diff options
author | Guenter Geiger <ggeiger@users.sourceforge.net> | 2002-07-29 17:06:19 +0000 |
---|---|---|
committer | Guenter Geiger <ggeiger@users.sourceforge.net> | 2002-07-29 17:06:19 +0000 |
commit | 57045df5fe3ec557e57dc7434ac1a07b5521bffc (patch) | |
tree | 7174058b41b73c808107c7090d9a4e93ee202341 /pd/doc/4.fft.examples/00.INTRO.txt | |
parent | da38b3424229e59f956252c3d89895e43e84e278 (diff) |
This commit was generated by cvs2svn to compensate for changes in r58,
which included commits to RCS files with non-trunk default branches.
svn path=/trunk/; revision=59
Diffstat (limited to 'pd/doc/4.fft.examples/00.INTRO.txt')
-rw-r--r-- | pd/doc/4.fft.examples/00.INTRO.txt | 65 |
1 files changed, 65 insertions, 0 deletions
diff --git a/pd/doc/4.fft.examples/00.INTRO.txt b/pd/doc/4.fft.examples/00.INTRO.txt new file mode 100644 index 00000000..b5218793 --- /dev/null +++ b/pd/doc/4.fft.examples/00.INTRO.txt @@ -0,0 +1,65 @@ +These patches demonstrate how to use Pd's short-time Fourier transform objects, +rfft~ and rifft~, to do a variety of things. The patches can be quite +expensive; the phase vocoder, for instance, requires a 300MHz Pentium 2 to run +at 44100 Hz. By default Pd runs at 44100 Hz, but you can specify a different +sample rate on the command line, for instance: + +pd -r 16000 09.pvoc.pd + +On SGI, Pd will check whether your audio system is running at the correct +sample rate and will print out a warning if not. In NT or W95, Pd's behavior +will depend on your audio driver. In Linux Pd usually tries to set the rate +of the conversion hardware. + +Included in this directory are: + +01.fftanalysis.pd -- does a windowed FFT analysis and resynthesis of a sine +tone, showing how to specify block size and overlap. + +02.noisefft.pd -- Fourier analysis of white noise. This patch also shows how +you can average power spectra over time. + +03.denoise.pd -- using the technique from the previous patch, finds the noise +floor in a recorded sample and attempts to scrub it away. + +04.shifts.pd -- tests the leftshift and rightshift "externs" used in the next +patch. + +05.sheepgoat.pd -- tries to discriminate between "pitched" and "unpitched" +components of a sound. + +06.sheepgoat2.pd -- another attempt at the same thing. + +07.tinbell.pd -- a spectral flattener, which can make the sound of a large bell +turn into the sound of a tamtam. + +08.convobros.pd -- spectral cross synthesis between two sounds + +09.pvoc.pd -- phase vocoder + +---------- after this point, the patches haven't been cleaned up ----------- + +10.phaselockedvoc.pd -- phase locked vocoder; see Laroche&Dolson in ICMC97 for +a discussion of something that works better than this. + +11.pianorev.pd -- an attempt at a phase-coherent reverberation algorithm to +imitate the "piano reverb" obtained by putting a speaker under the sound board +of a piano and picking up the sympathetic vibrations. + +12.sinedecomposer.pd -- estimate the frequencies and amplitudes of the +components of a sound + +13.tracemaker.pd -- show how to use Pd to combine snapshots of a spectrum into +continuous spectral traces. + +14.partialtracer.pd -- sinusoidal analysis/resynthesis of a time-verying sound. + +15.waveformgrab.pd -- bash a sample into phase-coherent windows and make a +pitched resynthesis. + + + + + + + |