The Music Genome vs Audio ID
The Music Genome Project began in the year 2000 as a way “to capture the essence of music at the fundamental level”, a process that takes a trained musician 20 to 30 minutes per song. Pandora uses your first song request to find music with similar attributes within the Music Genome. As you rate suggestions, your station is tuned based on aggregate results of your selections and ratings, therefore creating a playlist specifically tailored to your tastes.
The music genome has a few flaws. Despite the unbearably detailed categorization criteria in the genome, they all rely on human interaction to populate the data. Enter Audio ID, a technology owned by German Company Magix AG, that uses a different set of categories that are assigned by an automated system. From my experience with these two systems, Music Genome is superior at providing a variety of enjoyable music, while Audio ID is more efficient at finding mixable music.
For the average person, Audio ID will provide no usefulness whatsoever, but if you would like to see it in action, you can do so at Magix AG’s latest creation, http://www.mufin.com. Mufin will take a song suggestion, and give you a list of songs that sound similar to the one you entered. This is great for a House or Trance DJ that is trying to keep a consistent groove on the floor. Specify that you want a song with a similar bass line and equal tempo (commonly called BPM), and Mufin will do its best. The website seems to contain mostly low-quality samples of obscure songs, however you can import your own (non-iTunes purchased) music if you have the pro edition, included in the latest edition of Magix MP3 Maker. Which will even give you 3d view of your music library, which is highly customizable to your needs.