Recreating an Unreal Reality: Performance Practice, Recording, and the Jazz Rhythm Section

Introduction

This paper will discuss various topics relating to the recorded sound of jazz rhythm sections, with special attention given to how consumers of these recordings (specifically theorists of jazz rhythm and rhythm section musicians themselves) engage with the musical surface presented by jazz recording. The focus of discussion is on ‘live’ (live in-studio, or live on stage) recordings of ‘acoustic’ (using acoustic instruments and little or no in-studio or on-stage sound reinforcement) jazz. The first two sections will examine issues in recording technique and the ways that recording alters the sound of the rhythm section. The third section will discuss reception of such recordings as documents of live rhythm section performance, and relate this reception to Charles Keil’s discussion of ‘participatory discrepancies,’ Vijay Iyer’s 2002 paper on microtiming in African-American music, and some recent theoretical studies of timing in jazz music, especially the work of Matthew W. Butterfield (2010, 2011) and the team of Anders Friberg and Andreas Sundström (2002).

Bass and drums form a distinctive unit within jazz ensembles. In jazz recordings where bass and drums are both present the instruments are played together most of the time and work in a kind of polyrhythmic unison to create the signature swing and ‘straight eighth’ grooves of jazz music. This close relationship means that, as both developing and mature players, bassists and drummers must simultaneously consider their playing individually and as members of a tightly paired musical team. Within the community of jazz players, scholars, and fans these rhythm section dyads receive considerable attention, and practically every notable jazz ensemble has, at its core, a fabled rhythm section whose tropes and timbre are carefully scrutinized by serious listeners. Theorists and rhythm section players alike listen to recordings of noted rhythm sections using extreme precision, at times even using digital audio techniques to parse recorded performances at the millisecond level. The practice of transcription has been an established aspect of jazz connoisseurship since the late 1930s (Potter: 1990, 64) and serious students of jazz are expected to undertake considerable aural and notation-based transcription of canonical improvised solos (Baker: 1979, 36-48)

Jazz recordings play an essential role in the process of becoming a rhythm section player, and rhythm section scholars and practitioners uniquely modulate their use of recordings. Unlike other instruments in the jazz ensemble, bass and drums share a stylistic timekeeping role that demands their sounding be encountered both as two instruments and as a cohesive unit. In live playing situations listeners and players share acoustic space allowing participants and observers to use spatial, aural, and visual cues to parse the rhythm section’s activity. Since accurate transcription of live performance is impossible without recording, however, rhythm section musicians need recordings to create transcriptions and effectively learn performance technique. Recordings reveal sound based on microphone location and the multitude of factors that influence all recordings from original recording equipment to playback speaker settings. They reproduce a peculiarly situated analog of the sonic events present at the moment of recording. This disconnect between performative moment and recorded artifact has serious implications for both rhythm section musicians and scholars of jazz.

The Jazz Rhythm Section on Record

The modern drum kit evolved partly in response to the needs of jazz musicians; its range of timbre and textures are designed to ideally represent jazz rhythm. Through the middle and late 1930s various players, most notably Kenny Clarke, began to focus much of the timekeeping and rhythmic drive of their drumming on the ride cymbal while distributing other rhythmic elements to the hi-hat and snare drum (Brown: 1990, 42-45). This progress in drumming style was supported by a new school of pizzicato string bass playing, typified by Milt Hinton and Oscar Pettiford, that used two or four beat groupings along with characteristic rhythmic interruptions (sometimes called ‘skips’ and ‘drops’) to connect rhythmic and harmonic aspects of the musical texture. By the middle 1940s bebop rhythm section style had fully evolved into a transcendent combination of instrumental sounds that opened new avenues for musical exploration within the ‘black music tradition’ and completely fulfills the ‘predilections for conceiving music’ within that tradition outlined by Olly Wilson (Wilson: 1992, 330).

