Hi again,
Er, just realised - my granular objects are for r-t granular sampling, rather than synthesis per se. So basically, you sing into Pd and out comes your granul[ated]/[ised](can anyone advise on the correct term?) voice, or whatever. This is of course rather different to granular synthesis using sine bursts. The objects might still be useful though. I'll endeavour to put tar-balls up on our ftp server here at Glasgow (it'll take a couple of days for me to get it together).
regards, nick
Nicolas Leveille wrote:
Quoting Nick Fells nick@music.gla.ac.uk:
Hi,
I've got a couple of granular objects which I could email to you if you like. They're Linux only at the moment, if that's any good. They allow you to adjust the no. of granular streams, grain duration, 'randomisation' etc as they're graining away.
yes please, I'd be happy to try them! :)
By the way, is it just me or the output of [ osc~ ] sounds really bad? I think I hear some quantization noise when approaching low volumes. I didn't check the source code yet, but I suspect no dithering or interpolation is done.. is my assumption right?
-n