Granular grains are less than 50 ms long, typically 10-30 ms

A grain is a single, short acoustic event used as the building block of granular synthesis. Barry Truax’s definition (following Xenakis and Roads) sets the upper limit at 50 ms, with typical values of 10-30 ms. This brevity matters: heard alone a grain is ‘the merest click or point of sound’, so pitch and texture do not come from any individual grain but from assembling hundreds or thousands of them per second. Learners commonly confuse grain duration with note duration or with the attack time of an ADSR envelope — these are unrelated parameters. The grain’s shortness is what lets the granular model treat sound as a fluid of micro-events rather than a sequence of notes.

Examples

At 20 ms grain duration with 500 grains/second, the average overlap is 10 simultaneous grains. Raising grain duration to 40 ms doubles overlap and smooths the texture; dropping to 5 ms moves toward an icy, grainy, eventually noise-like timbre.

Assessment

State the maximum grain duration in the standard granular model and the typical range. Predict how changing grain duration from 10 ms to 40 ms affects the perceived texture of a granular pad (at fixed density). Explain how grain duration differs from note duration.