"Nature, Networks, Chamber Music" by"Mark Trayle"

Abstract: the author gives a brief history of his work with musical process, describes the convergence of scientific visualization and computer-human interface in his work, then makes prognostications about the future of interactive computer music.

"...the only worthwhile researchers: undisciplined ones. The ones who refuse to answer new questions using only pregiven tools." Jacques Attali, Noise.

I've been conducting undisciplined research with musical processes for a number of years. These processes were first constructed from commercially available analog synthesizer modules. Voltage-controlled oscillators, pseudo-random voltage generators, and other circuits were interconnected to make a self-attenuating feedback loop. A closed system, a musical 'machine' with a simple and periodic behavior.

Later, I made new processes out of digital circuitry and assembled groups of them into an idiosyncratic and highly specialized synthesizer that I dubbed Wild Kingdom(e).

These processes could pass signals back and forth and mutually influence each other. Although each of these processes performed a relatively simple electro-musical task, their behavior as a network was complex and not easy to predict from moment to moment. Another case of the whole being greater than the sum of the parts.

Intrigued by this dynamism, I wanted a means to reach in and participate in the music-making, to be another source of control (or chaos!) in the network. I opened processes up to input from the 'real world', through some simple (cursor keys on a keyboard) and not so simple (a do-it-yourself pitch follower) interfaces. These interfaces allowed performers to use hands and fingers to control processes.

Instead of being scored, the detailed characteristics of musical events arose in real-time from the interaction between network and performer. The general musical form, the "shape" of a piece, grew from the topology of the network and the nature of the processes that made up the network, it's nodes.

Musical decision-making was decentralized and spread throughout the network, the result of the delicate interdependencies between processes, this interplay within the network, and between the network and the performer, began to take on the character of a small ecosystem. Interactive, flexible, and non-hierarchical, it inspired the notion that networks can provide a rich metaphor for nature. "Ecology, in the widest sense, turns out to be the study of the interaction and survival of ideas and programs (i.e., differences, complexes of differences, etc) in circuits. (n. 1)

I moved the network from the hardware domain to the software domain to make its topology and the behavior of its nodes easier to modify and add on to. I programmed new processes that simulated natural processes and events... the growth and decline of populations, the dripping of a leaky water faucet, the harmonic motion of bouncing objects.

The real-time output of these simulations was two-dimensional, points scaled and mapped onto an x/y coordinate system. I created a simple scheme (a protocol) to pass this output to other processes, then wrote new processes and upgraded others so that they could make use of this x/y input.

At about the same time computer-human interface devices were becoming readily available; the mouse became standard equipment for new computers and pitch followers got cheaper. The x/y movements of the two-dimensional mouse fit my ad-hoc protocol and became the model for all my interfaces.

Using the mouse I could modify the variables of a simulation in real-time, and the network would react in delightful and surprising ways. Something in between a musical instrument, a computer game, and a written score, the network displayed a responsiveness and vitality missing from sequenced or taped computer music.

As the 2D mouse paradigm became prevalent in the computer-using community, it occured to me that other people could perform these pieces, foreshadowing a new, interactive, chamber music.

"Simulations, Interactions, Performance"

Our understanding of natural phenomena is enhanced through the creation and use of generalized mathematical models. When faced with very complex systems that defy generalization, researchers turn to simulation. These computer-generated models help create an "allegorical knowledge" of the processes at hand, providing imagery that might further intuition and lead to understanding (n. 2).

In some cases (fractals, e.g.) the imagery produced by scientific visualization becomes interesting in and of itself, making inroads into the art world and sneaking into popular culture. "Science itself is now the only field through which the dimension of mythology can be again revealed" (n. 3).

Simulations are used as the formal backbone for some of my recent pieces. A model of population growth and decline drives cc:BTWT, directing small and large-scale musical events. In Basic Harmony an algorithm that simulates harmonic motion in different domains acts as a musical guidance system. These pieces give voice to phenomena that are usually considered 'non-musical'. The simulations that guide them provide a formal/conceptual focus for the music and create audible analogs of natural processes, enhancing and revealing the innate musicality of natural systems.

Experience has taught us that the computer-human interfaces that 'catch on' fastest are those that are easy to learn and easy to use. The simple point and click technique of the mouse is fairly intuitive, hence it's popularity.

Generally, interfaces to computer music systems are based on existing (acoustic) instruments. Some form of instrumental technique, typically keyboard technique, is still required to compose and perform music on a computer. Of the many reasons why the keyboard became the de facto computer music interface (electronic keyboards are relatively easy to design and build, the target market for the first synthesizers was the home organ crowd, etc.), "easy to learn" is not one of them.

