The recent death of Milton Babbitt brought the avant garde composer’s life and work a tiny burst of Internet prominence. One interesting thread in his career was his involvement with the mighty RCA music synthesizer, a unique device that appeared most of a decade before the better-known systems developed by Robert Moog and Donald Buchla. The synthesizer’s advocates at RCA described it as being capable of duplicating any imaginable sound. Of course it was not, but its limitations had the positive effect of showing researchers how far they still had to go to duplicate the subtleties of acoustic instruments. For Babbitt, however, composing highly abstract music outside the traditions of classical instrumental performance, the big synthesizer’s simplistic and unfamiliar sounds were quite useful.
Twentieth century composers had already experimented with sounds produced by interconnected collections of manually controlled electronic oscillators, resonant filters, and amplifiers. The RCA Mark I and Mark II synthesizers went beyond this by linking its (vacuum tube) analog audio components together with a primitive (but enormously costly) digital sequencer built around something like a player piano roll. Each vertical column of holes punched in the paper tape was the time history of a single “bit” of control data, and each horizontal row contained the state of all of the control bits at one moment in time. As each hole passed through the reading device, an electrical circuit was closed. The pattern of “on” and “off” signals derived by the reader from each row of hole positions was therefore equivalent to the binary instructions used by a computer.
In the RCA synthesizers, a collection of electromechanical relay switches played the role of a computer’s circuitry for decoding instructions. The “bits” in the synthesizer’s “instructions” were used in groups of four according to the rules of binary math. Four columns of hole/no-hole positions could be used as a four-bit binary number. The sixteen possible values of this number (i.e., sixteen distinct on-off patterns) would then be used to select one of up to sixteen alternatives at the corresponding switching point in the system’s interconnections. In this way, the composer controlled the pitch, timbre, and loudness of each sound the system produced.
The RCA engineers made one insanely clever choice that was soon overruled by Babbitt and the other people who had to use the device: they supplied a direct mechanical connection – a flexible drive cable – between the paper tape transport and the turntable of a disc-cutting lathe. The special multichannel discs used by the machine could then be recorded in multiple passes for overdubbing, and everything would be in perfect sync – if nothing broke or jammed or otherwise went awry on any of the passes. If something went wrong, the users would start over with a new disk. The users decided instead that mid-1950s magnetic recording tape technology was a lot more practical than the direct-to-disc alternative.
Because the four-channel Mark II system cost half a million 1959 dollars to build, the one built for the Columbia-Princeton Electronic Music Studio – used mainly by Milton Babbitt – was the first and last to be produced.