On May 13, 2009, Clint Adams wrote: > On Wed, May 13, 2009 at 04:18:25PM -0600, Gordon Haverland wrote: > > Anyway, a few ideas to chew on. Maybe they have occured to > > some of you before. But is your database the place where > > this kind of information can be stored? Should it be stored > > there? > > I don't think we want to compete with MusicBrainz; certainly if > there are cases for libre.fm to use it as a data source, we > should explore those.
Certainly cooperating is something to do. The only thing I thought MusicBrainz was really trying to get from people listening to music, was a fingerprint of the music to recognize it (I have not investigated it, this just seems to be the only unique thing I remember of MusicBrainz). But, if people would like to play music "at random" from Libre.fm, they should be able to find an appropriate definition of random to choose from. The thing that I recognize about the fingerprints MusicBrainz is interested in, is that if the original source is a WAV, and we have listeners using MP3, Ogg, Flac and other kinds of encodings, I believe they all get different fingerprints. Even though, they are supposedly the same source. With things like bugs in floating point hardware, the same Ogg encoded on different platforms could show up multiple times. Users may want to have Libre.fm only send MP3. Or they might be able to handle anything except Flac at more than 16 bit. I know my Mom's hearing isn't great, but she enjoys things like Spanish guitar music. A 24k MP3 stream is just fine for her, she couldn't hear the difference with higher fidelity anyway. Some people have very good hearing, and don't want to listen to low bitrate streams. Gord _______________________________________________ Libre-fm mailing list [email protected] http://lists.autonomo.us/mailman/listinfo/libre-fm
