My private unreleased stuff & MusicBrainz

Tags: #<Tag:0x00007f05251c3d20> #<Tag:0x00007f05251b1b20> #<Tag:0x00007f05251a85c0> #<Tag:0x00007f05251a23a0>


I understand that MusicBrainz is only for released music recordings. Unreleased and bootleg albums reiterates that. But I have this private, unreleased stuff which I would like to have MusicBrainz help me with. Is there a way?

One example is my custom retracking of a Release. A DVD concert recording was released, and I added it to MusicBrainz. But the people who mastered it, bless their hearts, put the DVD chapter boundaries at curious locations, often unrelated to the movement divisions of the work. So I fired up ffmpeg and split the unitary soundtrack of the whole DVD into my own per-movement audio files. I would love to re-use the Release-level titles, artists, and relationships. I can re-use most of the Track-level artists and relationships, but I tweak the Track Titles and times.

What I want is to be able to tell Picard, "tag my per-movement audio files by taking my list of Tracks and Relationships, and falling back to Release cbc2c for everything else.

A second example is privately-shared audio files. An composer gave me some personal recordings of performances of his works, to introduce his work to me. I can be sure that there is an Artist entry and Work entries. But I don’t think these particular recordings are public enough to justify adding them to MusicBrainz.

What I want is to be able to tell Picard, “tag my private audio files as non-album tracks, by using my per-track data and Relationships, and falling back to Artist and Work entries from MusicBrainz”.

A third example is an Artist’s promotional recordings. I downloaded them as audio files from the Samples section of the website. The Artist didn’t clearly say, please download these and keep them, but also didn’t do anything to prevent that. Since then, the Artist remodeled their website, and the Samples section is gone.

What I want is to be able to tell Picard, “tag my promotional audio files as non-album tracks, by using my per-track data and Relationships, and falling back to Artist and Work entries from MusicBrainz”.

Here I may actually have a leg to stand on. Can I justify adding these audio files to MusicBrainz, because the Artist did at one time present them in a way that was almost publicly releasing them?

If I could author some metadata into an XML file, which could reference MusicBrainz via MBIDs; and if Picard could read this XML file, then that might give me what I want.

If I could make entries in to MusicBrainz but mark them “private”, that might give me what I want. It would keep private data out of MusicBrainz’s public database, but would still increase MusicBrainz’s size and demands.

If I could make a private MusicBrainz server instance, which stored my private entries but passed on all unresolved MBIDs to the main MusicBrainz, that might give me what I want.

Any idea how to accomplish this? Are there feature requests already written for this, for I should be voting?

Thank you!


I don’t have any suggestions for the rest of your post (at least not at 2AM), but…

That strikes me as clearly a release. It’s on the artist’s public web site—how would that not be released? Probably of the “promotion” type, but I don’t know where the definition of that is. (This presumes the samples were organized sort of like an album)

If not, Standalone Recording even mentions Internet-only releases. (And even recordings off the radio.)


Your first two cases, IMO, don’t belong in MusicBrainz, as they are not useful to the public at-large.

If you absolutely need to have them, I think setting up your own MB server and entering them there could do the trick. Whether that’s worth the effort, I don’t know.


Oh, I agree. But nevertheless I am looking for a way to handle them, and the publicly-available MusicBrainz is part of the answer.

The simple-minded approach of setting up my own MB server, then storing my own data there, amounts to a fork. I lose the benefit of later updates to the public MB data. Do you know of anyone who has set up a private MB server, in a way that stores only the private data, and forwards requests for other data to the public MB server? That’s the main question.


I have not tried it, but setting up a „normal“ slave server that gets replication data from MB-central, and then turning on editing could work.

The MBIDs of entities should not collide, but there may be other factors preventing this.

Unreleased and bootleg albums

The main objection I have with adding unreleased or otherwise not-intended-for-public-consumption content is a lack of checks and balances. If only one person or a select few have access to a given release, no one but them can verify any information about it. For all we know, the data on said release could be inaccurate or (worst case scenario) completely falsified, and no one but those with access would know. Additionally, as you are likely already aware, often times content like this is distributed on conditions of secrecy, meaning those who are granted access are forbidden from disseminating information about said content (or severely limited in what they can reveal). If we were to allow users to add data on unreleased or privately-disseminated content, it would not only undermine our internal checks-and-balances system, it would deal a major blow to the system of trust much of the music industry relies on to prevent leaks and potentially cause serious repercussions in the amount of content produced and officially released.