Flexistentialism Welcome Guest | Login or Register

A service everyone needs to check out is the new MusicBrainz. It is a service that identifies music, with a searchable database. In that respect it is very similar to services like FreeDB and Gracenote’s CDDB. However, a fundamental difference is that instead of just generating CD identification numbers, MusicBrainz analyzes the audio of the track and generates a TRM id. This ID number will be the same for a track on a CD, an mp3, an ogg file, a wav file, etc, because it is generated based on acoustic data, not just sector data from a CD track.

The organization is a non-profit, and all their software, server and client side, is GPL. Their entire database is free data as well. This helps alleviate concerns that they will turn evil, like CDDB did when it was acquired by Gracenote.

Currently, MusicBrainz software is rudimentary, but effective. It only exists for Windows at the moment, though the library that handles the actual generation of the TRM is available for unix, and development is in progress for other operating systems. I may attempt to port the library to OS X (I already tried compiling it, and it almost got there, so porting shouldn’t be too hard), and then write some software to go with it, but I’m afraid that I’m not a good enough programmer yet. We’ll see. However, even if I don’t do it, someone surely will. This service could revolutionize media identification, and they are even planning on expanding to video file identification, and more.

The software currently loads up a list of your MP3s, generates TRM ID’s for each song, and checks with the database to see if the song is already there. If so, then the tag information is downloaded, and your file is updated. If not, you are given a method to search the database for songs that have been inputted, but don’t have a TRM yet, and you can link them up. Finally, you can specify the artist, album and track name yourself, and MusicBrainz will create new artist and album entries in its database. You can also import albums from FreeDB. When you are done, you hit the submit button, and all the TRM ids you generated get sent to the server where they are added to the database to help other people identify their music. Then you hit the save button, and all your files get their ID3 tags updated, and optionally renamed.

Another benefit MusicBrainz has over other services of its type, is that the content is user-moderated. Say I go in and find that the track Lunar Cycle by Man With No Name is in the database, but some jerkoff couldn’t spell, and put it in as Luner Cicle. With other services, I just have to live with what it finds, but here, I can submit a change. The change gets added to a list of changes that people vote on. If my change is voted in, the title gets updated. Every user can make moderations, and vote on other moderations. There is a rudimentary karma system in place that rewards people who fix a lot of mistakes.

If you create new albums and songs in the database, they urge you to pay attention to the style rules, which include information on how to capitalize names, how to indicate multiple disc sets, how to indicate collaboration songs, etc. This is to keep the database searchable and consistent.

Right now I’m going through my files, which is quite a mighty undertaking, and updating tags and generating TRM IDs. My taste in music is quite eclectic, so I’ve already had to create about 15 different albums and artists on the database. I’ve taken the time to make sure all the data is accurate and that my entries follow the style rules. Its that kind of quality participation that will help the database for everyone.

Oh, and just so you know, AudioScrobbler and iScrobbler support for TRM IDs and the MusicBrainz server are planned for the future!

Feed on comments to this Post