Martin, that's pretty good and a lot closer to how akihiko's sound than I've been able to get.
Hans I've downloaded the 'many' lib and I get this error when running the 200-voices patch voicetest~ 0 0 $0 ... couldn't create $1 $2 $3 $4 $5 $6 $7 $8 $9 ... couldn't create $1 $2 $3 $4 $5 $6 $7 $8 $9 ... couldn't create $1 $2 $3 $4 $5 $6 $7 $8 $9 in the main patch [voicetest~ 0 0 $0] doesn't create and the variables from 1-9 are in the [voices-wrap~] instances, I'm sure you knew that. On Tue, Jul 12, 2011 at 4:57 PM, Jeffrey Concepcion < [email protected]> wrote: > Hans, > Thanks for letting me know about the 'many' lib, I'll check it out. > And Martin I'll try out your patch asap. I think I read somewhere > about a [polyosc~] object or it could be a figment of my imagination. > > On Tuesday, July 12, 2011, martin brinkmann <[email protected]> > wrote: > > On 07/12/2011 01:55 AM, Jeffrey Concepcion wrote: > >> Also, it seems to me that I would need > >> one oscillator per note, which is ridiculous if you want to make > something > >> isn't that CPU intensive. > > > > it depends on the cpu. i have made a 'quick and dirty' cluster-synth > > patch (attached) with 192 not very (or at all) optimized oscillators, > > which produces about 20 percent load on my machine (3 ghz intel core). > > it should run on a atom-netbook. if you need more instances, it might be > > too much though. (and less voices do not sound that awful too...) > > > > another possible approach i can think of would be using ifft as a > > oscillator bank. though it is probably not trivial to calculate the > > spectrum. (or you will have to use a very large blocksize, to get enough > > 'frequency-resolution') > > > > bis denn! > > martin > > > > -- > www.epicjefferson.com > www.avmachinists.org Puerto Rico based Art Collective/ Non-Profit Org > -- www.epicjefferson.com www.avmachinists.org Puerto Rico based Art Collective/ Non-Profit Org
_______________________________________________ [email protected] mailing list UNSUBSCRIBE and account-management -> http://lists.puredata.info/listinfo/pd-list
