MusicBrainz

• 1. The Database That Refused to Die

  ›

  MusicBrainz started as an act of defiance. When the Compact Disc Database, known as CDDB, placed restrictions on its data, a project rose to replace it with something no single company could lock away. That project, run by the MetaBrainz Foundation, now holds information on over 2.8 million artists, 5.4 million releases, and 38.7 million recordings. What began as a workaround to corporate data control became one of the most ambitious volunteer-driven music archives ever assembled. How did a scrappy alternative to a restricted database grow into the backbone of major broadcasting institutions? And what does it take to keep 38.7 million recordings accurately catalogued by unpaid volunteers?

• 2. Volunteer Editors and Style

  ›

  At the heart of MusicBrainz is a community of volunteer editors. These contributors maintain entries that cover, at minimum, album title, track titles, and the length of each track. They also follow community-written style guidelines to keep the database consistent. Beyond the basics, entries can store release dates and countries, CD identifiers, cover art, acoustic fingerprints, and free-form annotation text. The breadth of what a single release entry might contain reflects years of collective effort to capture music information in structured form. The MetaBrainz Foundation, a non-profit based in San Luis Obispo, California, took over stewardship of the project in December 2004 from its creator, Robert Kaye. That transfer formalized the community's ownership of what had become a shared resource, not a product.

• 3. Fingerprinting's Rocky Road

  ›

  In 2000, MusicBrainz began using a patented acoustic fingerprinting technology called TRM, a recursive acronym for TRM Recognizes Music, developed by a company called Relatable. The feature pulled in a large user base and drove rapid database growth. By 2005, TRM was buckling under the weight of more than one million stored tracks. In May 2006, MusicBrainz partnered with MusicIP, replacing TRM with a system called MusicDNS. TRMs were fully phased out in November 2008. Then, in October 2009, MusicIP was acquired by a company called AmpliFIND, throwing the future of the free identification service into uncertainty. That uncertainty proved decisive.

• 4. Chromaprint and AcoustID

  ›

  Lukáš Lalinský, a long-time MusicBrainz contributor, launched the Chromaprint acoustic fingerprinting algorithm in February 2010. Chromaprint analyzes the first two minutes of a track, detecting the strength for each of 12 pitch classes and storing those measurements eight times per second. A post-processing step compresses the fingerprint while preserving its identifying patterns. The resulting identifier feeds into the AcoustID search server, which matches recordings by similarity and returns both an AcoustID identifier and any associated MusicBrainz recording identifiers. Though AcoustID and Chromaprint are not officially MusicBrainz projects, both are open source and closely tied to the MusicBrainz ecosystem. The BBC announced on the 28th of June 2007 that it had licensed MusicBrainz's live data feed to enhance its music web pages, with BBC editors also joining the volunteer community directly.

• 5. Cover Art and Event Archives

  ›

  MusicBrainz allows contributors to upload cover art images of releases to the Cover Art Archive, a joint project with the Internet Archive that launched in 2012. Internet Archive supplies the bandwidth, storage, and legal protection; MusicBrainz stores the metadata and provides public access through both the web and an API. Until the 16th of May 2022, cover art was also sourced from Amazon and other online retailers, but the Cover Art Archive is now the preferred source because it gives the community greater control. Over six million images are now stored there. In June 2024, MusicBrainz and the Internet Archive launched a second joint project, the Event Art Archive, described as "the internet's greatest repository for event art" and holding over 9,500 images at launch.

• 6. Licensing and Open Data

  ›

  Since 2003, MusicBrainz's core data covering artists, recordings, and releases has been dedicated to the worldwide public domain under the Creative Commons Zero 1.0 Universal legal tool. Additional content, including user contributions and the live data feed, is published under the Creative Commons Attribution-NonCommercial-ShareAlike 3.0 Unported License. The database itself runs on PostgreSQL. The server software is covered by the GNU General Public License v2.0, and the client library libmusicbrainz is licensed under the GNU Lesser General Public License, version 2.1. That latter license specifically allows proprietary software products to build on the library. On the 20th of January 2006, the Barcelona-based company Linkara became the first commercial venture to use MusicBrainz data, through their "Linkara Música" service. The BBC's beta music site, launched the 28th of July 2008, went further: it published a dedicated page for every MusicBrainz artist in its catalog.

• 7. Picard and ListenBrainz

  ›

  MusicBrainz Picard is a free and open-source application for identifying, tagging, and organizing digital audio files. It matches audio files and compact discs against the database using either stored metadata or acoustic fingerprints, then writes new tags including recording artist, album title, record label, and release date. Picard supports a wide range of audio formats, among them AAC, ALAC, FLAC, Monkey's Audio, MP3, Ogg Vorbis, and WavPack. A sample set of tags generated by Picard for the album Piano Opera FINAL FANTASY VII / VIII / IX by Nobuo Uematsu illustrates the precision of this data: the entry records the barcode, catalog number, label, performer, release country, and even ReplayGain values down to two decimal places. ListenBrainz extends the MetaBrainz ecosystem in a different direction. It crowdsources listening data from media players and services, including Clementine, Music Player Daemon, and Spotify, and can import scrobbles from Last.fm and Libre.fm. ListenBrainz was designed to recover features lost when Last.fm was acquired by CBS, and its listening data is released under an open license available for music research in both industry and academic settings.