hi
i tried to create a convolution filter (if this is the correct expression for what i want), but i failed. i want to filter an audiosignal with an impulse response (e.g. of a room) or the transfer function of a loudspeaker. i read the docu about 'fft and pd' from marius schebella (http://www.parasitaere-kapazitaeten.net/Pd/fft_und_pd_en.htm) and i tried the patch from page 5. when i use an impulse response, where the first sample is 1 and the rest is 0, it works, which means that the output of the convolution filter is the same as the input. but when i change the position of the sample with value 1, i get bad glitches in the sound. the same, when i use any other impulse responses than the (dummy) one i described first. i also tried the patch from page 10 (convolution filter) which is the same as the pd-doc -> 4.fft-examples -> 08.convobros.pd. but this one seems to work more as a vocoder than like what i want. if somebody understands what i want and i am trying to explain --> hints or links are very welcome. also if somebody have already made a patch to generate reverberation with room impulse responses by convolution, i am very interested.
cheers
roman
Roman Haefeli wrote:
hi
i tried to create a convolution filter (if this is the correct expression for what i want), but i failed. i want to filter an audiosignal with an impulse response (e.g. of a room) or the transfer function of a loudspeaker. i read the docu about 'fft and pd' from marius schebella
either use [FIR~] (iemlib) or read about OLA (overlap and add); tm did an open-source tutorial on this, but i think he might have forgotten to put it somewhere....
mfg.a.sdra IOhannes
You might be interested in my partconv~ external:
http://puredata.org/Members/bensaylor/partconv~-0.1.tar.gz/view?searchterm=p...
Ben
On Wednesday 02 February 2005 07:52 pm, Roman Haefeli wrote:
hi
i tried to create a convolution filter (if this is the correct expression for what i want), but i failed. i want to filter an audiosignal with an impulse response (e.g. of a room) or the transfer function of a loudspeaker. i read the docu about 'fft and pd' from marius schebella (http://www.parasitaere-kapazitaeten.net/Pd/fft_und_pd_en.htm) and i tried the patch from page 5. when i use an impulse response, where the first sample is 1 and the rest is 0, it works, which means that the output of the convolution filter is the same as the input. but when i change the position of the sample with value 1, i get bad glitches in the sound. the same, when i use any other impulse responses than the (dummy) one i described first. i also tried the patch from page 10 (convolution filter) which is the same as the pd-doc -> 4.fft-examples -> 08.convobros.pd. but this one seems to work more as a vocoder than like what i want. if somebody understands what i want and i am trying to explain --> hints or links are very welcome. also if somebody have already made a patch to generate reverberation with room impulse responses by convolution, i am very interested.
cheers
roman
PD-list@iem.at mailing list UNSUBSCRIBE and account-management -> http://iem.at/cgi-bin/mailman/listinfo/pd-list
-
Ben Saylor -
IOhannes m zmoelnig -
Roman Haefeli