Evolutionary music

From Wikipedia, the free encyclopedia

Evolutionary music is the audio counterpart to Evolutionary art, whereby music is created using an evolutionary algorithm. Evolutionary sound synthesis is a related technique for generating sounds or synthesizer instruments. Evolutionary music is typically generated using an interactive evolutionary algorithm where the fitness function is the user or audience, because it is difficult to capture the aesthetic qualities of music computationally.

[edit] Hypothetical example of an evolutionary music system

1. Randomly generate 20 piano rolls.
2. Play each roll, and give it a score (high if you like it, low if you don't).
3. Throw away the 10 rolls with the lowest score.
4. Create 10 new rolls by copying parts of the 10 best rolls, occasionally introducing or deleting new notes at random.
   • Note: most people will want to emulate biological evolution here by taking pairs of piano roll 'parents' and performing a kind of sexual recombination on them. However, any kind of 'copy/combine with error' should work.
5. Go to 2 (except you only need to listen to the 10 new piano rolls).

After a while, the music will sound more pleasing.