Granular synthesis is a method of digital sound synthesis that uses many thousands of very short (usually less than 100 milliseconds) overlapping "micro-sounds", the so-called grains. The waveforms of these grains are often sinusoidal, although any waveform can be used. One alternative to sine waves is to use the grains cut from sampled, pre-recorded, or live captured sounds. By manipulating the time distribution of a large number of grains, arranging them in layers, forming the so-called granular cloud (grain cloud) and manipulation of many other parameters, such as the frequency, playback speed, amplitude envelope, volume, waveform or panorama arrangement, can create a huge number of very complex and changing over time sound colors.
The "granular cloud" can be understood as the fluctuating regions of sonic energy that appear to be moving in the space.
Many composers, such as Iannis Xenakis and Barry Truax, have considered granular synthesis and granulation techniques to be a means of shaping large masses of sound. Both of these composers are considered to be pioneers of this concept (Truax was the author of the first granular synthesis computer programs).