bozho

Arthur Clarke is one of the futurists that has imagined things right: http://www.youtube.com/watch?v=FxYgdX2PxyQ

strange. It works on chrome on windows.

I'm not missing it - all three things are encoded there. They might not be that "visible", which is a point of improvement, but the theory is there. As I wrote in the blog "random" generates noise. Th…

Creative Commons is the one I picked

sure, any feedback is welcome. The thing is still in an experimental phase, and I'm trying to improve it bit by bit

Will look for a better soundbank, indeed

I've read some papers about software used to help live performers, some sort of co-performer, or accompanying robot. But I can't find the paper right now..

they sound pretty nice :) will read the paper immediately

I deliberately made it generate/play tracks with 0 user input. This thing might be good, but I couldn't make it play anything meaningful

true that, I'll try to improve the general structure, so that it sounds complete

Nope. Pretty interesting, though. Will read about it more.

Great. The composer doesn't need to have gone through emotional breakdown, drugs, or whatever, to create beautiful music. He just needs talant. And I think talant may be at some point expressed in zer…

True that about the growing market. And true that few of these products/companies are really anything novel. But you have to do things better than the rest (I've discussed that more extensively here …

The idea of "liking" and "disliking" is to: 1. Get a collection of 'best' tracks that sort-of advertises the successful parts of the algorithm 2. To let me analyze which intermediate decisions in the …

I have indeed looked at some genetic algorithms and a lot of papers on the topic (in fact, my former university has a course about that, so I read the coursebook). Most of what I have read is either i…

My point is similar, but not exactly the same. The best human composers will still be the best. However, the other 90% may be replaced by computers. Or, computers may generate some "skeletons", which …

bad, or artificial? If "artificial" - that's the performance side, and it depends on the soundbank. If a real performer plays the same score, it would sound better. The point is to have nice compositi…

A lot of random() + a lot of composition rules :)

Currently the algorithm isn't learning. But I am storing a lot of parameters and then run analysis to see which ones are liked, and which - not