by the Slate Raven, starts
to demonstrate that mixing by mouse may become less prevalent and that
there could be a return to console-style mixing. The relative cost of mix-
ing using a touch surface such as the Raven is significantly lower than an
Avid C|24 for the same Pro Tools functionality. It is likely that this area
may develop further, meaning that more engineers will stop sitting at right
angles to the studio’s SSL or Neve that only has two channels up panned
hard left and right.
The main game changer is highlighted within this book. Joshua Reiss’
work at Queen Mary University will slowly pave the way to true auto-
mation of many of the functions of mixing. It is unlikely that automated
mixing using intelligent systems will ever take away the gloss an accom-
plished mix engineer can apply, but it is expected that a number of the
routine editing functions and application of processing on the channel strip
could be automated. Not automated by selecting just the Logic Pro chan-
nel preset, but actually the system appreciating what the sound is, how it
sounds and responding as though you had made the decisions yourself.
This could save considerable time and leave more time for mix refinement
by the engineer. As this book was going to press, iZotope announced the
development of Neutron, which starts to offer mix automation based on
the material played into it. It is anticipated that these intelligent system
developments, as with Reiss’ initial involvement with LandR.com will
slowly become more commonplace in the future of mixing music.
LandR currently only offers standard resolution masters at a time when
the industry has the option to move to higher resolution. One of the major
changes, intertwined with the development of technology, or the lowering
cost of hard drives, will be a common system resolution above current
CD/PCM standards of 44.1 kHz/16 bit. Many professionals record at 24
bit, but many have yet to make the move permanently to 96 kHz or above.
Given MFiT, HDtracks.com and devices such as the PonoPlayer, there
is a slow, but certain, move back to high fidelity playback. This has also
been given the shove to the forefront by the masses of people returning (or
young people flocking) to the turntable and vinyl.
While technically a mastering engineer’s issue, equal loudness stan-
dards are being adhered to in the broadcast industry much faster than ever
thought would be the case. EBU’s R-128 standard has been widely adopted
and is playing its part in broadcast. Similar principles are being employed,
but for the playback of material within software, whether that be a vol-
ume-levelling system within iTunes or now YouTube. Mastering engineer
Conclusion 277
Notes
1 Gary Bromham notes ‘sonic trends’ within his chapter for this book and expands this to
the ‘sonic signature’, which is related in principle to ‘watermarking’ as discussed in
Hepworth-Sawyer and Golding (2010).
2 It must be noted that Joe Meek was much more than just ‘a mix engineer’. Meek was per-
haps the first superstar producer.
3 Sir George Martin had already made moves away from EMI as a staff employee, acknowl-
edging his potential as a freelance producer.
4 We acknowledge that we’d have wished for more representation of this within the book, but
the contributions were not forthcoming.
5 Production Advice—http://productionadvice.co.uk/youtube-loudness/ accessed 29/04/2016
14:16.
Bibliography
Brown, P. (2010). Are We Still Rolling?: Studios, Drugs and Rock ‘n’ Roll—One Man’s Jour-
ney Recording Classic Albums. n.p.: Tape Op Books.
Burgess, R.J. (2014). The History of Music Production. Oxford, Oxford University Press.
Clarke, E., Cook, N., Leech-Wilkinson, D. and Rink, J., eds. (2009). The Cambridge Compan-
ion to Recorded Music. Cambridge: Cambridge University Press.
Cleveland, B. (2015). Joe Meek’s Bold Techniques. Nashville: ElevenEleven Publishing.
Day, T. (2000). A Century of Recorded Music, Listening to Musical History. London: Yale
University Press.
Emerick, G. and Massey. H. (2006). Here, There & Everywhere: My Life Recording The Beat-
les. New York: Gotham Press.
Hepworth-Sawyer, R. and Golding, C. (2010). What Is Music Production? Boston, MA: Focal
Press.
278 Russ Hepworth-Sawyer
Hodgson, J. (2010). Understanding Records: A Field Guide to Recording Practice. New York:
Continuum Books.
Horning, S. Schmidt (2013). Chasing Sound: Technology, Culture & the Art of Studio Recording
from Edison to the LP. Baltimore, MD: John Hopkins University Press.
Kealy, E.R. (1979). From Craft to Art: The Case of Sound Mixers and Popular Music. Work
and Occupations 6 (1): 3–29.
Millard, A. (2005). America on Record: A History of Recorded Sound, Second edition. Cam-
bridge: Cambridge University Press.
Shepherd, I. (2016). YouTube Just Put the Final Nail in the Loudness War’s Coffin. Available
at: http://productionadvice.co.uk/youtube-loudness/ [Accessed: 29 April 2016, 14:16].
Terrell, M., Simpson A. and Sandler, M. (2014). The Mathematics of Mixing. Journal of the
Audio Engineering Society 62 (1/2): 4–13.
Index
Note: Italicized page numbers indicate a figure on the corresponding page. Page numbers
in bold indicate a table on the corresponding page.
279
280 Index