We Are Sitting In (Another) Room: Improv with Architecture

Today marks the 40th anniversary of Nicolas Collins’s Pea Soup, a piece that uses electronics to “play” the signature acoustics of a space. In honor of that milestone, Collins today unveils Pea Soup To Go, a free virtual jukebox programed with recordings of 70 different versions of the work, iterations which span decades and continents.

Since the composition relies on linked microphones and loudspeakers in a “self-stabilizing feedback network” to map and respond to changes in the room and produce the sonic content featured in the piece, it might just be one of the purest forms of ambient music available. The jukebox shuffles the various collected recordings, masking transitions between each with long crossfades, allowing listeners to dip into this historic stock pot and feast until they are full.

Nicolas Collins: Technically it’s pretty simple. Everybody seems to have heard the squeal of feedback at some point, and most are familiar with the fact that moving the microphone (or electric guitar) usually changes the pitch of the feedback. I explain that the phase shifter (the electronic gizmo at the heart of the piece) emulates a hand moving the mike every time the feedback starts to swell. The piece has a sufficiently dreamy, non-threatening quality that most people don’t worry too much about the how and why.

MS: And that idea led you to the title Pea Soup?

NC: The immersive quality of the sound field brought to mind the cliché of a fog “as thick as pea soup.” Rather silly, in retrospect, but I was pretty young and now I’m stuck with it.

MS: While reading up on the history of Pea Soup, I was surprised to discover that the work can involve (or always does?) live musicians. This was something I didn’t quite pick out in the first few iterations of the piece I heard via the jukebox. They are charged with interacting with the electronics (or later the software) in some specific ways. Can you explain why you prescribe their actions in the way that you do? And then this of course made me curious about the impact of the audience in the space and therefore on thework itself.

NC: Left to its own devices the Pea Soup feedback network creates simple, languid melodies whose pitches are derived from the resonant frequencies of the room (and the tempo reflects the reverberation time–larger rooms play slower tunes.) A small change in the room acoustics can cause a pitch to be added to or dropped from the melody, like some slow hocket music. I ask performers to “play” the acoustics by walking around the room, since interfering with the reflecting paths of the feedback often causes a change in the patterns. They play notes as well: playing a unison with a feedback pitch, then bending slightly out of tune, can stop the feedback; playing an octave or fifth above a feedback pitch can cause the feedback to break to the upper interval; and introducing a pitch that hasn’t been heard in the feedback from several minutes often brings it back into the melodic pattern.
Audience sounds and movement obviously influence the patterns as well–a performance in a noisy bar unfolds very differently than in a quiet, formal concert hall. I’ve also installed the work in gallery settings, where interaction with the audience becomes central.

In performance I usually let the feedback system stabilize for a few minutes, as a sort of alap introducing the scale of the room, before the players start. The web app (Pea Soup To Go) shuffles a library of around 70 performance recordings, with long fade-ins and fade-outs. The sequence is random (or as close as I can get), as is the selection of in- and out-points for each file, so the recordings always start at different times–sometimes one drops right in on a musician’s sounds, but sometimes you have to wait a few minutes to hear a player. Plus the players are instructed to play “inside” the feedback texture, rather than soloing on top, so it’s not always easy to distinguish the instrumental voices.

MS: Okay, now for the gear snobs in the crowd, this piece offers some interesting insights into the punishment time can dish out on work that involves specific electronic components that can break down and become obsolete. This led you to some particular extremes—I especially loved the correspondence you exchanged with Carl Countryman, the maker of the phase shifter you originally employed in the piece. Can you tell us a bit about that evolution and how it affected the work?

NC: This will make me sound even older than I am, but back in 1974 there were no digital delays (or at least no affordable ones). The studio at Wesleyan had three Countryman Phase Shifters that Alvin Lucier had bought to do what’s called “Haas-effect Panning,” which is a way to pan sounds quite realistically using very short time delays. I had been working a lot with feedback, and discovered that changing the phase shifter’s delay setting could emulate moving a mike, opening up a whole new vista of quasi-automated feedback manipulation. Pea Soup emerged as one of the major products of my undergraduate education.
After college I moved on to other materials and technologies (early microcomputer music, live sampling and signal processing, collaboration with improvisers.) But I’d return to feedback from time to time, and when, through my day job in New York, I ran into Carl Countryman at trade shows I’d always ask if he had any of the Phase Shifters back at his warehouse. By the 1980s he was making very popular high-quality Direct Boxes and lavaliere microphones, and the phase shifters were long gone and, it seems, not missed–his answer was always “no.”

