This open-source software works by chopping samples up into a ‘brain’ of small, interconnected sections called blocks, which are grouped into a network based on their similarity.
This then processes a target sample, chopping it up into blocks in the same way, and attempts to match each block with one in its brain for realtime playing.
The idea is that you’re interpreting one sound with another. There are plenty of tweakable parameters - some with intriguing names such as ‘novelty’, ‘boredom’ and ‘stickyness’ - and, pleasingly, the software is said to be “slightly out of control”.
If that all sounds a bit techy and difficult, allow us to reassure you that the workflow is actually pretty straightforward. You simply need to load some samples into the brain, click the (re)generate brain button, load a loop sample into the target, click the (re)generate blocks button, press play and then tweak the brain parameters to taste.
Discussing the origins of the software (it’s actually been in development since 2015) and what it’s capable of, Aphex Twin said: “What if you could reconstruct source audio from a selection of other MP3s/audio on your computer?
“What if you could build a 303 riff from only a cappellas or bubbling mud sounds?
“What if you could sing a silly tune and rebuild it from classical music files?
“You can do this with Samplebrain.”
While Aphex Twin worked on the design of Samplebrain, it was brought to life by Dave Griffiths of non-profit organisation Then Try This. Nik Gaffney helped out with the Apple builds.
Samplebrain runs on PC and Mac (Intel and silicon). You can download it now via GitLab.