Fragment is a collaborative cross-platform audiovisual live coding environment. It uses a pixel-based real-time image-synth approach to sound synthesis.

Fragment is able to produce high-quality fast additive and granular synthesis simultaneously with re-synthesis support. According to the developer, ‘it has many features making it a bliss to produce any kind of sounds or visuals and is aimed at artists seeking a creative environment with few limitations to experiment with’.

The video, above, is a one-hour ambient soundscape, made of granular and additive synthesis patches with live compositing visuals produced from two videos. This demonstrate granular synthesis features, videos looping/compositing and mixing sound synthesis methods.

Features:

• Complete additive, spectral, granular synthesizer powered by WebAudio oscillators, a wavetable OR
• Fragment Audio Server
• Complete audio/visuals live coding environment with JIT compilation of shader code
• Real-time, collaborative app.
• Distributed sound synthesis, multi-machines/multi-core support (Audio Server with fas_relay)
• Stereophonic or monaural
• Polyphonic
• Multitimbral
• 32-bit float images data (WebGL 2 only)
• Multi-output channels per slice
• Shader inputs (webcam, images, videos with audio analysis, audio files analysis, drawing over textures…)
• MIDI in
• OSC in/out
• Spectral recording with export and re-import as a texture
• Audio synthesis can be done on a dedicated computer on the network
• Per-sessions discussion system
• Sessions based, no authentifications, ability to run locally

See the Fragment site for more info. 

via Sonic State