On 2011-01-19, Wen X wrote:
Convolution is equivalent to polynomial multiplication. Presumably we all know how to factorize a polynomial?
That's the normal way to do it, yes. But if you want to "factorize" a polynomial into an arbitrary number of parts, not necessarily equal to the number of its divisors, and you want to have all of the resulting parts translating into equally heavy computational parts after you translate the polynomial theory into actual filter designs, and then you furthermore want to have those smaller polynomials/convolutions/filter topologies feeding lots of parallel cores, using clean code, yet distributing well, working well in order so as not to muck up your cache, etc, ad nauseam...
It's enough to make a stronger man both laugh and cry. -- Sampo Syreeni, aka decoy - de...@iki.fi, http://decoy.iki.fi/front +358-50-5756111, 025E D175 ABE5 027C 9494 EEB0 E090 8BA9 0509 85C2 -- dupswapdrop -- the music-dsp mailing list and website: subscription info, FAQ, source code archive, list archive, book reviews, dsp links http://music.columbia.edu/cmc/music-dsp http://music.columbia.edu/mailman/listinfo/music-dsp