Ok, thanks. I guess if i don't care much about invertibility, nor speed, I can just look for resonances -- in other words, make myself a table: notes=: ^0j2p1*(^.2^1+(i.61)%12)*/t
and take the inner product of this with some sound sample (which uses the same t), and then find magnitude: %:(*+)notes +/ .*sample In principle, at least, the more energy there is at some frequency, the greater the resonance amplitude. Well, it's a theory, anyways... now I just have to dig out some code to import sound samples and try this out. Anyways, thanks again for the tips and links, -- Raul ---------------------------------------------------------------------- For information about J forums see http://www.jsoftware.com/forums.htm
