Freesound's API supports content-based search filtering by automatically extracted audio descriptors

Beyond keyword/tag search, the Freesound API accepts filters on audio descriptors extracted automatically from every uploaded sound — including pitch, BPM, and many spectral/timbral features (listed in the Audio Descriptors Documentation). This allows queries like ‘give me all sounds with a fundamental pitch of 440 Hz ±0.1 Hz’ or ‘find sounds with a BPM of 120’. Crucially, these features are machine-extracted and might not always be accurate, so results should be treated as approximate and auditioned. This descriptor-based search lets you source samples by musical criteria — a kick near a target pitch, a loop at a target tempo — without relying on uploader-supplied tags.

Examples

curl “https://freesound.org/apiv2/search/?filter=pitch:[219.9%20TO%20220.1]” Finds sounds whose extracted fundamental pitch is near A3 (220 Hz).

Assessment

A live coder wants only sounds whose BPM is between 118 and 122. Write a filter query-parameter string using the descriptor filter. Then explain why they should audition results rather than trusting the extracted BPM blindly.