GRAIN

mother misty · Posted: Wed Dec 15, 2004 6:07 am Post subject: GRAIN

hello,

I've noticed that several people (including myself) are very interested in granular synthesis. Wouldn't you like to see some grain modules like a grain delay (maybe similar to the one in reaktor)?

zynthetix · Posted: Wed Dec 15, 2004 10:39 am Post subject:

would this be achievable through some clever use of S&H modules?

mosc · Posted: Wed Dec 15, 2004 10:52 am Post subject:

Be sure to see Rob's granular synthesis patch here: http://electro-music.com/forum/topic-2718.html

The Kyma system has tremendous granular synthesis capability and I've played around with it quite a bit, but I have never really gotten too turned on by this technique. Rob's patch does a pretty good job though.

jksuperstar · Posted: Wed Dec 15, 2004 11:14 am Post subject:

I think at the least I'd like a delay module that has mod inputs for BOTH the write head location and read head location (a red or blue signal scaled to cover the range of the delay, with the delay having a range similar to the current dealys--5ms, 20ms, 100ms, etc). Combine this with a couple oscillators, and you can make many things. This would help with granular synthesis, odd tape delays, pitch shifting, etc.

ian-s · Posted: Thu Dec 16, 2004 8:42 pm Post subject:

Does anyone have links to interesting examples of granular synthesis? As a technique, t's been around for quite a while now, someone must have done something good.

Rob · Posted: Fri Dec 17, 2004 3:45 am Post subject:

jksuperstar · Posted: Fri Dec 17, 2004 9:35 am Post subject:

http://www.granularsynthesis.live.com.au/index2.html

Hadn't thought of it before, but in essence, true granular synthesis is very much like the Fourier based analysis/synthesis (Fourier Series, Convolution, FFT, etc). Except 1 thing: it replaces the use of sine waves with any waveform as the basic building block of all waveforms. So, you don't have an infinite number of waveforms needed to represent a signal, which means that time is now part of the equation (which it is not in the Fourier relations).

I didn't know grains were developed back in the old communications days. Seems everything that is DSP these days is based on work done back in the 30's to improve the telephone system.