David Cope, composer and pioneer in the field of algorithmic composition, which in the 1980s developed a computer program to write music in the style of Johann Sebastian Bach, Wolfgang Amadeus Mozart and other classic masters, died on May 4 at his home in Santa Cruz, California. He was 83 years old.
The cause was congestive heart failure, said his son Stephen Cope.
Before the proliferation of IA music generators, before the emergence of Spotify and the advent of the iPod, before Brian Eno even invented the term “generative music”, Mr. Cope had already understood how to program a computer to write classical music.
It was in 1981 and, struggling with the writer’s block after being responsible for composing an opera, he was desperate with a composition partner. He found one in a disk disc.
The process was simple but tedious. Mr. COPE began by quantifying the musical passages of his own work, making them as numbers in a database that could be analyzed by an algorithm for identifying the models he created. The algorithm would then reset the “signatures” – the name of Mr. COPE for the models he found – in new combinations, and he would convert these combinations into a score.
It was not the first time that someone used a computer to create music. In 1957, Lejaren Hiller and Leonard Isaacson had used a five Supercalculator at the University of Illinois to compose “Illiac Suite”, widely considered as the first score generated by computer. But Mr. COPE’s program goes further: by digitizing and reproducing unique signatures, his algorithm could mainly reproduce the style.
After years of troubleshooting and fine adjustment, the program, known as experiences in musical intelligence, was able to produce a complete opera in a few hours. Emi, or Emmy, as Mr. Cope affectionately called, was officially born. It was one of the first computer algorithms used to generate classical music.
Thank you for your patience while we check the access. If you are in reader mode, please leave and connect to your Times account, or subscribe to all time.
Thank you for your patience while we check the access.
Already subscribed? Connect.
Want all the time? Subscribe.