Although rhythm section interplay and vocabulary was well established during the 1940s, jazz recordings from this period tend to have a lack of focus on the bass and drums. It was not until about 1950 that recording technology and technique advanced and the definitive recorded sound of the jazz rhythm section developed. Noted recordist Rudy Van Gelder engineered many of the seminal recordings of modern jazz. An early adopter of new technologies, Van Gelder was among the first people in the US to obtain Neumann U-47 condenser microphones. This ‘industry standard’ condenser microphone, the first to have a switchable (omni-directional to cardioid) capsule, was designed to capture ensembles from a distance. Working with audio engineer Rein Narma, Van Gelder modified the U-47’s amplifier so that the microphone could be placed close to instruments, capturing great detail.  By 1952 or 1953 Van Gelder was using several U-47 microphones at close range to capture the sound of small group jazz (Skea: 2001, 61-62).

This close-microphone approach to recording jazz has since become standard practice. While recordings of classical music often employ reverberant spaces, jazz recordists starting in the 1950s and continuing to today prefer to use ‘dead’ spaces and close microphone placement. In the hands of an expert engineer this can lead to beautiful, intimate, and detailed recordings of the rhythm section. Bass and drums are each distinctly audible yet blend, to use drummer Billy Higgins‘ term, like a ‘family’ (Stern: 2000). Through the 1950s and 1960s Van Gelder and others recorded live on-stage and in-studio performances of defining figures in jazz music, including all the most respected rhythm sections, and this body of recorded work comprises the core literature of jazz music. Musicians, scholars, and connoisseurs alike refer to jazz recordings from this era frequently, and many of the most notable recordings have been re-issued, repackaged, and remastered several times.

While it is generally agreed that the Van Gelder approach to recording ensemble jazz is close to ideal there has been less discussion of how this production style affects the expectations of rhythm section musicians and other participants in jazz music. The discrepancies between recorded artifacts and their live performance precursors are also rarely engaged. As Simon Trezise points out, recordings are not ‘mimetic’ of the performance that yielded them. ‘The record and associated equipment are telling us about a performance, but it is not the performance itself; it is filtered through a large number of processes and contexts with which the original performer has nothing to do’ (Tresize: 2009, 207). In the case of canonical jazz recordings, which function as the type and template for developing players and scholars of jazz, this lack of integrity between performance and recorded artifact is especially notable.

How Common Jazz Recording Techniques Distort the Rhythm Section Sound

Looking at the rhythm section, there are several issues that problematize the use of recordings to understand the interplay of jazz musicians. One is the issue of microphone placement. Given that a good recordist would work to find the ‘ideal’ placement of two or more microphones to capture bass and drums, it’s notable that the bassist and drummer themselves (at least until isolation and headphone mixes became more common in the 1970s) would not hear the microphone’s perspective. Instead their experience of the musical texture would be uniquely situated based on their own position and aural awareness. For example, Gunther Schuller demonstrated that legato or detached bass articulations have a distinct affect on their perceived ‘swing’ (Schuller: 1989, 855-859). While aspects of Schuller’s work on this topic have been criticized (see Prolger: 1995, 29-31 for an overview), there is a consensus among bassists and other jazz musicians that the degree of connectedness from one pitch onset to another is a key component of overall bassline swing feel (Reid: 2000, 8; Richmond 1983, viii). Applying this idea to recordings, a microphone close to the bass will likely capture more of the bass’ decay envelope than is apparent to a drummer even 2 or 3 feet farther away. The drummer is behind their kit and, unlike the microphone, their ears are less ideally situated to perceive the overall acoustic signature of the bass. As a result the amount and kind of connectedness or smoothness of bassline a drummer hears (and is responding to musically) will vary from that which is captured on record.

Another issue with microphone placement concerns the tapped ride cymbal pattern, a key component of swing rhythm. Whatever microphone configuration is used to record an ensemble, it’s almost certain that the exact timbre, attack transients, and loudness of the ride cymbal (or, often, cymbals) is markedly different live than on record. The ride tap patterns will have a particular quality for the bassist depending on his or her orientation vis-à-vis the drummer. They may be hearing the top of the cymbal, the more resonant tone emanating from beneath the cymbal, or some combination. The sound of the ride-tap is one of the primary components of a drummer’s style, and it is worth noting that the cymbal sound a bassist hears will deviate, sometimes significantly, from the recorded sound.

