Audio: granular synthesis (0:10)
granular synthesis plays overlapping streams of grains. The granular synthesis figure shows the waveform:
Granular synthesis is a synthesis technique for manipulating grains.
A grain is a sample of microsound, a sound with a very short duration. The typical duration of a grain is between 10 and 100 milliseconds. This is the region where two discrete sounds blur into a single sound.
A grain is a sample of sound. It is created the same way as any other sample. The grains in granular synthesis are digital samples of a sine wave created by computer synthesis.
Each grain has an amplitude envelope. It is a simple envelope with two segments, a linear or exponential fade in and fade out. The envelope prevents the sound from clicking.
A group of grains is called a cloud. A cloud can contain hundreds or even thousands of grains.
Grains can be generated at a fixed rate or a variable rate. A grain cloud can consist of discrete grains or overlapping grains or a mixture of both. Clouds can be discrete or overlapping too.
grain has minimal sound processing. You can really go to town by adding other effects such as filtering, phasing and flanging.
Manipulating grains and clouds and adding sound effects gives granular synthesis its distinctive sound. The result is a subtle, shifting soundscape.
Granular synthesis is treated here as a method for writing synthetic harmony. It could equally well be treated as a method for writing noise. The two approaches to writing music are the same. In this way, the guide has completed a full circle back to where it started.