Then in the late 1990s I was in Berlin with a DAAD fellowship, and an ensemble with which I was working (Kammernesemble Neue Musik Berlin) asked if they could revive Pea Soup. At first I tried to reconstruct the original analog circuit. I emailed Mr. Countryman, who obviously still remembered my unwanted nagging, and he sent me the schematic with the explicit understanding that I was never to bother him about this device again. The circuit is not complicated, but it has one odd custom-made part that was difficult to duplicate. I did a few performances with my best attempt in the analog domain, but after a few years I wrote a software emulation of the original analog boxes that, with enough code tweaking, evolved into a pretty convincing substitute.

Software has allowed me to add a few features that would have been great to have back in 1974 but were out of reach then (such as a filter that automatically nulls pitches that would otherwise dominant in the texture.) Programs are not as cute as little metal boxes, but they’re lighter and can be distributed more freely, like old-fashioned paper scores: I’ve posted the program on my web site, where anyone who’s interested can download it and perform Pea Soup without the need fly in Nic and his gear.

MS: How does the experience of Pea Soup via this clever website relate the performance experience of hearing it live for you?

NC: In a big space with big speakers Pea Soup can be a very immersive and interactive experience—“church of sound,” as one friend once called it. The web app (Pea Soup To Go) is obviously more like listening to a recording of a concert than experiencing a live event, but this is a record that never ends, never repeats—a multi-disk CD changer in “shuffle mode” with a twist: the long crossfades knit the 70 files into one continuous performance. Since every room is in a different, architecturally determined “key,” you end up hearing a series of odd, vaguely modal chord changes that stretch out over an almost glacial time scale.

MS: Even before I started reading the background on Pea Soup, I kept thinking of Cage and Lucier associations related to “hearing” a space–using a space and its contents as so essential to the end sonic result. Do you hear this piece as in that evolutionary line? In what ways does it intersect and/or diverge?

NC: Yes, it certainly is in that line. I was a young, impressionable student of Lucier’s at the time I made Pea Soup. I was drawn to feedback under the twin influences of Lucier and Cage. I loved Lucier’s extraction of musical material from fundamental acoustical phenomena (think of Vespers and I Am Sitting In A Room). My parents were both architectural historians, and the link between music and architecture was critical to my finding a comfortable place to work. And feedback became the solution to my Cage-induced ambivalence about making personal musical decisions in a world where all sounds could be “musical sounds”: turn up the volume and let nature/god/architects do the rest—a sort of acoustical I Ching.

Divergence? I think my generation of musicians and composers is (and always was) much more comfortable with the idea of improvisation than our teachers were: Cage hated it; Lucier kept trying to come up with other words to describe it. In Pea Soup and most of my other work I embrace improvisation, I hand a lot of responsibility off to my players, and live with the consequences.

I also see each musical generation incorporating a new generation of technology. My peers and I embraced synthesizers, effect boxes, homemade circuitry, computers. And technological shifts often beget stylistic changes – some modest, some significant. There’s a certain kind of technological interactivity that I believe is, for better or for worse, the gift of my generation of experimental music composers.

MS: Even though this was originally a student piece, you note that the lessons of architectural acoustics have continued to engage you, making this piece of ongoing interest even 40 years later. What have some of those lessons been?

NC: I still have difficulty making certain musical decisions, and I often return to acoustics to clarify the edges or underpinnings of a piece. In the end no sound gets to the ear without engaging with acoustics, and the physical reality of sound keeps me grounded. There’s a certain primordial consonance or orderliness or reassuring “rightness” in it, that I find helpful when I’m feeling lost.

Recently, while tweaking the software for Pea Soup, I discovered a simple way of mapping the resonant frequencies of a room to conventional music notation. I’ve written a piece (Roomtone Variations) that uses this technique to create a site-specific score for any concert space, in real time, in the presence of the audience. The score is projected on a screen for all to see as it unfolds, and after the analytical intro (which takes about two minutes) an ensemble performs purely acoustic variations on this “architectural tone row” – a kind of “Pea Soup Unplugged.”

Another new piece, Speak, Memory, uses room reverberation as short-term memory for image files and sound bites. In the course of the performance I display the transformation of the original pictures and sounds as they are “forgotten” by the room. (I hope to include both these pieces on my first concert in New York in many years, at Roulette on March 9.)

You could look at this obsession in one of two ways, I suppose: either I am somewhat pathetic for, at the age of 60, still being hung up on my first true love from age 20; or it’s a sign of deep commitment to one’s fundamental beliefs. Take your choice.