Other interfaces exist for performing music on a computer: the mouse, touch-sensitive pads, knobs and wheels of all sorts, data gloves, etc.. Whatever these interfaces might lack in the way of instrumental 'touch' and subtlety, software can make up for with 'intelligence', making complex responses to simple commands and/or gestures. At the low-bandwidth end of the interface spectrum a potentiometer can be made to change a single musical parameter, either exponentially or linearly.

At the other end of the spectrum, a gestural interface can change many parameters at once, and provide the physicality missing from some forms of performance interface (n. 4). And in between these two extremes lie other alternatives for physical and virtual interfaces to musicmaking machinery.

This range of interactivity suggests new musical languages (gestural or otherwise), then proposes a new kind of composition: flexible and responsive forms programmed by composers and guided by performers. The migration of technique from the performer to the interface also suggests that amateurs can more fully participate in making music with computers.

"...music - a channelizer of violence, a creator of differences, a sublimation of noise, an attribute of power - creates in festival and ritual an ordering of the noises of the world." (n. 5)

Computer technology has changed the way we make music, travel, and watch the weather. It facilitates the creation and recording of music, quickening the pace at which music is stockpiled as a commodity. The rapid proliferation of the MIDI-based "home studio" has helped democratize the process of making music, but has had the side-effect of removing composition and performance from a social context, from it's ritual roots. This "economy of repetition" (n. 6) reinforces the artistic status quo and further mediates our experiences of festival and nature. Performance, as opposed to making and consuming recordings, provides a direct connection to ritual. So, rather than using computer music technology to further refine archaic notions about composing, performing, and listening to music, perhaps it's better to approach it as an entirely new instrument. An instrument that can reanimate the electronic media by blurring the distinction between producer and consumer, encouraging the consumer to be the composer and performer as well as the audience.

"Processes, The HUB, Software Instruments"

The kind of music I do has its roots in the 'process music' of the 1960's, stemming from the idea of creating a closed system and listening to its behavior (n. 7). When the system is opened up to input from performers and other processes, a more free-wheeling kind of behavior can be heard.

The rigid temporal and harmonic conventions of 'minimalism' are bypassed, while maintaining the focus of attention (and the locus of musical form) on the system's behavior. The map of a particular piece may remain the same, but the scenery and the path taken will change from performance to performance... a constant source of surprise.

Says John Cage, 'A structure is like a piece of furniture, whereas a process is like the weather' (n. 8).

Many of the pieces composed and performed by the HUB are in the 'process music' tradition. The HUB is a computer music band, its members compose and perform music with a network designed specifically for passing musical messages between the players (n. 9).

The HUB is a direct descendant of the League of Automatic Music Composers (n. 10). John Bischoff and Tim Perkis played in The League for several years. They looked for a way to simplify the ad-hoc, error-prone technology of The League and came up with the network server, the 'mailbox', that is the center of The HUB.

The members of The HUB are all composers as well as designers of their own hardware and/or software instruments. They adapt existing compositions or design new ones to be played in the network context.

I use the word 'design' intentionally. Most HUB compositions are algorithmic by nature; dynamic systems running on one composer's computer affect the behavior of another's, and in turn make the network into a larger dynamic system with a very complex behavior.

A HUB composer can specify the musical application of messages passed through the mailbox, but given the lack of a socio-musical hierarchy in the group, it's difficult (and not particularly desirable) to provide the level of detail and control that Western ears expect from a composition. By distributing musical intelligence throughout the group, the HUB avoids the 'solipsistic quest' for the perfectly crafted composition, musicmaking remains the 'imperfect and social process' that it is (n. 11).

In 1987 I started designing a software instrument to make network music... sort of a one-man HUB. Running in the Amiga's multi-tasking environment, this instrument would allow autonomous processes to interconnect, pass messages, and make music. Some processes had real-world interfaces, others had fixed functions, generating or transforming messages and passing them along to MIDI boxes and the Amiga's sampler.

I made a prototype from the design, and made a few pieces using it (Breathing, cc:BTWT, White Science)... but I wasn't crazy about it. The process of changing the network configuration from piece to piece was tedious and error-prone, not good during a performance. Coercing the instrument to fit the needs of a new piece was difficult. So I took the instrument apart and made it into a set of software tools.

What started off as an integrated package ended up as a library, a set of functions written in C . I ditched the notion of a generic compositional tool in favor of better performances and a clearer compositional outlook.

"Basic Harmony"

Basic Harmony was constructed with these tools. This piece is a set of three studies of simple harmonic motion through different domains (time, amplitude, pitch). Each of these studies links a musical process and the performer with a different interface.

In the first, Cant, I use an electric guitar to feed pitch and amplitude information to the process. In Fulcrum, events are triggered by tapping a pencil on a piece of wood that's been wired for sound. The third and last, Lachrima, uses the sound of running water and a pitch-follower for navigation.