The issue of divergence between recorded cymbal sound and a bassist’s participatory experience of that sound is compounded when the drummer switches from one ride cymbal to another. Many drummers use two cymbals (or three, or more) and move their ride pattern from one to the other to underline shifts in texture, section breaks, etc. This switching is a highly effective way to shape the overall musical texture, and the resulting change in ensemble sound can be marked. When a drummer changes where the ride pattern is being played the sound source of that pattern shifts, possibly one metre or more, and the timbre also changes. While microphones will capture the timbral shift, a bassist’s experience of that shift will vary from the recorded representation of the cymbal sound. For example, assuming a drummer’s sound is being captured by a pair of ‘overhead’ condenser microphones (a common practice), no matter which cymbal a drummer plays the microphones will be exposed to the sound emanating from the stick-head and the vibrations of the top of the cymbal. A bassist standing just to the right of a drummer will hear the top sound of the leftmost ride cymbal and a combination of the top and bottom sound of the rightmost. In addition the right cymbal will be closer to the bassist, leading to other perceptual effects relating to loudness and proximity. Such subtleties can have a major impact on participatory impulses for both bassist and drummer, and, of course, there will be no corresponding participation on the part of the microphones.

There are quite a number of additional factors at the recording-input stage that can lead to significant change in a rhythm section’s sound, some of which are outlined in Table 1. There are many other post-production techniques that will also alter the rhythm section sound, most notably equalization and limiting. All these techniques may lead to great results in a recorded product, and, in addition, will generate a rhythm section sound that is increasingly divergent from the ‘natural’ or ‘room’ sound experienced by musicians themselves. Musicians develop an idealized rhythm section sound that is determined by reference to well-known recordings, one that has an exaggerated emphasis on bass frequencies and detail resulting from close microphone placement and a blend of bass and drums that is created by the interaction of multiple microphones rather than the acoustic interaction available only during live performance. Returning to Trezise’s work on the topic, ‘we may regard a record as forming a diegesis with and within its domestic or other environment’ (Trezise: 2009, 207).  Among the environments inhabited by jazz recordings are the practice spaces of developing musicians, the desks of music theorists and historians, and the studios of experienced musicians and engineers. Each of these environments is one of deep engagement, and the diegetic activity regarding rhythm section sound in these environments, to a remarkable degree, tends to forget that the recorded product will differ from the participatory moment of a recording’s creation.

Recording Technique Effect on Recorded Rhythm Section Sound
‘Fader riding’ by engineers changes the rhythm section dynamic
The proximity effect created by placing directional microphones close to instruments Unnatural emphasis of bass frequencies
Input signal compression or limiting Alters the perceived ADSR envelope of acoustic instruments.
Use of isolation booths during recording (increasingly common in jazz after about 1980) Completely changes how bassist and drummer hear each other during recording
Use of ‘direct’ magnetic or piezoelectric pickups either alone or together with microphones to record the bass Yields a recorded bass sound that is markedly different than the acoustic sound
Extreme stereo panning Obscures rhythm section ‘blend’

Table 1 Some common recording techniques and their effect on rhythm section sound.

Creative Entrainment with Rhythm Section Recordings Leads to Participatory Discrepancies

The phenomena associated with rhythm section recording and the reception of these recordings resonates with the concept of ‘participatory discrepancies,’ or PD, Charles Keil’s terminology for the sub- and supra- syntactical interactive aspects of music performance. Musicians and scholars of music refer to canonical jazz recordings as an immutable reference documenting playing style without acknowledging the fact that their existence as documents is sonically distinct from the circumstances of their origin. In fact the sonics of jazz recording are not true to the original performance—no recording is. Just as rhythm section players are performing microrhtyhmic PD during a recording, the microphones, engineers, and all recording arts introduce PD of their own. Keil advises that musicologists ‘Listen and look for participatory discrepancies in the process and texture of a music and in its wider contexts’ (Keil: 1987, 278, italics original). Musicians and musicologists who analyze jazz recordings are participants in the ‘wider contexts’ of jazz music as they work to understand, describe, and enact music based on their participatory engagement with jazz discourse and performance. Their studies of jazz are often based on a recording of a performance. That recording is discrepant from the original sounding, so active engagement with the recording leads to PD between the experience of people present at the recorded event and present day interpreters of a recording. Similar PD can occur between people listening to a recording on different sound systems, in different rooms, under different social conditions, and so on.

