Granular synthesis is a method used to create sounds by working with very short pieces of audio. It is similar to sampling, which involves recording and playing back sounds. However, in granular synthesis, the recorded sounds are divided into tiny parts called grains, each lasting about 1 to 100 milliseconds. These grains can be combined together and played at different speeds, starting points, volumes, and pitches.

When the grains are played slowly, they create a type of sound often called a "cloud," which feels different from regular sounds. When played quickly, the grains form notes with unique qualities. By changing the shape, length, position, and how closely the grains are spaced, a wide variety of sounds can be made.

This technique has been used in music for effects, as a base for other sound changes, or as complete musical pieces. Common effects include changing the loudness over time and stretching sounds to be longer. More creative uses include spreading sounds across multiple speakers, rearranging them randomly, breaking them apart, or blending them into new sounds.

History

In 1947, Dennis Gabor proposed that sounds can be broken down into small parts called "grains," each grain being a short burst that includes information about time and pitch. Greek-French composer Iannis Xenakis is credited with developing granular synthesis, building on Gabor's ideas.

In 1974, American composer Curtis Roads became the first person to use granular synthesis on a computer. In 1986, Canadian composer Barry Truax created real-time granular synthesis using the DMX-1000 Signal Processing Computer.

Microsound

This includes all sounds that are shorter than musical notes, the sound object time scale, and longer than the sample time scale. Specifically, these sounds are shorter than one tenth of a second and longer than 10 milliseconds. This range includes part of the audio frequency range (20 Hz to 20 kHz) and part of the infrasonic frequency range (below 20 Hz, such as rhythm).

These sounds include short, sudden sounds and are known in acoustics and signal processing by names such as sound particles, grain, glisson, grainlet, trainlet, microarc, wavelet, chirplet, fof, time-frequency atom, pulsar, impulse, toneburst, tone pip, acoustic pixel, and others. In the frequency domain, they may also be called kernel, logon, and frame.

Micromontage is a type of music that uses very small sounds.

Microtime refers to the level of "sonic" or aural "syntax," which describes how sound energy changes over time in different frequency ranges.

Related software

• Csound – complete music software that includes tools for creating sounds by breaking them into tiny pieces (includes information about these tools)
• Max/MSP – software that uses visual elements to create programs for working with audio and video as they happen
• Pure Data (Pd) – a visual programming language that lets users create programs for working with audio and video as they happen
• SuperCollider – a programming language used to create sounds electronically in real time
• ChucK – a programming language designed for creating music with precise timing
• EmissionControl2 – a tool that creates sounds by using tiny sound pieces

Related hardware

• Mutable Instruments Clouds – a digital, open source eurorack music module that includes four preset modes. The first mode, and the one that starts automatically, is a granular processor.
• Make Noise Morphagene – a eurorack music module that uses microsound and granular synthesis techniques, as well as methods inspired by Musique Concrète for manipulating audio by combining sounds.
• Tasty Chips GR-1 – a polyphonic granular synthesizer that can produce 128 grains per note, allowing up to 1000+ grains to play at the same time.
• Instruō Arbhar – a eurorack music module released in 2019 that uses the 'Lexer Method' of granular synthesis. It automatically detects the start of sounds to sample into one of six 10-second audio buffers, supports mono or stereo settings, can produce up to 88 polyphonic grains, controls granular settings using voltage signals, and allows users to create custom configuration files.

Discography

• Curtis Roads (2004). CD with Microsounds. MIT Press. ISBN 0-262-18215-7. Includes parts of nscor and Field (1981). Microsounds is available on Discogs. nscor (1980), New Computer Music (1987), and Wergo 2010–50 are listed on Discogs with details about their releases.
• Iannis Xenakis. Analogique A-B (1959) is available on Discogs under Alpha & Omega and Music For Strings.
• Truax, Barry (1987). Digital Soundscapes is available on Discogs.