On 09-20, Fons Adriaensen wrote: > The thing to be avoided is code that e.g. generates a heavy load > every fourth period and does almost nothing in the three periods in > between. Or a synth that generates a peak load whenever a voice is > started and much less while the note lasts. IIRC yoshi/zyn use FFTs > in some of the algorithms, so this sort of thing could happen.
Yeah, the cost of starting an additive synthesis note can be fairly expensive as with a worst case instrument and default settings you'd need to do 256 1024-point IFFTs per part and a bunch of math leading upto the transform (in addition to all sorts of other initialization stuff (like the memory allocation which I just finally removed)). I've got this cost almost removed in my local version, but it's not release ready. --Mark
pgpYukn1_bs57.pgp
Description: PGP signature
_______________________________________________ Linux-audio-dev mailing list Linux-audio-dev@lists.linuxaudio.org http://lists.linuxaudio.org/listinfo/linux-audio-dev