Orm Finnendahl wrote (10 Apr 2002 10:53:20 +0200):
I'm just asking because my project will probably contain some heavy fft processing and I'm trying to find out whether it's necessary to implement other fft routines than the existing ones to minimize cpu-load.
FFT in PD was killing my old PC, so last year I was planning to hack PD with a fast FFT library. Now that I have a better PC it's less of a priority. There's a couple good FFT libraries out there: http://cr.yp.to/djbfft.html http://www.fftw.org/~skral/ http://www.fftw.org/pfftw-0.03.tar.gz The last one is optimised for MMX and is fast; it's experimental stuff, but I tried it once with some C application and it was working very well. Then I tried to replace the mayer fft calls in d_fft.c with the ones from the pfftw library. It didn't work so maybe someone have an idea how to do it.
Hi,
Thanks a lot for the information Mark. I will check this out when I'm getting to the audio part of the project (sometime next month probably).
Yours, Orm
Am Donnerstag, den 11. April 2002 um 13:11:32 Uhr (-0400) schrieb Marc Lavallée:
Orm Finnendahl wrote (10 Apr 2002 10:53:20 +0200):
I'm just asking because my project will probably contain some heavy fft processing and I'm trying to find out whether it's necessary to implement other fft routines than the existing ones to minimize cpu-load.
FFT in PD was killing my old PC, so last year I was planning to hack PD with a fast FFT library. Now that I have a better PC it's less of a priority. There's a couple good FFT libraries out there: http://cr.yp.to/djbfft.html http://www.fftw.org/~skral/ http://www.fftw.org/pfftw-0.03.tar.gz The last one is optimised for MMX and is fast; it's experimental stuff, but I tried it once with some C application and it was working very well. Then I tried to replace the mayer fft calls in d_fft.c with the ones from the pfftw library. It didn't work so maybe someone have an idea how to do it.
-- Marc
-
Marc Lavallée -
Orm Finnendahl