Sir Mustapha
Group: Musicians
Posts: 2802
Joined: April 2003 |
|
Posted: Oct. 27 2007, 10:27 |
|
There are already computer programs that can generate music, starting from certain kinds of input from the user. Pete Townshend has the Lifehouse Project, inspired by an idea he had back in the 70's, which produces a "musical portrait" of a person, based on stuff like an image, a sample of voice, etc.. The results are interesting, and in my opinion, exactly because they don't focus on sounding "pleasant". Music can't be pleasant all the time without becoming pedestrian and monotonous. Music exists in dissonance, resolution of tension, movement of harmony, etc.. Even ambient music does that.
But anyway... well, Boards of Canada titled one of their tracks as Music Is Math. It is, to a certain extent. If you treat music as a matrix of notes, it is math. And just as it is possible to "teach" a computer to play chess (and even beat the world #1 champion at it...), it's perfectly possible to teach it to write music, and teach it to learn to do so. It's computationally possible, and no doubt there's a lot of research going on in that area. The problem lies in what is not in the notes - i.e., how relevant and important a piece of music is, given the social, historical, regional context and so on. Would Stravinsky's Rite of Spring cause a riot if it premièred today, for example? Would Sgt. Pepper's Lonely Hearts Club Band cause the globe to spin the other direction (metaphorically!! ) today like it did back in 1967? To me, that's the heart of the matter. Art is a social, cultural phenomenon. Maybe the idea of an electronic chip producing music out of nowhere could cause a cultural revolution, but the results of the process themselves? I'm not so sure.
-------------- Check out http://ferniecanto.com.br for all my music, including my latest albums: Don't Stay in the City, Making Amends and Builders of Worlds. Also check my Bandcamp page: http://ferniecanto.bandcamp.com
|