While this level of analytical and didactic participation goes beyond the scope of most discussions of PD in jazz music, there are important considerations presented by the current line of inquiry that inform both the dialectics of the jazz rhythm section and the study of record production. The excellent recorded documents of jazz are a wonderful reference but they misrepresent the sound of the rhythm section and lack data key for understanding the performance practice of jazz. The act of listening and transcription, central to jazz analysis and pedagogical practice, is in part a mapping of a musician’s gesture from recorded media onto the body of the transcribing musician, a mimetic process wherein a transcriber attempts to embody the transcribed. This kind of ‘musical perception involves an understanding of bodily motion–that is, a kind of empathetic embodied cognition’ (Iyer: 2002, 394). ‘Empathic embodied cognition’ is problematized in this case by the fact that the recording process has distorted the subject being embodied by a musician or empathically analyzed by a theorist.

For one example, a bassist transcribing Scott LaFaro playing with Bill Evans and Paul Motian on Sunday at the Village Vanguard is hearing, over and over again, the sound of a bass recorded at close proximity (Goldsby: 2002, 19). This sound, possibly accentuated by use of a directional microphone, is much different than what Evans, Motian, the audience, or even LaFaro himself heard. A theorist analyzing the microtiming of Motian and LaFaro’s interaction is using the microphone’s situated interpretation and the mastering engineer’s normalization, not the rhythm section’s perception of time or dynamics. Such examples can be found frequently in the reception of jazz recordings; located microphones and engineering processes yield a sonic palette that is engaged with as real, but is in fact unreal or hyperreal.

PD occur during the intense process of listening and transcription, when developing and mature musicians alike work to completely engage with recordings. This process may be the type of enveloping participation Keil waxes poetic about in ‘Participatory Discrepancies and the Power of Music.’ The discrepancies are found in the engaged listener’s creative entrainment with recorded materials. People listen into recordings and allow themselves to be subsumed by the sonic experience. When paired with a focused desire to learn about a recording’s content these listeners become entrained in the experience. To transcribe and fully learn or analyze a recorded jazz performance is an act of embodiment that requires imagination as well as technical listening. Transcription is, in no small part, a creative process designed to, borrowing the words of educational philosopher Maxine Greene, ‘…release imagination to open new perspectives, to identify alternatives. The vistas that might open, the connections that might be made, are experiential phenomena; our encounters with the world become newly informed’ (Greene: 1995, 18). Canonical jazz recordings hold an oracular position within the culture of jazz performance and analysis; musicians engage with them in a participatory fashion, using them as a model upon which to develop an individual sound, forgetting that these recordings are simulacra.

The PD of creative entrainment with recordings are more intense for the rhythm section than for other jazz instruments. Rhythm section players and scholars study bass and drums as a dyad. Other instruments inhabit more individualized or soloistic sonic spaces on the surface of recorded jazz, their phrasing is often slightly delayed partly in order to stand out from the rhythm section, and their role is not as relational (Friberg and Sundström: 2002, 345-346). In many ways the hyperreal and distorted sound of recorded jazz rhythm sections is a positive creative development, an ideal that musicians and engineers can work toward creating in live situations. It’s also likely that this ideal leads to PD among performing musicians that can be difficult to negotiate. For musicians in the jazz tradition it’s possible that the sound of iconic jazz recording often surpasses the sound of actual live jazz. As a result, the real thing—live and unplugged—can sound thin, muddy, and disappointing.

Creative entrainment PD are also a consideration when interpreting work by theorists of rhythm. As mentioned, scholarship on the topic of swing rhythm in the past ten years has become very fine-grained, including analysis of waveforms to determine inter-onset-intervals (IOI) and statistical evaluation of the microtiming of swing as recorded musicians express it. Studies by Matthew W. Butterfield (2010, 2011), J. A. Prolger (1995), and Anders Friberg and Andreas Sundström (2002) have provided insight into the perception, execution, and purpose of jazz phrasing at the level of anacrusis, tactus, and meter. With the exception of J. A. Prolger’s 1995 paper (which used musicians recorded in isolation specifically for research purposes) these scholars have based their study on recorded jazz performances for which germane information such as the position of microphones in the room, location of musicians in relation to each other, and the various production and post production techniques used is unavailable. This scholarly work does contribute to our understanding of improvised rhythm and swing and presents well-reasoned arguments. It does not account for the fact that the microtiming and amplitude of events as microphones capture them may be different from what the participating musicians perceived.

