Hacker News new | past | comments | ask | show | jobs | submit login

Working on hard stuff is in my bones, but it'd be nice at times to be able to live in a la-la land where monetization was completely irrelevant.

If I, say, "solved the money problem" in the sense that Paul usually uses the term, i.e. had a few million piled up, I'd probably get a small team together to rethink the way that music composition on computers is done. Start with a lot of research and interviews about how composers actually build up a piece and just build the best system for expressing musical ideas. I'd like to get around to that someday. It'd be cool being able to ignore the fact that it's a small, poor customer base. ;-)

The fact that computer interfaces are still designed to mimic tape recorders or wire cabling is a embarrassing.




That'd be cool. One music related pie-in-the-sky project I would like do would be to create software that could in real time give my voice the timbre of a rock star. I'd change a setting and switch from Axl Rose voice, to Bono, to Robin Wilson.


Well, the thing there is that you'd have to re-synthesize the voices, wave-shaping wouldn't be enough, so what you'd basically want is a vocoder. The problem, of course, is that synthesis of human voices isn't advanced enough to handle those kind of dynamics, so the problem where you'd end up focusing is on voice synthesis.


There is a research group at Cambridge (Oxford??) that is working on a problem like this.


I've done this - you've got the right idea in solving the "money problem" first. ;-)




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: