# An example of why digital platform metadata should not be treated as canonical
**URL:** https://community.metabrainz.org/t/an-example-of-why-digital-platform-metadata-should-not-be-treated-as-canonical/644927
**Category:** MusicBrainz
**Tags:** deezer, digital-media, applemusic, metadata, spotify, musicbrainz
**Created:** [July 11, 2023, 10:56am UTC](https://community.metabrainz.org/t/an-example-of-why-digital-platform-metadata-should-not-be-treated-as-canonical/644927 "2023-07-11T10:56:14Z")
**Posts on this page:** 11
**Page:** 1
### Author:  [@atj](https://community.metabrainz.org/u/atj)
#### Post date: [July 11, 2023, 10:56am UTC](https://community.metabrainz.org/t/an-example-of-why-digital-platform-metadata-should-not-be-treated-as-canonical/644927/1 "2023-07-11T10:56:15Z")
So we have 2 different artists who both at one point or another used “Adeline” as their performance name.
Adeline Michèle - a US based musician, composer and vocalist who now uses “Adi Oasis” as her primary artist name:
- [MusicBrainz](https://musicbrainz.org/artist/9aeab082-b567-477d-ad27-8475481bf150)
- [Discogs](https://www.discogs.com/artist/961867-Adeline-2)
Adeline Hedmar - a Swedish house music producer and vocalist based in London:
- [MusicBrainz](https://musicbrainz.org/artist/35e3036f-ddb1-4e66-8bbb-9f23b1c92d12)
- [Discogs](https://www.discogs.com/artist/2695516)
- [Resident Advisor](https://ra.co/dj/adeline)
An example release featuring Adeline Michèle from 2019:
**JKriv featuring Adeline – Yo Love**
- [Discogs](https://www.discogs.com/master/1602320-JKriv-Featuring-Adeline-Yo-Love)
- [Spotify](https://open.spotify.com/album/3zNi7uzTFoe0DPFa5DEy5t)
- [Apple Music](https://music.apple.com/gb/album/yo-love-feat-adeline-single/1475894480)
- [Deezer](https://www.deezer.com/en/album/109720832)
Notes on this release:
- Spotify credits the artist as “Adi Oasis”.
- Apple Music includes “feat. Adeline” in the title but only has an artist credit for JKriv.
- Deezer includes artist credits for both JKriv and Adeline, but the Adeline artist is a mix of various artists named Adeline (including Adeline Hedmar).
An example release from Adeline Hedmar from 2012:
**Adeline – Love Handles You**
- [Discogs](https://www.discogs.com/master/1652473-Adeline-Love-Handles-You)
- [Spotify](https://open.spotify.com/album/5UBGVRzymKFziMSJZiyf3X)
- [Apple Music](https://music.apple.com/gb/album/love-handles-you-single/1645156402)
- [Deezer](https://www.deezer.com/en/album/356607977)
Notes on this release:
- Apple Music credits it to Adi Oasis.
- Deezer credits it to the same Adeline as credited on “Yo Love” (as noted above).
I mentioned at the top of the post that Adeline Michele has changed her performance name to Adi Oasis. It appears that on at least Apple Music and Spotify, an artist can change the name of their ID on the platform, which changes the credits for all existing releases:
- [Spotify](https://open.spotify.com/artist/5RRfTrwXUGYiBB0DMV4hyh)
- [Apple Music](https://music.apple.com/gb/artist/adi-oasis/1316203858)
So now we have this strange situation where releases which were explicitly released under one artist name (and is still present on the cover art) are now listed under another artist name, e.g.
- [“Adeline - Adeline” on Apple Music](https://music.apple.com/gb/album/adeline/1612302115)
- [“Adeline - When I’m Alone” on Spotify](https://open.spotify.com/album/2pDTgwhHKhvVzx4G0TUMiF)
OK, so you may be wondering why am I telling you all this? Well it’s for a few reasons:
1. I frequently see “_but it is credited like this on streaming platform X_” as a justification for submitting edits that are clearly incorrect.
2. To demonstrate that the data that is submitted to these platforms is often wrong or not specific enough.
3. To highlight the different (and often significant) metadata limitations of the various platforms, and how they can compound, conflict and result in a completely incoherent mess.
