Hi,

I am trying to build a granular audio-file processor which allows independent control of speed & pitch, and which allows alteration to both parameters throughout continuous audio file playback.

I have copied over a patch described in Farnell's Designing Sound which offers most of what I am after. However in Farnell's patch, every time an adjustment to speed is made, it also needs to restart audiofile playback for the new parameter to function (see [r $0-speed]).

My patch is attached. Does anyone have any ideas on how I can modulate the speed parameter without restarting audiofile playback?

pd_grains.zip

Instructions:

1. load "grain_host.pd"
2. click "load_wav"
3. select audio file "voice1.wav"
4. click "play"
5. modulate parameters.

Thanks !!

Peter

One simple way is to use a phasor~ to generate the playback position.

The speed of the phasor~ should be inversely proportional to the length of the audio file segment in seconds.

Then, the output of the phasor should be multiplied by the length of that segment in samples, and then multiplied again by the speed factor

cool, thanks mod!

might you have a simple technique to make speed modulation sound less choppy?

[edit]
updated patch. . . .
pd_grains2.zip

Hello, i downloaded you patch and the object [grainvoice $1] can not be created. Could you give me a tip?

The phase vocoder from the Pd help examples (Audio examples -> I07) does exactly what you are looking for.