The Genealogy of Transfigured Wind
Working with computers, as most people have learned, can be exhilarating but also deeply frustrating. It is as easy to enjoy the rush associated with what they sometimes enable as it is difficult not to resort to (often unsavory) anthropomorphic imaginings when they fail to do as we wish.
I have set out here to create an illustrated history -- a genealogy of sorts -- of the computer components of a series of my compositions entitled Transfigured Wind. My aim has been to assemble a body of information (notes, diagrams, sound recordings, and linking commentary) paralleling what a musicologist would seek (with an exclusively paper trail) in order to illuminate the creation of a particular musical work. So far as I am aware, this has not been attempted before in an electroacoustic context.
It did not seem that this would prove an unwieldy task, but in fact it did. The underlying problem is that work of this sort -- harnessing computational power in order to organize, transform, and synthesize musical sounds -- has had only a brief history. Its processes, nomenclature, and means (hardware and software) have been in continuous flux for decades. As a result, documentation can be promisingly specific while remaining irritatingly uninformative.
I have opened a metaphoric window on a process I and my collaborators undertook in the early 80s. What follows is by no means a comprehensive record. Neither are the materials presented fully explained, nor could they be. I have selected, from countless pages, a collection of documents that provides a set of stepping-stones along a path through the year-long process that began in the spring of 1983. It culminated the following Spring with performances of both Transfigured Wind II, the full orchestra version (at the New York Philharmonic's Horizons '84, with the American Composers Orchestra, conducted by Charles Wuorinen, Harvey Sollberger flute soloist), and Transfigured Wind III, the chamber orchestra version (at the 1984 Los Angeles Cultural Olympics, again with Sollberger and this time a California Institute of The Arts Ensemble conducted by Jean-Charles François).
Without venturing an elaborate history, here are a few relevant details:
- During 1981-83, in a series of residencies at Pierre Boulez' Ircam facility in Paris, I completed Archipelago, a 32-minute composition for chamber orchestra and 8-channel computer-processed sound. It was premiered by the Ensemble InterContemporain, under conductor Peter Eötvös, at Ircam's Festival, "The Concept of Musical Research."
- Particularly helpful in the realization of Archipelago were my Musical Assistant at Ircam, Thierry Lancino, and David Wessel, a senior staff member there.
- During the following year, I set out to replicate the scale of the Ircam work in San Diego, but to simplify the formal design and to consolidate the associated process (e.g., my editorial algorithms, SPLITZ and SPIRLZ, phase vocoding analysis and resynthesis, and spatialization).
- At UCSD, we had recently hired F. Richard Moore, one of those involved with the earliest digital sound synthesis experiments at Bell Telephone Laboratories. This work was done by a group led by M. V. Mathews. Within the Center for Music Experiment and Related Research on the UCSD campus, Moore established CARL (the Computer Audio Research Laboratory). His chief lieutenants were programmer Gareth Loy and researcher Mark Dolson. My Musical Assistant was a UCSD doctoral student, Richard Boulanger.
- I distinctly remember the elevated level of skepticism that accompanied my assertion that I was -- somehow -- going to produce a 35-minute composition for orchestra, solo flute, and computer processed sound at our fledgling facility.
- That it was, indeed, an arduous process will be clear to anyone who ventures across the stepping-stones provided here and scans the terrain they make accessible.
The assembled evidence includes:
- Note book entries and other jottings
- Graphic designs
- Musical notation
- Mixing diagrams
- Sound recordings (from both instrumental and computer-generated sources)
- Binaurally encoded examples (allowing the simulation of quadraphonic space)
A limited commentary is also supplied, but there is much that will remain cryptic and obscure. This is not so unnatural as might at first be thought. Any complex human undertaking -- all the more an intensively collaboratory and inter-disciplinary one as this necessarily was -- involves an intricate dance with specific proposals, proscriptive limitations, conversations, intuitive leaps, and the seemingly endless trials which elicit some successes but also many, many failures.
This history is divided into four large sections:
This section deals with the initial instrumental and computer-mediated explorations that served to ignite my imagination sufficiently to keep the whole enterprise (the creation of the Transfigured Wind series) in motion for more than a year. Most of the resources were newly created, but an earlier flute solo, the 1965 Ambages, written for Karen Reynolds, was also an important compositional resource.
This section describes the proposed form of the whole composition -- where the seminal flute solos would be situated in the large design -- as well as the digital signal processing strategies that were conceive and realized. An abundance of sound examples allows a visitor to experience even now, at a remove of more than two decades, some of the sense of discovery we had then.
Although, in the early 80s, computers were capable of processing sounds that had been converted to digital form, they were still far too massive and delicate to serve musicians under actual performance conditions. Thus, the four primary "computer cues" assembled from multiple layers of multi-channel sound files, were transferred back into the analog realm for dissemination in concert using magnetic tape recorders. This section documents the planning, structure, and mixing operations necessary to realize the "tape cues." One can listen to individual sub-components, and then hear them progressively montaged.
Recordings of two of the Transfigured Wind series are provided in their entirety in this section:
Transfigured Wind III, for solo flute, chamber orchestra, and computer sound, and
Transfigured Wind IV, for solo flute and computer sound.
Aligned with these recordings at precisely the correct moments in time, are the computer cues (or what the documentation refers to as Tape cues: I, II, IIIa, IIIb, IVa, and IVb) meant to be joined with either a solo flutist or a chamber orchestra and soloist in order to evoke the composer's imagined result.
A visitor can listen to particular sections, comparing the computer sound alone with the way it functions musically in two contrasting contexts (chamber orchestra and solo flute)
Finally, there are several documents that describe the creative and administrative task undertaken then -- in medias res -- as well as the resources required. It is accurate and also revealing to note that this work was undertaken at a moment in history when the collaborative and multi-disciplinary field of computer music was in its relative infancy. Many processes (such as sound file editing) that now seem elementary and are available on almost any computer were then – in the time-sharing world of the DEC PDP 11-55 computer -- still inaccurate and clumsy.
-- Roger Reynolds
Del Mar, CA, July-September 2006