hi i placed a spectral-filtering patch (and some others) at ftp://www.parasitaere-kapazitaeten.net/pdpatches marius.
Am Donnerstag, 31. Januar 2002 00:00 schrieb Miller Puckette:
It's not the same approach, but there's a simple spectral-envelope-drawing example in
http://man104nfs.ucsd.edu/~mpuckett/172.01s/week1/
in 02_draw_spectrum.pd and the abstraction osc-voice.pd ... I'll think about cleaning it up to stick in the example patches.
Another possible approach is to use the envelope as a filter; rfft~ any signal, multiply right and left outputs by the "tabreceive", recombine via rifft~.
I don't think that directly synthesizing from the desired FFT spectrum will work very well, but you could try it by just sticking the tabreceive~ straight into the left-side input of the rifft~...
cheers Miller
On Wed, Jan 30, 2002 at 11:01:21PM +0100, Frank Barknecht wrote:
hi all,
Miller Puckette hat gesagt: // Miller Puckette wrote:
I agree. Spectral envelope drawing, anyone?
Yes, I'm playing with this. And while this is brought up, I have a probably simple question, which does bother me anyway:
Say I want to make an additive synthesis instrument, that generates its sound by an rifft~ object. This object should receive information about the magnitude of certain frequencies. This should be done with a (drawable) array, where the x axis is frequency and y is magnitude. The array gets read by a tabreceive~ and has a size = window blocksize. Now: How do I have to preprocess this xy-axis data to let rifft~ generate the matching sound composed of the input frequencies.
Any hints?
[I have some understanding of the FFT, but for this problem I just don't get a solution. Call me stupid.]
Ciao,
__ __
Frank Barknecht ____ ______ ____ __ trip\ \ / /wire ______ / __// __ /__/ __// // __ \ / / __ \ ___
/ / / ____/ / / / // ____// /\ \ ___\____
/_/ /_____/ /_/ /_//_____// / \ \_____\_____
/_/ _\