Sampling composers

Morey, J., & McIntyre, P. (2014). The Creative Studio Practice of Contemporary Dance Music Sampling Composers. Dancecult, 6(1), 41–60.

There is so much to love about this paper, starting with the title. You can read it the way it was intended, that dance music producers are composers. Or you can creatively misread it to mean that the dance producers are using samples of other composers. It works equally well either way.

Sampling consists of acts of listening, selecting and editing

In the age of the internet, effectively any sound that has ever been recorded becomes available raw material for new music. The challenge with sampling isn’t so much identifying possible sample sources as it is managing the vast universe of possibilities. The listening and selecting steps in the sampling process are really the hard parts. The editing and looping are comparatively easy.

Morey and McIntyre identify three main techniques used by sampling artists:

  1. Traditional sample manipulation: looping, flipping, scratching.
  2. Use of a sample as an initial building block for a composition from which the sample is then removed. This is a widely used technique for avoiding sample clearance fees.
  3. Live performance in the studio which is subsequently cut up and treated as a sample library. This is a technique that I like to use.

Where is the creative content of the music?

If you mostly engage with music via scores, the most salient parts of music will seem to you to be the ones that can be notated: melodies, chords, and lyrics. If you mostly engage via recordings, then the most salient parts of the music will be the sounds themselves: the timbre and space, the performance nuances, the arrangement and production. Melody and harmony are part of the experience as well, but it takes some work to pull them out of the sound mass. (Also, plenty of recorded music of the past few decades doesn’t use melody or harmony at all.) Lyrics can be perfectly unintelligible without diminishing the pleasure of hearing them. D’Angelo is a case in point.

In the notation paradigm, the work is complete when it’s written down. In the recorded paradigm, a score is neither necessary nor sufficient to realize a work. The work is the mixed and mastered recording. This idea is commonsensical in our culture everywhere except in the music academy. In a grad school composition seminar, I was required to submit a score for my final project. It was a funny request, since the project was a sample collage I made in Ableton Live, and was never notated in any form. I ended up just submitting screencaps of my session. If someone else wanted to perform or adapt the piece, I’d just give them the session file. Written instructions would be superfluous.

Roland Barthes and Julia Kristeva made a distinction between the geno-song and the pheno-song, analogous to the genotype and phenotype in biology. The geno-song consists of the musical structure and stylistic conventions–the song and arrangement. The pheno-song is concerned with “the voluptuousness of [the song’s] sound signifiers” (Barthes 1990, 295). Barthes calls the pheno-song the “grain” of the music. Albin Zak would refer to it as the “track.”

In Barthes’ framework, we react to the geno-song formally and analytically, whereas we react to the pheno-song ecstatically and physically. This is a gross oversimplification, but it’s a useful one. The art of hip-hop sampling is all about the pheno-song. The rhythms, melodies, and harmonies in a sample are easily altered. A sample’s true and unique value is its timbre. The same drum pattern will sound quite different depending on the specific kicks, snares and cymbals you use. Anyone can put together a four-on-the-floor techno pattern, but it requires a specific kick drum sound to fill the dance floor.

The case of Bob James is a perfect illustration of the way that sampling musicians prioritize sound over harmonic, melodic or structural content. Hip-hop musicians love to sample Bob James. This surprised me, because his geno-songs are lightweight jazz fusion, seemingly far distant from the tastes of hip-hop listeners. But James’ pheno-songs have timbral qualities that suit hip-hop perfectly.

The recordings most prized by samplers might only have five good seconds of music on them, but those five seconds are more than enough to create a great sample. Jazz fans like me usually prefer John Coltrane to Bob James, but Coltrane isn’t a terrific source for groove samples because the bass and drums on his records were often recorded poorly–Elvin Jones sounds like he’s playing a pile of cardboard boxes half the time. For a top-to-bottom listen, Coltrane is still  better, but if I’m looking for a good snare drum, I’ll go to Bob James, whose recordings are produced immaculately.