I’m not suggesting that we should disregard **all** metadata presented by these platforms but the idea that it should be treated as canonical is frankly ludicrous. The goal of MusicBrainz is accuracy and unquestioningly replicating errors from other sources does not help to achieve this.
Unfortunately I can’t offer any real solutions to these issues. However I do think that we, as editors, need to be more open to the idea that digital platform metadata is unreliable and there is nothing inherently wrong with using our best judgement to fix errors, better reflect artist intent and workaround the limitations of the metadata available on the platforms being used as sources.
---
### Author:  [@Victini](https://community.metabrainz.org/u/Victini)
#### Post date: [July 11, 2023, 11:04am UTC](https://community.metabrainz.org/t/an-example-of-why-digital-platform-metadata-should-not-be-treated-as-canonical/644927/2 "2023-07-11T11:04:04Z")
I had a similar issue adding the Deemo soundtrack to my library. The song “[Moon Without the Stars](https://musicbrainz.org/recording/c7a0a732-caa2-4f23-a840-1ce4b40b7346)” is canonically credited to Jerry Barnes feat. Quiana, but on streaming/iTunes it’s a combined artist for whatever reason and it’s been added to MusicBrainz with a singular artist entity “Jerry Barnes Quiana”.
---
### Author:  [@atj](https://community.metabrainz.org/u/atj)
#### Post date: [July 11, 2023, 11:41am UTC](https://community.metabrainz.org/t/an-example-of-why-digital-platform-metadata-should-not-be-treated-as-canonical/644927/3 "2023-07-11T11:41:10Z")
Thanks, that looks like a another example of incorrect metadata being submitted to the digital platforms. Funnily enough I don’t think the MusicBrainz artist being credited is correct and should be a separate artist:
[Quiana (2)](https://www.discogs.com/artist/1786848) vs. [Quiana (4)](https://www.discogs.com/artist/12241210-Quiana-4)
Edit: I’ve gone ahead and fixed this issue - see [open edits](https://musicbrainz.org/artist/2b9b4e12-798f-4b52-b8d7-0304a8f46261/open_edits).
---
### Author:  [@elomatreb](https://community.metabrainz.org/u/elomatreb)
#### Post date: [July 11, 2023, 2:58pm UTC](https://community.metabrainz.org/t/an-example-of-why-digital-platform-metadata-should-not-be-treated-as-canonical/644927/4 "2023-07-11T14:58:29Z")
Personally I treat the metadata provided by the platforms as secondary to the cover art, where I know that the artist and/or label signed off on it and it doesn’t change out under me (usually).
Streaming platforms (especially Spotify) also create entire bogus single releases IMO (see [other thread](https://community.metabrainz.org/t/digital-releases-how-to-deal-with-single-track-compilation-excerpts-on-streaming-platforms/623302)).
---
### Author:  [@atj](https://community.metabrainz.org/u/atj)
#### Post date: [July 11, 2023, 3:05pm UTC](https://community.metabrainz.org/t/an-example-of-why-digital-platform-metadata-should-not-be-treated-as-canonical/644927/5 "2023-07-11T15:05:34Z")
I agree cover art is a good and reasonably canonical source of information, although it can be difficult at times to determine what is artist intent versus just design / typography.
However I will point out that I’ve seen several instances where the artwork for a release has changed months / years after the original release date…
---
### Author:  [@IvanDobsky](https://community.metabrainz.org/u/IvanDobsky)
#### Post date: [July 11, 2023, 5:26pm UTC](https://community.metabrainz.org/t/an-example-of-why-digital-platform-metadata-should-not-be-treated-as-canonical/644927/6 "2023-07-11T17:26:25Z")
> [@atj](#):
>
> the idea that it should be treated as canonical is frankly ludicrous. The goal of MusicBrainz is accuracy
TOTALLY agree. Digital platforms are shops. They do what they do to sell items. They will always sell using any data they have to hand that makes a sale. This should never make them a perfect reference source.
