g r a n u l a r   s y n t h e s i s

 

 

definition:

 

Granular Synthesis is a technique that was pioneered, most noticeably, by Curtis Roads and Barry Truax, both of whom have fully explored the possible realms of this 'brassage' method. This implementation is technically speaking a synchronous single-voice time-domain granulation algorithm. So there.

What is a grain? It is simply an enveloped 'slice' of sound lasting in the vicinity of 20-60ms. By overlapping enough of these grains, which can stretch the original sound-file out to many times its original length, allowing the composer to 'get inside the sound', revealing its inner complexity.

parameters:

 

As with Sample Hose, Granular Synthesis 'inbetweens' initial and final values and variances. The possible values are:

Grain Length: The length of each grain. Values between 20-60ms are the norm, but try longer values for a 'pulsed' effect
Density: The percentage of grains which are created. A pretty useless parameter really. I'm sorry I put it in there...
Grain Shape: A number between 0 and 200. Represents the curvature of the enveloping (see how the graph changes when you change the number).
Grain Skew: Skews the enveloping to one side. Doesn't seem to affect the sound that much. Again, apologies for this parameter....
Grain Pan: Allows you to specify whereabouts in stereo space each grain is to be created. 0% = left, 50% = middle, 100% = right. For a complete random scattering, set both value and variance to 50%.
Spacing: The length of time after the previous grain has been created. For overlap, choose a negative value. For space between each grain, choose a positive value. I apologise for the lack of consistency with the Sample Hose.
New Length: The new file length. For a 500% stretch, for example, choose old length times... and enter 500.