Full text

The discovery was born from a mistake. In 1967, while attempting to model the singing voice on a Stanford mainframe, John Chowning accelerated vibrato into audio frequencies (above 20 Hz). Instead of a rapid fluctuation in pitch, the ear perceived a complex, rich timbre with discrete sidebands. Unlike additive synthesis, which required dozens of oscillators for each overtone, the FM algorithm calculated highly complex inharmonic spectra of percussion and brass using just four parameters. In 1973, Chowning published the paper "The Synthesis of Complex Audio Spectra by Means of Frequency Modulation," which is still considered the standard reference material for this technology.

Working with early digital systems demanded absolute calculation: sound was not generated in real-time, and compiling a few seconds took hours. For his piece "Stria" (1977), Chowning wrote an algorithm in the SAIL programming language, where double modulation spectra were aligned not to classical notes, but to strict mathematical proportions of the golden ratio, creating a "crystalline," alien sound.

The academic world did not immediately grasp the commercial potential of the discovery. American manufacturers ignored the patent, but a visiting Japanese engineer from the Yamaha Corporation understood the principle in 10 minutes. Released in 1983, the DX7 synthesizer revolutionized the market. The record-breaking licensing royalties (the largest in Stanford's history until the invention of Google's algorithms) made the CCRMA center for computer acoustics, founded by Chowning, completely independent.