XIM Timbre · measured 25 Oct 2026
Training data for sound matching and preset inference
Sound-matching models need the preset, the notes that drove it and the audio that came out as one triple; the presets exist by the hundred thousand, the sessions show them in use, and the audio is rendered on order.
Each song comes with
- the preset file
- the MIDI that drives it
- on order: the rendered audio per preset and note set, with the synth and version named
Built from
- Synth presets Patches for the synths dance producers use, from sold sound banks and sessions.
- MIDI files Note data from sessions and MIDI packs: chords, leads, basslines, drums.
- DAW sessions The producer's project file with its audio and MIDI, as it was when the track was finished.
How it ships
Readiness: built on order; first delivery scoped per synth family on the brief. Delivery to a private S3 prefix or your bucket, with a manifest, checksums, creator-level splits and a datasheet.
Loader: one folder per song with data.json.
Rights
From sold banks and sessions under terms that name AI training. Read the rights page.