The Machine that Learned to Sing
A look at the first computer music
4/4i — Episode 1
Machines That Learned to Sing
The Birth of Computer Music (1950s–1970s)
Hosted by The Composer
Episode Description
How did computers learn to make music?
Long before modern AI music generators, a small group of engineers and composers were teaching room-sized computers to produce sound. In 1957, Bell Labs engineer Max Mathews wrote the first widely used computer program for digital sound synthesis, launching the field of computer music.
Over the next two decades, breakthroughs at Bell Labs, the University of Illinois, and Stanford University’s CCRMA transformed computers from calculating machines into musical instruments.
In this episode of 4/4i, we explore the pioneers, machines, and ideas that created the foundation for modern digital music and AI composition.
What You’ll Learn in This Episode
- How the IBM 704 computer generated the first digital music
- Why Max Mathews is called the father of computer music
- How the MUSIC programming language shaped modern synthesis software
- The first algorithmic composition created with a computer
- The invention of FM synthesis
- The founding of CCRMA at Stanford
- The role of the Samson Box digital synthesizer
- Early computer music compositions like Dreamsong
Chapter Guide
Chapter 1
Max Mathews and the First Computer Music
Computer music began in 1957, when Bell Labs engineer Max Mathews wrote the program MUSIC I for the IBM 704 mainframe computer. The system generated digital sound waves directly from mathematical calculations and produced a short 17-second composition.
This experiment demonstrated that computers could synthesize audio instead of merely playing recorded sound.
Key takeaway
Digital music began as an engineering experiment.
Sources
Computer History Museumhttps://www.computerhistory.org/revolution/computer-graphics-music-and-art/15/222
History of Informationhttps://www.historyofinformation.com/detail.php?id=3886
Stanford Newshttps://news.stanford.edu/stories/2011/05/max-mathews-father-computer-music-dies-84
Chapter 2
The MUSIC Programming Language
Mathews expanded his early experiments into the MUSIC-N family of programming languages, which allowed composers to define digital instruments and control them with musical scores.
Versions included:
- MUSIC II (1958)
- MUSIC III (1960)
- MUSIC IV and V
This modular approach introduced the idea of unit generators, software building blocks for oscillators, envelopes, filters, and other synthesis components.
Modern audio programming languages like Csound evolved directly from this system.
Source
https://en.wikipedia.org/wiki/MUSIC-N
Chapter 3
The First Computer Composers
Bell Labs soon became a meeting place for composers and scientists exploring new musical ideas. Engineers collaborated with experimental composers interested in algorithmic structures, spectral analysis, and new digital timbres.
Computer music emerged as a hybrid field combining:
- acoustics
- signal processing
- mathematics
- composition
These collaborations laid the groundwork for digital music studios.
Chapter 4
The ILLIAC Suite — Algorithmic Composition
While Bell Labs focused on sound synthesis, Lejaren Hiller and Leonard Isaacson at the University of Illinois used computers to generate musical scores.
Their 1957 composition “The ILLIAC Suite for String Quartet” used probability and rule-based logic to create musical structures.
This experiment is widely considered the first algorithmic composition created with a computer.
It demonstrated that computers could influence musical decision-making, not just sound generation.
Chapter 5
Computers Become Instruments
During the 1960s, the MUSIC systems evolved into powerful synthesis platforms. Composers could design entirely new sounds using mathematical descriptions of waveforms and envelopes.
Instead of recording instruments, composers could construct sound itself.
This approach formed the foundation for modern digital synthesis and sound design.
Chapter 6
Computer Music Spreads
By the mid-1960s universities began building computer music laboratories.
Researchers explored:
- digital sound synthesis
- algorithmic composition
- signal processing
- acoustic modeling
Computer music was no longer a single research project. It had become an international field combining science and art.
Chapter 7
FM Synthesis at Stanford
In 1967, Stanford composer John Chowning discovered frequency modulation synthesis (FM synthesis).
FM synthesis allowed complex timbres to be created by modulating the frequency of one oscillator with another.
Stanford later licensed the technology to Yamaha, which used it in the famous DX7 synthesizer, one of the most influential instruments of the 1980s.
Sources
https://www.soundonsound.com/people/john-chowninghttps://en.wikipedia.org/wiki/John_Chowning
Chapter 8
The Creation of CCRMA
In 1975, Stanford founded the Center for Computer Research in Music and Acoustics (CCRMA) with John Chowning as director.
The center brought together composers, engineers, and computer scientists to study:
- digital synthesis
- psychoacoustics
- sound modeling
- algorithmic composition
CCRMA quickly became one of the most influential computer music research centers in the world.
Chapter 9
The Samson Box
In 1977, CCRMA installed the Systems Concepts Digital Synthesizer, commonly known as the Samson Box, designed by Peter Samson.
This specialized digital synthesizer allowed composers to generate sound much faster than earlier mainframe methods and became the primary synthesis system at CCRMA for over a decade.
Source
Computer Music Journal researchhttps://www.researchgate.net/publication/262389654_Life_and_Times_of_the_Samson_Box
Chapter 10
Early Computer Music Compositions
One of the early landmark works created at CCRMA was Michael McNabb’s “Dreamsong” (1978).
The piece combined digital synthesis with electroacoustic composition techniques and became one of the earliest widely discussed computer music works.
Source
https://en.wikipedia.org/wiki/Dreamsong
Key Figures
Max MathewsJohn ChowningLejaren HillerLeonard IsaacsonPeter SamsonMichael McNabbJean-Claude RissetKarlheinz Stockhausen
Recommended Listening
John Chowning — Turenas (1972)
John Chowning — Stria (1977)
Michael McNabb — Dreamsong (1978)
Next Episode
Next time on 4/4i:
We travel to Paris to explore IRCAM, the groundbreaking research institute founded by Pierre Boulez, and the European tradition of computer music that helped transform digital synthesis into a global art form.