Roland0 wrote: > That may be more involved than you'd expect. > - first, you'd need to identify the audio features suited to this task > (see 'A Review of Physical and Perceptual Feature Extraction Techniques > for Speech, Music and Environmental Sounds' > (https://www.mdpi.com/2076-3417/6/5/143/htm) for an overview, and 'this > page' (https://essentia.upf.edu/streaming_extractor_music.html) for > essentia_streaming_extractor_music's capabilities) > - then, you'd find a way to compute the similarity for each feature > (e.g. musly uses the Kullback-Leibler divergence for MFCCs) > - finally, you'd have to combine the similarity of all features and find > a way to compare songs > > Note that if you use essentia_streaming_extractor_music, you'd also have > to configure it to store all frame values (by default, only the > statistics computed over frames are included in the output). At this > point, using Essentia's API directly would make more sense.
Thanks, but that is probably a bit too much for me! Think I'll stick with the high-level descriptors. I've analysed ~20k tracks - which took quite some time at roughly 15 tracks/min (still quicker than MIP). To locate 'similar' tracks I'm checking the difference in 'danceable', 'aggressive', 'electronic', etc between a seed track and tracks in the DB. Tracks are then sorted based upon average difference of these parameters, and a random selection of the top few chosen. This seems to work reasonably well for me. Code is not ready for release, as I'm still playing with it - but can be found in my github page. *Material debug:* 1. Launch via http: //SERVER:9000/material/?debug=json (Use http: //SERVER:9000/material/?debug=json,cometd to also see update messages, e.g. play queue) 2. Open browser's developer tools 3. Open console tab in developer tools 4. REQ/RESP messages sent to/from LMS will be logged here. ------------------------------------------------------------------------ cpd73's Profile: http://forums.slimdevices.com/member.php?userid=66686 View this thread: http://forums.slimdevices.com/showthread.php?t=108278 _______________________________________________ plugins mailing list [email protected] http://lists.slimdevices.com/mailman/listinfo/plugins
