SOURCE treats a CC-licensed sound library as a live instrument rather than a static sample bank

Most samplers work from a fixed local library loaded before a session. SOURCE inverts this: its ‘instrument’ is Freesound’s 500,000+ Creative Commons sounds, fetched in real time by text query during use. The player types a search term, SOURCE downloads matching results and maps them to MIDI notes, and the set immediately changes timbre. Because every sound in Freesound carries a CC license (CC0, CC-BY, or CC-BY-NC), SOURCE is designed to remain license-compliant from the start — there is no moment where the user ‘forgets’ to check a license. This approach converts a sound library into a generative palette that is never the same twice: running the same query returns a randomised subset of results. The design insight is that treating a queryable online corpus as the raw material of an instrument blurs the line between sampling, searching, and live improvisation.

Examples

A performer types ‘rain’ into SOURCE before a set; it downloads 8 Freesound rain sounds, maps them across a two-octave keyboard range, and the performer plays them. Repeating the search next night yields a different 8 sounds from the same pool.

Assessment

Explain why SOURCE is described as ‘a sampler that does not sample’. What would a session look like if Freesound’s servers were offline? What is the key difference between SOURCE’s workflow and loading a sample pack?