Apparatus and method for determining a prominent tempo of an audio work
View Patent ↗The prominent tempo of an audio data is determined by detecting a plurality of beat rates of the audio data. One or more audio data characteristics are used to filter through the beat rates to determine the prominent tempo. Systems, methods, and apparatuses to determine the prominent tempo are discussed herein.
1. A method of determining a prominent tempo of audio data, the method comprising:
using a processor to derive a mood profile of the audio data;
comparing the mood profile to a plurality of mood-tempo maps;
determining a tempo class of the audio data based on the comparing of the mood profile to the plurality of mood-tempo maps;
determining a baseline tempo measurement;
calculating a differential by comparing the baseline tempo measurement to the determined tempo class; and
correcting the baseline tempo when the differential exceeds a threshold.
2. The method of claim 1 , wherein the comparing of the mood profile with the plurality of mood-tempo maps comprises comparing mood scores of the audio data to mood profile scores for a plurality of mood profiles, each mood profile being associated with at least one mood-tempo map.
3. The method of claim 1 , wherein the comparing of the mood profile includes:
assessing a similarity between the mood profile and at least one mood-tempo map of the plurality of mood-tempo maps; and
assigning a tempo class to the audio data in reference to the assessed similarity.
4. An apparatus comprising:
a processor configured to execute a mood analysis module configured to determine a mood profile of the audio work;
a tempo classification module configured to classify the audio work in a tempo class;
a beats per minute (BPM) estimation module configured to determine a baseline BPM of the audio work; and
a BPM correction module configured to determine a corrected BPM of the audio work based on a comparison of the tempo class and the baseline BPM,
the BPM correction module being configured to correct the baseline BPM when the tempo class and the baseline BPM exceed a threshold.
5. A non-transitory machine-readable medium comprising instructions that, when executed by one or more processors of a machine, cause the machine to perform operations comprising:
deriving a mood profile of audio data;
comparing the mood profile to the plurality of mood-tempo maps;
determining a tempo class of the audio data based on the comparing of the mood profile to the plurality of mood-tempo maps;
determining a baseline tempo measurement;
calculating a differential by comparing the baseline tempo measurement to the determined tempo class; and
correcting the baseline tempo based on the differential exceeding a threshold.
6. The non-transitory machine-readable medium of claim 5 , wherein the comparing of the mood profile with the plurality of mood-tempo maps includes
comparing mood scores of the audio data to mood profile scores for a plurality of mood profiles, each mood profile being associated with at least one mood-tempo map.
7. The non-transitory machine-readable medium of claim 5 , wherein the comparing of the mood profile with the plurality of mood-tempo maps includes
assessing a similarity between the mood profile and at least one mood-tempo map of the plurality of mood-tempo maps; and
assigned a tempo class to the audio data in reference to the assessed similarity.