On Tue, May 25, 2010 at 3:33 PM, Chris Cannam <[email protected]> wrote:
> I think the point Neils has is just that the outcome of your noodling > is somewhat independent of your explicit intention. Notes that sound > satisfying together are probably going to sound satisfying largely > because of some intrinsic mathematical relationship, or at least > something that is probably open to analysis to some extent but that > you don't yourself understand or plan. Quite an interesting > philosophical avenue here, and one that's fairly well trodden in other > fields (ask an English theory student about Wimsatt and Beardsley). I've been reluctant to weigh in, because I just know I'm going to blah blah about math, waste time, and no one will care :) I'm good at it. When we consider the analysis of what sounds good, we are probing a psychological question. The mind, being completely un-observable and distinct in study from the brain itself, is impossible to measure directly. We can model the mind, and analyze whether or not our model fits with observed behavior. The actual intrinsic "math" that goes on in the brain and its counterpart, the mind, cannot be exactly known. So, what we do is model the various interacting processes that make up our direct experience of music and sound. This approach is, in fact, objective despite the fact that we may not be able to explain all of the significant interpersonal and moment-to-moment sources of variation that affect our experience. It's really an exciting time for the study of music psychology (and I've been saying so for 10 years). The degree to which computers can compose music depends on the success of modeling musical experience in humans. As musicians and composers, we approach the "tiling problem" with a set of techniques, instruments, and vocabulary. We are able to get direct, immediate feedback on the effectiveness of a giving tiling, which computers, at present, cannot. Currently, computers have expanded our techniques and instruments while people have expanded their vocabulary to compose new and novel music with them. The point I'm getting at: the structure isn't in the music itself, it's in the mind of the listener. I've been toying around with the idea of modeling high-dimensional psychoacoustic spaces as non-linear manifolds (I'll skip the subject for now, b/c I'm not sure I can describe it). Because I intend to work on it, I do think that successive approximations through modeling are possible that will push the outer boundary of musical vocabulary and instruments further than musicians and composers alone could. Chuck _______________________________________________ Linux-audio-dev mailing list [email protected] http://lists.linuxaudio.org/listinfo/linux-audio-dev
