Audio Fingerprinting

What Is Audio Fingerprinting?

Audio fingerprinting is a technique that identifies a recording by converting a short snippet of its sound into a compact digital signature, then matching that signature against a database of known tracks. It works from the audio itself, so it can recognise a song from a few seconds captured in a noisy room.

Audio Fingerprinting in Practice

Shazam made the technology famous: hold up a phone in a bar, and within seconds the app names the song by matching its acoustic fingerprint against a vast catalogue. The same principle now runs quietly behind the scenes of the music business. YouTube’s Content ID fingerprints uploads to detect copyrighted recordings, and monitoring services use fingerprinting to track where and how often music is broadcast. For brands this matters twice over: fingerprinting is how rights holders detect unlicensed use of commercial music in campaigns, and it is also how a brand can verify where its own commissioned tracks and idents are appearing. It identifies the recording itself, unlike an audio watermark, which requires embedding a signal beforehand.

Famous Example

Shazam, which names a song from seconds of audio in a noisy room