About

Fragment is a collaborative cross-platform audiovisual live coding environment with pixels based image-synth, the sound synthesis is powered by pixels data produced by live GLSL code.

Fragment has a progressive learning curve, anyone can start to experiment with sounds and visuals really quickly.

Fragment has many features making it a bliss to produce any kind of sounds which can be associated with visuals, it is aimed at artists seeking a creative environment with few limitations to experiment with, a programmable noise-of-all-kinds software, Fragment potential is almost boundless.

Many videos of most features are available on the official YouTube channel

For any questions, a message board is available here

The old single page documentation (still updated to some extent) is available here

The PDF documentation can be found here

All software is free of charge, and distributed according to free software/open source principles and licenses.

Requirement

  • Web. browser such as Chromium (recommended), Opera, Safari or Firefox (MIDI is not supported by Firefox at the moment)
  • Mid-range GPU, Fragment was made and regularly used with a GeForce GTX 970
  • Mid-range multi-core CPU, a beefy CPU may be needed for advanced features
  • Not necessary but a MIDI device such as a MIDI keyboard is recommended

A high amount of memory is needed to load many samples/distributed sound synthesis with the audio server, this is because all samples are pre-loaded for each instances.

Performances

Fragment has excellent performances with a modern multi-core system and a browser such as Chromium.

If you experience crackles or need advanced audio features, it is recommended that you use the audio server available as a download on the homepage.

The audio server is able to provide fast sound synthesis, it also provide many settings such as audio device selection, sample rate settings etc.

Moreover, the audio server is able to do distributed real-time audio synthesis on any machines or cores over the network by splitting the workload between servers instance, fas_relay is needed to use this feature