switch to the U.K site >
Granular Synthesis uses small slices of sounds (‘grains’) to compose new sounds from existing material. By combining multiple grains of differing lengths, amplitude, pitch and speed creates very characteristic sounds of modern music. Xenakis claims to have invented the technique and indeed his ‘Analogique A-B’, composed of tiny tape splices of pure tones, is credited as the first piece of granular music in 1959.
Tape editing proved extremely time consuming, but by the 1970s digital processing could take the place of tape splicing. Curtis Roads dove into the early computer based granular synthesis and made some of the classic techniques known through his recordings, teaching and texts like ‘Microsound’. Today, most computer audio programs have some sort of granular synthesis engine or plugin. Dr. Richard Boulanger has used granular synthesis in Sound to great effect and he is also a beta tester for this card.
The French phrase for the cartridge is “Grain de Folie” which could be translated as “seeds of madness”, but in French “grain” also translates to “grain”, and “madness” evokes the strange disassembling/reassembling granular process. Also, “avoir un grain de folie” is a typically French expression to describe people behaving in a non conventional way, thus a fitting play on words. How it works: Granular processing requires a block of memory to hold digital samples for playback, and the Z-DSP has one second of memory for the audio used in processing.
From this audio buffer the grains will sample smaller sections for playback. The number of grains in the process determine how dense the overall output sounds. These programs have 3, 4 or 6 grains for playback. Each grain plays from a random point in the audio buffer and have an independent envelope controlling their duration. The envelope time is the ‘grain size’ parameter in many of the programs. In the context of the Z-DSP, the FV1 (the DSP brain) is really not designed for grain synthesis (due to technical choices like a “circulating” delay memory, and the lack of indirect memory access), but the chip also has other design niceties that help overcome its limitations…
This cartridge implements a simple and customised granular synthesis with a limited number of grains, and parameters that mainly control the size of grains and their positions in the sample. One nice aspect of the Z-DSP is that it uses live inputs (granular synthesis is usually based on a pre-recorded sample), so it can disassemble live input and reassemble it in real-time into a different order resulting in a (usually !) nice sonic transformation of both texture and the rhythm.