Given that the subject of this research is microtiming it is worth looking at some numbers. Sound travels about 34cm per millisecond (MS). At this rate rhythm section musicians offset 3-4 metres from a sound source will experience a 9-12 MS delay between instrumental onset and aural perception. While we can’t be sure how far apart the members of the rhythm sections were in the recordings analyzed by Butterfield or Friberg and Sundström it is likely that a correction to their bass and drum IOI calculations based on inter-musician distance could be statistically significant. The same is true of amplitude; it is not safe to assume that, because a microphone captures instrumental sound and that sound is audible on a recording at a certain amplitude, musicians present could hear it (or were responding to it) with that same amplitude. The same issues are at play with the use of waveforms to calculate IOI. While I have no doubt that all the scholars mentioned used the utmost care in locating the pertinent attack points it is possible that equalization, limiting, or normalization applied by engineers have affected the data being analyzed.

Conclusion

Keil writes ‘I suspect that every culture has its own blinders that protect participatory discrepancies and keep them as fully mysterious and as fully participatory as possible’ (Keil: 1987, 279). Among rhythm section players and scholars of jazz there is a high level of creative engagement with recorded music. This engagement is participatory and mysterious, often done alone with headphones, and invites various discrepant points of view about what is happening in a recording. The PD associated with the jazz rhythm section on record are still largely unexplored experiences shared by many musicians but rarely engaged with as PD. Rather, the creative entrainment that serious listening and transcription engenders can result in misapprehensions (discrepancies) about what the sonics of a live jazz rhythm section are. Such misapprehensions can motivate players to find an idealized sound, can influence scholarly work, and contribute to the language used among jazz musicians, theorists, and connoisseurs, but have yet to be fully understood or ethnographically studied.

The highly developed improvisational lexicon of jazz is problematized through engagement with recording. The “liveness” intrinsic to improvisation, interaction, and timekeeping are transformed through recording, rendering fluid momentary experiences into fixed documents. As improvised music has developed since the inception of recording these documents of improvisation have become central to the scholarship and evolution of jazz music. This paper has made an initial investigation into one small and possibly inconsequential aspect of the transformation of live music into recorded document—the effect of recording on the perception of microtiming—further and interdisciplinary research will reveal many valuable avenues for exploration.

Thomas Porcello’s paper on the language of novice and experienced recordists (Porcello 2004) and Ingrid Monson’s ethnographic investigations of jazz performers (Monson 1996) present templates for investigation into the effect of recording on jazz musicians’ approach to music making. The “shop talk” of jazz musicians is peppered with references to recordings, recording techniques, and the recorded sound of musicians. An analysis of this emic language would likely reveal surprisingly intertwined ideas of “liveness” and artifice. Further, such research would begin to unravel how language codes power and register within the world of jazz improvisation. In addition study of recorded improvisation generally opens a rich conversation about the meaning of improvised musical gestures that are created in the moment but, through recording, are captured and coded permanently. Live improvisation is embodied, situated, and fleeting. Recorded improvisations present a kind of graph of live improvisation and are disembodied, situational, and fixed. They can be returned to or remixed any number of times, and stand at a distance from their live progenitors. While staff-based notations of jazz improvisations are generally understood to be incomplete, lacking truly accurate rhythmic, pitch, and timbral information, this understanding does not extend to recordings of improvisation. Further research could begin to unpack the signs and significance of improvisation-on-record and illuminate the development of improvisation in the 20th C.

Next steps in this research can also include analysis of how consumer audio systems (especially headphone listening) affect the sound of jazz recordings.  Such systems tend to “sweeten” or otherwise distort the audio spectrum. Low-end enhancement (or cutoff), midrange boost, and “noise reduction” filters burnish the sound of recordings, and may influence the expectation of musicians performing acoustically. The values of audio system designers have a role in the development of jazz inasmuch as musicians use their products to understand and decode recorded documents.