These are the interfaces that I've chosen to use when performing these pieces, but they can be performed with any MIDI-stream source or just a mouse. No special "instrumental" technique is required, so anyone with the interest and access to an Amiga computer can play them.

Formally, study by study, here's what happens. Cant has two sections. The computer starts off the action by generating tables of harmonics based on scaled random numbers, using (insert equation here) to generate the harmonic series for each table.

The random numbers are derived from the date and time as reported by my computer's real-time clock, so each performance should be different (I'm skeptical of random numbers generated by a machine).

In the first part of Cant these tables control the rate of ascent or descent of square-wave glissandi. The initial pitch of the square-waves is determined by the incoming MIDI-stream (see Figure 1). In the second part the harmonic tables determine the timing of (slightly-tuned) percussive events, while the incoming data stream determines pitch and timbre.

Fulcrum uses the same type of tables as Cant, but here they're the determinant of amplitude. The harmonic motion of a bouncing pencil triggers ever more florid (and diatonic) melodic streams. The 'bounce' of the pencil is mic'ed, then processed to give it pitch.

Lachrima takes the timbres of Cant, then uses them to articulate the diatonic world hinted at by Fulcrum. The tables of harmonics now drive the processing of samples of running water, as diatonic chords accompany, first arpeggiated then blocked. As the piece progresses the chords start drifting out of tune, suggesting a return (a nasty compositional habit?) to the glissandi of the first part of Cant.

"Chamber Music"

The quality, variety and availability of computer-human interface grows while it's price slowly decreases. The time it takes the interface to get from the lab to the toy store shrinks. Creators of computer graphics increasingly use simulations of natural processes to generate images instead of drawing them out by hand. New microprocessors ease the task of performing these simulations in real-time.

While not everyone can afford a virtual reality system for the home, the number of people familiar with the 'look and feel' of this technology will continue to grow. So will corporate interest in the possibilities of 'high-end' digital technology (VR, hypermedia, etc.). It seems inevitable (global economic collapse notwithstanding) that interactive digital media will be a primary channel of distribution for art and entertainment.

Composers can use this new means of distribution to bring their work into the home. "The consumer, completing the mutation that began with the tape recorder and photography, will thus become a producer and will derive at least as much of his satisfaction from the manufacturing process itself as from the object he produces". (n. 12) By moving the experience of music from the passive stockpiling of recordings to the active creation of original performances, composers can re-introduce ritual to music. Straddling the border between the amateur and the music elite, interactive chamber music will question the notion of the masterpiece and the perfect performance. Traversing the distance between the computer and the world of natural phenomena, interactivity and simulation can reanimate the electronic media and guide and inform our experience of nature.

"Notes and References"

1. Gregory Bateson, Steps to an Ecology of Mind, New York, Ballantine Books, (1975), p. 483.

2. Richard Wright, "Computer Graphics as Allegorical Knowledge: Electronic Imagery in the Sciences", Leonardo Supplemental Issue Digital Image Digital Cinema (1990), p.70.

3. Joseph Campbell, The Masks of God: Primitive Mythology, New York, Viking Press (1959) ,p. 468.

4. Kurtenbach and Hulteen, Gestures in Human-Computer Communication in The Art of Human-Computer Interface Design edited by Brenda Laurel, Addison-Wesley (1990), p.312.

5. Jacques Attali, Noise, University of Minnesota Press (1985),p.23.

6. see Attali [3], p.144.

7. Steve Reich's Four Speakers is a good example of this genre. Four loudspeakers were placed in a circle on the floor. Four microphones were hung from the ceiling so that they rested a few feet above each speaker. The performers pushed the microphones so that they swung back and forth over the speakers like pendulums, causing audio feedback every time they were directly over the speaker. Each microphone swung back and forth at a different rate and amplitude, creating the phasing effect that was a hallmark of the process music "style".

8. John Cage, Empty Words, Wesleyan University Press (1979), p.178.

9. The HUB is John Bischoff, Chris Brown, Scot Gresham-Lancaster, Tim Perkis, Phil Stone, and Mark Trayle.

10. for more information about The League of Automatic Music Composers, see the article Music for an Interactive Network of Microcomputers in Foundations of Computer Music, MIT Press (1985).

11. Tim Perkis, "The Future of Music", compiled by Larry Polansky. Leonardo Vol. 20, No. 4, p.365.

12. see Attali [3], p.144.

"Discography"

Simple Degradation: Imaginary Landscapes, Elektra/Nonesuch CD 79235-2 (AAD).

Simple Degradation: The HUB, Computer Network Music, Artifact CD Art 1002 (ADD).

Mark Trayle, composer, performer, programmer. PO Box 192014, San Francisco, CA 94119 USA e-mail: met@well.sf.ca.us