A basic guide to Music Metadata and why it matters
Byta in partnership with MusicBrainz is creating a guide to understanding and utilizing music metadata. This post covers the basics.
Announcing a three-part article series, Metadata Matters. Byta and the team at MusicBrainz know that metadata is fundamental to digital music. Each article will tackle the importance of metadata when creating, sharing, and promoting music and how it relates to publishing, licensing, and getting paid.
What is Metadata?
In the modern music world that increasingly plays out online, music metadata becomes increasingly important. But what does metadata mean and how does it affect your music and how people can discover it?
The most basic music metadata for a track that an artist releases comprises the name of the artist (e.g. a band or solo performer), the name of the track and, if at all possible, the name of the release (single, EP or album) that the track appears on.
But, this is really absolutely the bare minimum data required to release a recording. Ideally it would have a lot more data associated with it, such as the release date, duration of the track, artist disambiguation (enough info to keep similarly named artists separate), copyright information or even lyrics.
When music fans listen to a recording online, however, they might like to know more information about the recording, album and/or artist. MusicBrainz tracks a lot of information to help music fans discover more about their favorite artists. Home pages, social media pages, artist support pages (e.g. Patreon) download/streaming locations are all good examples of supplemental metadata that should be present if you would like your music to be discoverable online.
However, the metadata rabbit hole goes a lot deeper – so far we’ve only discussed what we call public metadata. There are many more pieces of private metadata about music that the public never or only rarely sees. Some of the most important data is about which artists actually performed on the recording and their portion of the revenues that the recording generates, which are often called the splits. All too often these splits are never properly documented and quickly forgotten after everyone heads to the pub for a pint after the recording is done. This has led to many lawsuits as recordings get popular over time and people feel that they didn’t earn their fair share from the recording.
There are many more pieces of private metadata than we can realistically talk about in this short article. Copyrights, publishers, a full list of all the contributors and a way to legally identify them are among those, but since they are not publicly available, they don’t tend to help budding artists much, so let’s not focus on private metadata.
Why should I care about this? How does Metadata help me?
Sites like Discogs and MusicBrainz have user maintained databases of music metadata – all of the public metadata we’ve already touched on and a lot more. MusicBrainz in particular assigns each data element a unique identifier, so that the public can tell one artist from another (MusicBrainz tracks nearly 50 artists with the name “Void”).
These identifiers allow us to keep track of the Japanese release of an album with an extra track separate from the French version that has one track missing, even if they both still have the same cover art and name. Without a unique identifier it would be impossible to keep track of this mess that is so common in the music industry today.
Music metadata is simply often confusing, wrong or incomplete. It is rather important that resources like MusicBrainz exist to let anyone (artists, fans, labels, etc) look up what is known about a piece of music and how to contact the artist (through the official website and/or social media). Consider this: You may release the next amazing recording that captures the interest of the public for a moment. Great! But, what if a label that wants to sign you cannot find you? Will you miss out on that opportunity?
The MusicBrainz data is used by Google/YouTube, Universal Music, the BBC, Microsoft, Amazon and Pandora/SiriusXM among many others. If you get your metadata into MusicBrainz, soon thereafter your metadata will also be available to all of these companies!
The data is also used by sites like Last.fm and ListenBrainz, both of which have music fan communities who find new artists, new music and other music lovers. These sites recommend music to users who are interested in discovering new music and if your metadata is not available in MusicBrainz these sites may not be able to recommend your music.
More Metadata challenges
As we’ve already learned, metadata makes it possible to find your music. But data can go stale surprisingly quickly – for instance if your official web page URL changes, what can you do? Clearly it is not possible to edit all the audio files that you may have already distributed, so now what?
And worse, what happens when private metadata changes that you don’t even have access to? Copyrights are sold from music label to label frequently, which makes it hard for people to find the copyright owners for a piece of music. The best possible thing to do is to update your data on sites like Discogs and MusicBrainz to ensure that someone who is really trying to find you will have a good chance to do so.
But, keeping metadata up to date is not the whole picture either – there is also bad metadata and there are bad names! Bad names? Yes, in this world where we rely on search engines and increasingly on AI to find information, bad names can make artists impossible to find. Search engines normally throw out words that do not convey a lot of meaning in (English) texts, such prepositions and articles. (e.g. “the”, “a”, “on”).
Bad names then include some famous artists, such as “The The” and “A”. The same is true for naming recordings or releases – so we suggest that you avoid using simple words to name your recordings or releases. I would be very happy to never encounter another track named “A”. Similarly, you can imagine that fans of the band “!!!” may have had some difficulty finding information about the band.
Another case where bad metadata can easily be created is mixing languages or language scripts. Consider the third release of the Japanese band Morning Musume (モーニング娘。):
This release title mixes English text that uses the Latin alphabet with Japanese text that uses Hiragana/Katakana letters. There is a good chance that users will have a hard time finding this release if they search for it, because computers are poorly equipped to handle such mixed script/languages metadata.
We hope that you’ve had a chance to understand what we mean when we continue to talk about metadata. Our current modern world is awash in music with more being released every day. As an artist, if you release music without sufficient amounts of metadata attached to the release or discoverable in sites such as Discogs and MusicBrainz, few people will ever listen to it. Or worse, people will listen to your music, but know nothing about you or how to support you.
Stay tuned for our next article in which we’ll dive deeper into why metadata matters in our globally connected world.