Synthesizing sounds from scratch is all well and good, you just use a bit of maths. However, the latest build from [Daisy] eschews such boring concepts as additive or subtractive synthesis, instead going for a sample-based approach.

This build is based around the Daisy Seed microcontroller platform. It was actually inspired by an earlier project to create a ribbon synth, which we covered previously. In this case, the ribbon potentiometer has been repurposed, being used to control the playback position of a lengthy recorded sample. In this build, the Daisy Seed is running its audio playback system at a rate of 48,000 samples per second. It’s capable of storing up to 192,000 samples in memory, so it has a total of 4 seconds of sample storage. The Daisy Seed uses an analog-to-digital input to record two seconds of audio into the sample buffer. It can then be replayed by placing a finger on the ribbon at various points. Playback is via granular synthesis, where small sections of the overall sample buffer are used to synthesize a new tone. The video explains how the granular synthesis algorithm is implemented using the Plugdata framework. Design files are available for those eager to replicate the build.

Once you start tinkering in the world of synthesis, it’s easy to fall down the rabbit hole. Video after the break.