Abstract: A method, apparatus and computer memory are provided for recognizing an audio fingerprint of an unknown audio recording. A database stores a plurality of audio recording identifiers corresponding to a plurality of known audio recordings, where the audio recording identifiers are organized by variation information about the audio recordings. A processor searches a database and identifies at least one of the audio recording identifiers corresponding to the audio fingerprint, where the audio fingerprint includes variation information of the unknown audio recording.

Abstract: A method, apparatus and computer memory are provided for generating an audio fingerprint of an audio recording. A memory stores stable frequency family data corresponding to a plurality of stable frequency families. A processor curve fits audio recording data to at least one of the stable frequency families, extracts at least one variation from the curve fitted audio recording data, and creates the audio fingerprint of the audio recording from the at least one variation.

Abstract: A system and method are provided for matching a track set from a digital audio recording to metadata relating to the recording. Track duration data is obtained for the track set, and the track duration data is rounded. A search is performed for matching records in a first database based on the rounded track duration data, each resulting matching record having an identifier. Track duration data is retrieved from a second database based on the identifiers associated with the matching records. If more than one matching record is found, the track duration data retrieved from the second database is compared to the track duration data obtained for the track set to find a best matching record of in the second database. Metadata contained in the best matching record of the second database is output.