History of Electroacoustic Music

Bell Labs in the 1960s

Mathews and Pierce continued in their enthusiasm and advocacy of the computer as a musical instrument. In 1960, Pierce contacted Lejaren Hiller at the University of Illinois.

James Tenney
During a visit to Hiller's facilities, Pierce met composer James Tenney (1934-2006), who was finishing his master's work there in composition. In 1961 Tenney came to Bell Labs to work as a composer in residence in psychoacoustics and computer music. He remained there until 1964. Tenney became the first composer to take an interest in computer music research (in contrast to Mathews, who was a computer researcher with an interest in music).

During the period of his residency, until 1964, Tenney created a series of tape pieces, and wrote software that used the computer to make musical decisions in ways that had not been explored in music previously. While Mathews' software allowed music to be created in a traditional form -- a printed score, encoded on punchcards -- Tenney is credited with being the first to develop the computer as a means of realizing compositional intelligence, and Mathews was more than happy to provide Tenney with software to bring computer music to new levels.

Tenney had a variety of "extra-musical" interests that he brought to the creation of music with the computer. One was the nature of human perception of sounds, the application of Gestalt psychology to auditory perception. Another was in information theory, to which he had been exposed during his studies with Hiller. Another was in stochastics, the application of probability to the generation of many events, an interest shared by Xenakis, though the two did not work together. Finally, there was his embrace of the theories of John Cage, randomness, and chance operations.

Among Tenney's work created at Bell Labs were:

  • Analog #1: Noise Study (1961)
    This was inspired by the traffic noises Tenney heard when passing through the Holland Tunnel, and the varying nature of the noise of ocean surf. Tenney described "learning to hear these sounds more acutely, to follow the evolution of single elements within the total sonorous 'mass,' to feel, kinesthetically, the characteristic rhythmic articulations of the various elements in combination, etc." Tenney created the piece with a noise generator, filters, and an amplitude modulator, and used the computer to interpolate changes in characteristics, such as center frequency or bandwidth, over a set period of time.
  • Dialogue (1963)
    This piece juxtaposes sine tones and noise bursts, with the relationships of characteristics such as note duration, density of notes produced, and loudness, controlled according to probability distributions.
  • Phases (for Edgard Varèse) (1963)
    Although one of the more difficult of Tenney's pieces for those unaccustomed to computer music, this piece has been described as "not composed by either man or machine, but some goblin-hybrid of the two...Listening to it often seems like eavesdropping on a silkenly beautiful alien conversation...The subtle use of noise, pitches just on the level of audibility, miniscule glissandi that remind one of the inner patterns in rice paper, are all in their ethereal way quite moving. It is as if Phases is a surreptitious window on an ongoing perceptual and sonic communication from another dimension."
  • Ergodos II (for John Cage) (1964)
    Also created as a series of parameters and probabilities that change over set periods of time, the resulting analog tape could be played either forwards or backwards

image courtesy of Max Mathews

Music IV
In 1962 Music IV was written, also for the IBM 7094 computer. It was a re-coding of Music III in a new assembly language. It also included implementations of resonant (regenerative) digital filters. Princeton University also had an IBM 7094, and Mathews was more than willing to share the program with other who were interested in exploring it. So Princeton also began to use Music IV. They made some improvements, and called their version Music IV-B.

In 1963 Mathews published "The Digital Computer as a Musical Instrument" in Science:

There are no theoretical limitations to the performance of the computer as a source of musical sounds, in contrast to the performance of ordinary instruments. At present, the range of computer music is limited principally by cost and by our knowledge of music appears to be very promising technically. However, the method will become significant only if it is used by serious composers. At present, our goal is to interest and educate such musicians in its use.

Courtesy of Max Mathews

In 2006, Mathews was to recall that this was the most important article he had ever written, because it was read by two future innovators, Jean Claude Risset and John Chowning.

Creating music with these programs had a number of limitaitons. Speed and memory capacities of computers limited the range of available frequencies to under 10 kHz. Furthermore, realizing music was not real time by any means. The output samples were stored on computer tape. These tapes then had to be fed to another computer that could convert the samples to audio to be stored on audio tape. People from Princeton would bring their tapes to Bell, with usually a two-week turnaround time for development. It was extremely time-consuming. It was often hard to know whether a piece actually worked, or whether it sounded as it was intended, until it was transferred to audio tape at Bell and then returned to Princeton or Columbia to be played. Simple tests could take weeks to perform.

Computer music reached popular culture with Jonathan Pierce's "Portrait of the Machine as a Young Artist," which appeared in the June 1965 issue of Playboy. Science fiction writer Arthur C. Clarke heard a vocal synthesis demonstration prepared by Mathews of the computer singing "A Bicycle Built for Two." Some years later, this arrangement was quoted in the film 2001: A Space Odyssey, the screenplay of which was written by Clarke.

Jean Claude Risset
In France, Jean Claude Risset (1938-) read the Science paper. At the time, he was doing graduate work in physics in Paris. Enthused by Mathews' article, Risset obtained a grant from a French research agency to do his thesis research at Bell Labs. In 1964 Risset succeeded Tenney as research composer in residence. He was impressed with the interdisciplinary environment at Bell, which he found far less territorial than in European research centers. There was also no pressure for an immediate application, as they were involved in long-term, speculative research.

Risset became one of the first to research timbre with the new digital tools. He began with the trumpet, making digital recordings, and then spectral analyses. He found that describing the timbre was not a simple matter, and in fact there was no one timbre for the instrument at all. For one thing, the relative amplitudes of the harmonics was not fixed, but dynamic. Thus, to synthesize the trumpet effectively, the spectrum had to change in the right frequencies and at the right rates. The higher harmonics entered later than the lower harmonics. Also, the spectrum varied with loudness of the instrument -- when it was played louder, there were more higher harmonics. Risset's work is a fascinating study of acoustics applied to music, and his compositions remain early classics in the field of computer music.

Liner notes, James Tenney: Selected Works, 1961-1969, Artifact Recordings, 1992.
Max Mathews, keynote address at International Computer Music Conference, New Orleans, 2006.