Re: re[2]: granular looping

["Dennis Leas" <dennis@mdbs.com>]

> ...the definition of granular looping is easy...take a bunch of grains of
> sand (or small rocks), and arrange them in a loop.  the worlds
> pre-eminent granual looper (imho) is andy goldsworthy...and excelent
> example may be found here:
>
> http://cgee.hamline.edu/see/goldsworthy/gold_pebbleshl.gif

That seems like a good definition to me.  (Where's Dr. Z?)

Speaking of granulation - Kyma has a "Sample Cloud" goober that performs a
kind of granulation on a sample file.  I went nuts when I first found it.
Here's a description for the benefit of your mental health...

>From Symbolic Sound's description of Sample Cloud:
"Generates a cloud of short-duration grains, each using GrainEnv as an
amplitude envelope on a short
segment of sound taken from the specified Sample at a point in the sample
given by the TimeIndex. The
density of simultaneous grains within the cloud is controlled by Density,
with the maximum number of
simultaneous grains given by MaxGrains. Amplitude controls an amplitude
envelope over the *entire*
cloud (each individual grain amplitude is controlled by GrainEnv).
Similarly, Duration is the duration of
the entire cloud, not of each individual grain. You can control the
Frequency, stereo positioning, time
point within the sample, and the duration of each grain as well as
specifying how much (if any) random
jitter should be added to each of these parameters (giving the cloud a more
focused or a more dispersed
sound, depending on how much randomness is added to each of the
parameters)."

(I'll explain a little more...)
So the grains are extracted from the Sample file.  You can pick a totally
random selection of grains (occuring anywhere in the file) or totally
ordered selection of grains in whatever order you want.  The duration of
each grain can likewise vary from a fixed value to a random value.  Next, 
an
amplitude envelope is imposed on each grain.  Then each grain is placed
somewhere in the left-to-right stereo field.  Again, this can be totally
deterministic, totally random, or somewhere in between.  Finally, you can
determine the maximum number of simultaneous grains that you want to hear.
You can easily produce more than 28 simultaneous grains; each grain is
individually selected from the sample file, with an individual duration and
pan.

Cool stuff!

Dennis Leas
-------------------
dennis@mdbs.com