Sample-based music represents a titanic shift in the creative process, the natural conclusion of the separation of studio recording from live music. Jeff Chang says it best:

Sample-based music uses sounds instrumentally, rather than using instruments to make sounds. In sampling, sound marks the beginning of the creative process, and is accordingly treated as raw material. Instrument-based music treats sound as an ontological object, in which sound is considered the end of the process (2009, 145).

For instrumentalists and the composers who write for them, the timbre is a given. The expressive content is in the notes. Sample-based music cares about notes, to a limited extent, but they are neither necessary nor sufficient to the finished work.

Sampling is enforced collaboration

Even after thirty-plus years of hip-hop, a lot of people continue to feel moral discomfort about sampling, especially when it happens without permission. There’s a sense in which sampling is stealing, as the samplers themselves wryly acknowledge–see “Rhymin’ and Stealin‘” or “Jackin’ for Beats.” But samplers’ intentions are usually benign. Andy Carthy puts it well:

[Y]ou’ve created this little band and you’re trying to get them to talk to each other and have fun playing together. And once they’re all happy, that’s when you know that the track’s working. It’s almost personalising something that’s quite mechanical. Every element you pick is there for a reason and has its own little personality anyway, so it’s getting all these different musicians who’ve never met each other to talk to each other and have a good time, probably without anyone ever realising it.

There’s a sense in which samplers are bending the sampled musicians to to their will. But traditional composers do that too. George Grella goes so far as to compare the great classical composers to psychopaths in their desire to control the bodies of other humans. Seen this way, samplers’ coercion of unwitting participants looks more benign. At least the musicians who we’re sampling don’t have to be in the room when we’re exerting our uncaring will upon them.

Simon Reynolds likens an audio recording and a sample taken from it to a “residual ghost”, that is “unaware of change in its surroundings and continues to play the same scene repeatedly” (Reynolds 2012, 313). Reynolds also points out that most of the discourse around sampling is arguing in its defense. Why does sampling need so much defending, when everyone long ago made peace with collage in other media? Maybe because sampling amplifies the unreal qualities that all recorded music shares.

Recording is pretty freaky, then, if you think about it. But sampling doubles its inherent supernaturalism. Woven out of looped moments that are like portals to far-flung times and places, the sample collage creates a musical event that never happened; a mixture or time-travel and séance.

This phrase, “a musical event that never happened,” is a good one. It doesn’t just apply to sampling! Modern recorded music has grown increasingly distant from the sound of humans playing instruments in a room. A late-period Beatles song is just as alienated from real-time performance as a hip-hop track, though it’s easier to convince yourself of the Beatles’ physicality. Here we get to the nub of why sampling bothers people who otherwise have no interest in music copyright issues. It confronts us with the intrinsically illusory nature of studio music. We don’t like being fooled.

Though actually, some of us do like it. For fans of electronic music, the dissociative anxiety of listening to “unreal” sounds is the whole attraction. It mirrors and validates our lived experience. Nothing in America in 2015 is genuine or authentic. Better that musicians confront and engage that fact than try to wish it away. Reynolds presents sample-based composition as a simulacrum, a binary opposite to “a sequence of human actions happening in real time” (2012, 314). Most recordings are closer to the former than the latter. Simon Frith points out that that recordings become “real” performances just by existing.

I listen to records in the full knowledge that what I hear is something that never existed, that never could exist, as a ‘performance’, something happening in a single time and space; nevertheless, it is now happening in a single time and space: it is thus a performance and I hear it as one (Frith 1996, 211).

I have no desire to split hairs over the definition of a “performance,” but I’m perfectly happy to regard the playback of a recording as being just as “real” as humans playing instruments.

Critical listening is the fundamental musical strength of samplers

Where is the musicianship in sampling? Even if you think the results are cool, can you really respect someone who loops chunks of recordings the same way you respect a violinist or drummer? Here we get into some interesting territory: the relationship between technical skill and musicianship, or the lack thereof. Morey and McIntyre find that sampling composers are more concerned with product than process.

