Nuance vs Subsample

Nuance has no write-up on Specifiction yet; its figures are above.

Subsample is a standalone sampler that turns live input or recordings into collections of sounds that can be played from MIDI. It separates and sorts sound events, then maps them to MIDI notes by acoustic similarity without labels, training data, or manual tagging. In live sampling, it captures, trims, analyses, and adds distinct events to an instrument library in real time. For longer samples, onset detection can identify hits and place them on a beat grid at a target BPM. Its processors include filters, dynamics, distortion, radio, vocoder, pitch shifting, time stretching, and quantisation. It accepts WAV, BWF, FLAC, AIFF, OGG, and MP3/MPEG files, and supports MIDI input, OSC, and MIDI control bindings. Audio runs from mono through 7.1, with per-instrument routing and first-order ambisonic capture and playback. Subsample has no graphical interface and runs from a terminal, using YAML files and a MIDI map. It is an early release, and those formats may change. The open-source AGPLv3 license requires source availability for modified versions used as network services; proprietary use requires a commercial licence.