These kinds of investigations will benefit from the combined use of tools from linguistics, anthropology, political science, and musicology. One tool could be frame analysis, first theorized by Erving Goffman (Goffman 1974) and introduced into the academic study of music by Steven Feld (Feld 1984). Applied to the language and mores expressed by improvising musicians, sound engineers, and scholars of music, frame analysis could enhance research into the function and role of recorded improvisation. Jazz as a style forms a specialty within musicological research but the role of improvised gesture and interaction in the history of jazz music, and its integral role in the development of recording technology, can inform broader studies of the recording studio as location.

About The Author

John Crooks, University of California, Irvine

Bibliography

Baker, David (1979) Jazz Pedagogy. Van Nuys, California: Alfred.

Brown, Anthony 1990 ‘Modern Jazz Drumset Artistry’. In: The Black Perspective in Music. 18, 1-2, pp 39-58.

Butterfield, Matthew W. (2010) ‘Participatory Discrepancies and the Perception of Beats in Jazz’. In: Music Perception: An Interdisciplinary Journal. 7, 3, pp. 157-176.

Butterfield, Matthew W.  (2011) ‘Why Do Jazz Musicians Swing Their Eighth Notes?’. In: Music Theory Spectrum, 33, 1, pp. 3-26.

Feld, Steven. (1984) ‘Communication, Music, and Speech About Music’. In: Yearbook for Traditional Music, 16, pp. 1-18.

Fribers, Anders, and Andreas Sundström (2002) ‘Swing Ratios and Ensemble Timing in Jazz Performance: Evidence for a Common Rhythmic Pattern’. In: Music Perception: An Interdisciplinary Journal, 19, 3, pp. 333-349.

Iyer, Vijay (2002) ‘Embodied Mind, Situated Cognition, and Expressive Microtiming in African-American Music’. In: Music Perception, 19, 3, pp. 387-414.

Keil, Charles (1987) ‘Participatory Discrepancies and the Power of Music’. In: Cultural Anthropology, 2, 3, pp. 275-283.

Goffman, Erving (1974) Frame analysis: an essay on the organization of experience. New York: Harper & Row.

Goldsby, John (2002) The Jazz Bass Book: Technique and Tradition. San Francisco: Backbeat.

Greene, Maxine (1995) Releasing the Imagination: Essays on Education, the Arts, and Social Change. San Francisco: Jossey-Bass Publishers.

Monson, Ingrid (1996) Saying Something: Jazz Improvisation and Interaction. Chicago: University of Chicago Press.

Porcello, Thomas (2004) ‘Speaking of Sound: Language and the Professionalization of Sound-Recording Engineers’. In: Social Studies of Science, 34, 5, (Special Issue on Sound Studies: New Technologies and Music), pp. 733-758.

Potter, Gary (1990) ‘Analyzing Improvised Jazz’. In: College Music Symposium, 30, 1, pp. 64-74.

Prolger, J. A. (1995) ‘Searching for Swing: Participatory Discrepancies in the Jazz Rhythm Section’. In: Ethnomusicology, 39, 1, pp. 21-54.

Reid, Rufus (2000) The Evolving Bassist. Teaneck, New Jersey: Myriad Limited.

Richmond, Mike (1983) Modern Walking Bass Technique. Englewood, New Jersey: Ped Xing Music.

Skea, Dan (2001) ‘Rudy Van Gelder in Hackensack: Defining the Jazz Sound In The 1950s’. In: Current Musicology, 71-73, pp. 54-76.

Stern, Chip (2000) ‘Shop Talk: Billy Higgins’. In: Jazz Times. [Online] November. Available at: http://jazztimes.com/articles/20257-shop-talk-billy-higgins (accessed November 2011).

Schuller, Gunther (1989) The Swing Era: The Development of Jazz 1930-1945. New York: Oxford University Press.

Trezise, Simon (2009) ‘The Recorded Document: Interpretation and Discography’. In: Nicolas Cook, Eric Clarke, Daniel Leech-Wilkinson, and John Rink (editors) The Cambridge Companion to Recorded Music. Cambridge (UK): University Press, Cambridge, pp. 186-208.

Wilson, Olly (1992) ‘The Heterogeneous Sound Ideal in African-American Music’. In: Wright, Josephine (ed.) New Perspectives on Music: Essays in Honor of Eileen Southern. Michigan: Harmonie Park, pp. 327-340.

Discography

Evans, Bill, Sunday at the Village Vanguard. [Compact Disc] Riverside/OJC 1990 (original release 1961).