Lyric Variations for Violin and Computer, part 2 of 2, J.K. Randall, 1968
Paul Zukofsky, violin
\"Lyric Variations for violin and computer was commissioned by Paul Zukofsky in the summer of 1965; and after three years of extensive collaboration among composer, violinist, and recording engineer, was completed in the summer of 1968. Since the violin frequently plays several parts simultaneously -- and anyway since the piece was conceived as sound emanating solely from two widely separated loudspeakers -- there is no distinction between a \"live performer\" and a \"recording\" of this piece: in concert, a tape-recording is played through speakers.
\"Variations 1 - 5 are for violin alone. Variations 6 - 10 are for computer alone; and are structurally analogous to variations 1 - 5 in reverse order, so that the last variation for computer alone (var 10) -- the 2-minute \"jungle\" of variable rates of change (which takes 9 hours to compute) corresponds to the opening 2-minute violin melody (var. 1). Variations 11 - 20, for violin and computer together, are a transformation of variations 1 - 10; with variation 20 presenting in a quite direct way the total pitch-time configuration which the variations vary.
\"Throughout the computer part, I have tried to impose upon conventionally peripheral aspects of sound (vibrato, tremolo, reverberation, waveform transformation, etc.) the same degree of elaborate structuring that I impose (and that any composer imposes) upon pitch, attack-rhythm, and duration. For this reason, the listener ought provisionally to lay aside the obsolete and vague notion of \"timbre\" -- a bushel-basket for whatever aspects of sound we may in the past have relegated (however mistakenly) to the role of subliminally \"lushing-up\" pitch-production -- and ought instead to follow the individual participation of such aspects in the unfolding of the piece. (For a detailed discussion of these and allied problems, see my three articles in Vol. 5, no. 2 of Perspectives of New Music.)
[I should perhaps emphasize that the computer has been exploited in this piece solely as an instrument of performance, and not as a composer-surrogate. I used the MUSIC 4B computer-program (written by Godfrey Winham and Hubert S. Howe), a revised and greatly expanded version of the original MUSIC IV computer-program (written at Bell Laboratories by Max Mathews and Joan Miller). This program enabled me to invent and specify, in complete detail, every aspect of every sound in the computer part. The task of the Princeton University Computer Center's IBM 7094 computer was to numerically simulate the resultant total soundwave called for by my coded musical specifications -- in effect to numerically simulate the wiggly groove on this recording. The computer wrote this numerical simulation (in which each second of the total sound is represented as a single sequence of 40,000 numbers) on a large magnetic tape, which was in turn read by a separate machine at Bell Laboratories (in Murray Hill, N.J.) known variously as a \"data translator\" or a \"digital-to-analog converter.\" This machine impressed the 40,000 numbers per second alternately onto the two tracks of an ordinary musical tape as fluctuating voltage. Thus, strictly speaking my \"instrument of performance\" was not just the computer, but the computer plus the converter plus some ordinary stereo tape playback equipment which introduces actual sound into the process (for the first time) by playing back the converter-produced musical tape.]
The rhythmic relations of violin to computer (and of violin to violin) is nasty enough to require, if an intolerable burden is not to be imposed on the violinist during recording, a special time-beating arrangement. We adopted the expedient of having the computer produce a set of \"metronome\" tapes, whose sole use was to be played to the violinist through ear-phones while he played his part into the microphone. All violin and computer tracks were then transferred to sprocketed film (whose 96 holes per second guarantee accuracy of synchronization to within 1/2 of 1/96th of a second), from which the final master-tape was produced. The violin recording, as well as synchronizing, was done by the Herbert H. Hagens Recording Studio in Princeton, New Jersey; and financed, as was most of the recording, by Princeton University.\"