While our interviewees certainly acknowledged sampling composers who undertook considerable manipulation of samples (known as “chopping” and “flipping” in hip-hop production), it was interesting that they generally placed no greater values on these technical skills than they did on the ability to identify a sample that would work within a given context. As such, the ability to listen and select was considered as compositionally significant as any production or technique-based skills.

Timothy Warner says that the sampling composer “need not be a physical virtuoso, only a virtuoso of the imagination with an expertise in the manipulation and organization of sound” (2003, 96–7). This virtuosity is the same thing required of any composer, really. But samplers develop it by listening analytically to recordings, a lot of recordings. They know how to listen deeply and really hear the music. More importantly, they know how to listen past the music and hear its latent possibilities. Morey and McIntyre point out that producers of sample-based music need to be able to listen both passively, as “consumers,” and actively, as artists. This skill needs to be learned, just like any other.

In the past, we’ve conflated musicality with technique, because instruments are hard, and music is hard, and by the time you’ve learned one, you’ve probably spent a ton of time learning the other. Music editing software is comparatively easy, but you still have to learn the music. I like the Microsoft Word analogy: any reasonably bright person can quickly learn how word processing software works, but learning how to write is another ball of wax entirely. So it is with music software. I can take any motivated student and have them chopping up samples in an hour. But are the results going to sound good? That’s where the long hard work comes in.

All digitally recorded music is sample-based

Once you make enough computer music, you start to realize that the genre of music hardly matters, you can use the same workflow on all of it. You start with a bunch of recordings, and you edit and process them until they sound good. Those recordings might have just been recorded in the same studio where you’re sitting, and maybe you hardly edit them at all; nevertheless, you’re manipulating digital samples, same as any dance music or hip-hop producer. David Byrne and Brian Eno came to this realization while working with tape at the dawn of the 1980s, when sample manipulation was quite a bit harder than it is now. Morey and McIntyre quote Byrne at length talking about the creation of Remain In Light, and I’ll pass that quote right along here:

We gave ourselves two weeks to build this instrumental scaffolding . . . the vocals would have to come later. . . . We worked rapidly. One or two people would lay down a track, usually some kind of repetitive groove that would last about four minutes, the presumed length of a song. Maybe it would be a guitar riff and a drum part, or maybe a sequenced arpeggio pattern and an intermittent guitar squeal. Others would then respond to what had been put down, adding their own repetitive parts, filling in the gaps and spaces, for the whole length of the ‘song’. As we’d listen to one part being recorded, we’d all be scheming about what we could add—it was a kind of game. This manner of recording had the added advantage that we weren’t trying to replicate the sound of the live band. We hadn’t gotten attached to the way these songs and their instruments and arrangements sounded in performance, so in some way the conflicts we had confronted when we’d first entered a recording studio years previously were bypassed . . . . After the tracks began to fill up, or when the sound of them playing simultaneously was sufficiently dense, it was time to make sections. While the groove usually remained constant, different combinations of instruments would be switched on and off simultaneously at different given times. One group of instruments that produced a certain texture and groove might eventually be nominated as a ‘verse’ section, and another group—often larger sounding—would be nominated as the ‘chorus’. Often in these songs there was no real key change. The bass line tended to remain constant, but one could still imply key modulations, illusory chord changes, which were very useful for building excitement while maintaining the trance-like feeling of constant root notes. Up to this point, there was still no top-line melody, nothing that the singer (me) would put words to. That came later (Byrne 2012: 157–58).

Any computer with a DAW installed is a giant sampler. The only difference between “sampling composers” and everyone else is the sample sources and the editing style.

One more thing

While putting this post together, it occurred to me that my sample-based music process has informed my writing process. My preferred note-taking method is to copy and paste the source text into an editor, delete everything that’s uninteresting, paraphrase the parts that need paraphrasing, and leave the juicy quotes intact. Then I go through, responding to my textual “samples,” and voila! It feels just like building a track from loops.

2 thoughts on “Sampling composers

  1. Pingback: De la música como escritura a la música como sonido: el caso del Amen Break (y II) | La columna de aire

Comments are closed.