Mixing Music

Russ Hepworth-Sawyer, Jay Hodgson, Russ Hepworth-Sawyer, Jay Hodgson

• Publisher: Routledge
• Year: 2016
• ISBN:
  9781317295501
• Edition: 1
• Topic: Media & Performing Arts
• Subtopic: Music Theory & Appreciation

This series, Perspectives On Music Production, collects detailed and experientially informed considerations of record production from a multitude of perspectives, by authors working in a wide array of academic, creative, and professional contexts. We solicit the perspectives of scholars of every disciplinary stripe, alongside recordists and recording musicians themselves, to provide a fully comprehensive analytic point-of-view on each component stage of record production. Each volume in the series thus focuses directly on a distinct aesthetic "moment" in a record's production, from pre-production through recording (audio engineering), mixing and mastering to marketing and promotions. This first volume in the series, titled Mixing Music, focuses directly on the mixing process.

This book includes:

• 
  
  

• References and citations to existing academic works; contributors draw new conclusions from their personal research, interviews, and experience.

• 
  
  

• Models innovative methodological approaches to studying music production.

• 
  
  

• Helps specify the term "record production, " especially as it is currently used in the broader field of music production studies.

1

Exploring Potential of the Mix

Historical Milestones and Expanded Perspectives

Martyn Phillips

Introduction

Working beyond the plentiful suggestions from experts, copying others or trusting their art to happenstance, how can mix engineers find their own unique way of deepening their own craft? Is it possible to find new approaches? This article looks at conceptually what is possible in this regard and suggests some practical ways of achieving this.

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 significance of the increasing finesse that can be applied to the process of accessing 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 represented audio was created, committed to a medium of soot on paper. Parisian Édouard-Léon Scott de Martinville recorded an incomprehensible 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 Laboratory 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 the signal and noise, something that can be addressed with current technology (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 musicians 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 conceptual 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 beginning; ‘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. Recording multiple vocal parts inspired both controversy and inspiration. Richard 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 being considered important by Snyder (Petersen, 2005). While monitoring previously recorded parts via the record head (henceforth known as the sync head), additional recordings could be made it time with them. Earlier recorded parts could be individually re-attempted or patched up with reference to what had been originally later recorded ones; with the Sound on Sound technique, subsequent recordings would have been lost.

The development of Sel-Sync development had another crucial significance: recordings could now be to be separately manipulated in a number of ways, such as attenuating or equalizing. Parts could be re-contextualized or even discarded.

The process of deferment of the final mix from the recording of the parts starts with Les Paul and the forward planning of his Sound on Sound recording. In Les Paul’s case, this might be the time it took his wife, Mary Ford, to cook macaroni cheese for their dinner but can now span to decades as has been the case with the Beatles’ ‘Fly Like a Bird’ (Roger, 1994). This ability to be being able to examine the work at different times is not trivial; perspectives change as the listening environment does, also with the weather, the time, the cultural context. A mix then can make sense in many different environments and at different times has the potential to resonate with a greater number of listeners.

In more recent decades, the use of automation has enabled the engineer to refine a process in a number of different ways, to go beyond relying on manual control or the time constants of outboard processes to effect dynamic changes over time and to apply a number of simultaneous changing processes to a single part.

Up until the advent of digital, there was a one-to-one correspondence between all links in the audio chain: sound pressure, microphone capsule displacement, electrical current, capacitor charge, tape magnetization and speaker cone displacement could all be mapped to each other by some bijective function. The introduction of digital techniques in audio, such as random access, digital filters and other mathematical techniques, enabled this correspondence to be broken. A raft of new processes of manipulation could now be employed. For example, Melodyne Editor plug-in’s Direct Note Access, created by Peter Neubäcker, is able to change individual notes within polyphonic audio, something that would have been undreamed of just a few years ago (Celemony, n.d.). iZotope’s RX5 software editor can access individual harmonics or areas of the spectrum within complex audio for processing (iZotope, 2016). Other processes that go beyond the traditional techniques of sound manipulation include Sound Radix’s Surfer EQ plug-in (Sound Radix, 2015), which can track harmonics within a melodic monophonic part, and Pi plug-in, which can access hitherto opaque phase relationships between different parts to create a more phase-coherent mix (Sound Radix, 2015).

Artificial intelligence is now increasingly being used, not only for composition, but also for audio manipulation and can be utilized in the span of the recording process from reducing spill on drum kit microphones with Accusonics’ Drumatom (Accusonics, n.d.) to the mastering of the final stereo mix with LANDR (LANDR, n.d.).

The scope of what is possible in the mix is practically unlimited. Given sufficient artistic permission, any audio result can be produced. It has not been unknown for re-mixers, the author included, to have work accepted that has nothing of the original recording left, a metaphorical spade with a new blade and a new handle. Whether it was his intention or not, the new possibilities that Les Paul created for the mix engineer have developed sufficiently so that the job is comparable to that of the performer.

The Platonic Mix

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 actually 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 something 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 sublimely perfect, joyous and beat-less melody, an orchestra of voices” (Williams, 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

Can these revelations of sublime music be authentically considered perfect? Is the concept of a perfect mix useful?

A possible solution to whether perfection is in fact possible comes from the Austrian-born logician, mathematician and philosopher Kurt Friedrich Gödel, whose two incompleteness theorems, which discuss the limitations of mathematical systems, have implications beyond the purely theoretical. Smith (2013: 3) reports that “Gödel’s first incompleteness theorem shows that the entirely natural idea that we can give a complete theory of basic arithmetic with a tidy set of axioms is wrong”. He goes on to summarize the second theorem thus: “nice theories that include enough basic arithmetic, can’t prove their own consistency” (Smith, 2013: 6).

Recent research at University College London on the spectral gap (Cubitt et al., 2015) has demonstrated that Gödel’s theorems do indeed have real-world effects, that certain material properties are ‘undecidable’—they are neither true nor false (Knight, 2015).

Gödel’s proofs have also been extended beyond the realm of mathematics 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 produced, a new perception can be applied to it, which negates its perfection.

Artistic Exemplar

So why could it be that the music experienced in other states of consciousness come across as so perfect if they are not actually so? Williams’s NDE experience and following research (2014) leads him to comment,

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.

These experiences are rich in synesthesia, the involuntary stimulation of one sense by another and comparable links between the sensory experience and emotional responses. Perhaps these strong emotional links persuade the witness that what is being heard is on a more sublime level that anything in the physical realm. Alternatively, could it be that the emotions being experienced are more truly represented by the melodies, arrangements and textures experienced in these rarefied states of consciousness than by most of what can be created from everyday consciousness?

Wherever this music may be generated, some paragon of a completed work, even if not perfect, with associated emotions, does appear to be accessible, apparently instantaneously, to the human mind. By acknowledging its existence, even if not readily available, the mix engineer can view the process of mixing as a process of discovery as opposed to one of creation. Without entering an altered state of consciousness or having the refined discipline to mentally construct a complex audio artifact, how might this objective be revealed?

Unlike the musician, who must distill what they can of their lifetime’s worth of art into the moment’s performance, the mixer’s art does not need to be similarly condensed. The same piece of work may be sequentially revisited, thanks to Les Paul’s legacy, from a multitude of perspectives and interrogated using different mindsets, intuitions or emotions, here collectively termed as perceptions. It is not necessary to simultaneously be aware of all the various aspects of the artifact, but each dynamic can be sequentially addressed.

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 ...