---
### Author:  [@reosarevok](https://community.metabrainz.org/u/reosarevok)
#### Post date: [July 12, 2023, 10:33am UTC](https://community.metabrainz.org/t/an-example-of-why-digital-platform-metadata-should-not-be-treated-as-canonical/644927/7 "2023-07-12T10:33:30Z")
That’s without even counting issues where music that is clearly not related to a relatively famous artist has ended somehow in Spotify as credited to them and then in people’s new release queues (IIRC my release queue kept giving me fake “Run the Jewels” music at some point).
---
### Author:  [@chaban](https://community.metabrainz.org/u/chaban)
#### Post date: [July 12, 2023, 8:35pm UTC](https://community.metabrainz.org/t/an-example-of-why-digital-platform-metadata-should-not-be-treated-as-canonical/644927/8 "2023-07-12T20:35:45Z")
`https://musicbrainz.org/release/a83aee98-8ac6-4d08-9d26-a4b69c25ea58` lists “We Came As Romans & Brand Of Sacrifice” as secondary artists on digital service providers which doesn’t seem quite correct to my interpretation therefore I’ve used the music video as reference:
- [https://www.youtube.com/watch?v=\_VRrk\_\_m0iA](https://www.youtube.com/watch?v=_VRrk__m0iA)
> Harper - Weight Of The World featuring Dave Stephens from We Came As Romans and music Written by Brand Of Sacrifice.
- [Harper releases new music video for 'Weight Of The World' - Distorted Sound Magazine](https://distortedsoundmag.com/harper-releases-new-music-video-for-weight-of-the-world/)
> The new single was written by Leo Valeri, guitarist and producer of deathcore heavyweights BRAND OF SACRIFICE and features the guest vocal talents of Dave Stephens from WE CAME AS ROMANS.
- [https://twitter.com/EricDrkngl/status/1667027201227894786](https://twitter.com/EricDrkngl/status/1667027201227894786)
“Screenshot of SiriusXM showing Harper/Dave Stephens in the artist field”
* * *
[Another example](https://musicbrainz.org/release/8de625ad-c9f5-4668-a05c-1f34d28c9425) with a cover performance When it was released it also listed Ryuichi Sakamoto as secondary artist. He is written on the cover art too but since it’s a cover performance I’ve dropped him and only linked the work.
This [seems](https://musicbrainz.org/edit/100883038) a common occurrence.
* * *
I no longer trust cover art too much for artist credits. [Sometimes](https://musicbrainz.org/edit/67682695) the normalized name is intended.
* * *
> [@\ as \](https://community.metabrainz.org/t/real-band-as-fictional-band/638796/16):
>
> I don’t remember whether only the fictional names were shown in-game, if at all. There were a couple [radio stations](https://cyberpunk.fandom.com/wiki/Cyberpunk_2077_Radio_Stations) to [choose from when driving](https://static0.gamerantimages.com/wordpress/wp-content/uploads/2021/01/cyberpunk-radio.jpg) (there were also [some](https://i.redd.it/jwy4fl98599a1.png) scattered around throughout the game word). The [end credits](https://youtu.be/UgKVqwZtfwQ) had more details on performers, etc. and as can be seen some [didn’t have a fictional name](https://cyberpunk.fandom.com/wiki/91.9_Royal_Blue_Radio) or had [multiple fictional names](https://cyberpunk.fandom.com/wiki/103.5_Radio_PEBKAC). If I had to guess, both the fictional and real artists are credited on music platforms due to statistics, royalties or something like that. Maybe it…
* * *
Crediting the game as artist doesn’t seem proper either when the writers/performers are known:
[https://musicbrainz.org/release/236f6b12-f512-41d6-9474-5c3352edc11b](https://musicbrainz.org/release/236f6b12-f512-41d6-9474-5c3352edc11b)
And if no artist is on the front cover it should use Various Artists per [soundtrack guideline](https://musicbrainz.org/doc/Style/Specific_types_of_releases/Soundtrack#Release).
Another pattern is the label as artist on soundtracks like notorious [EA](https://open.spotify.com/artist/1gsT5GesFZ2qkDkgPxMUA1). Fortunately that mess has been cleaned up not too long ago by multiple editors.
---
### Author:  [@atj](https://community.metabrainz.org/u/atj)
#### Post date: [July 13, 2023, 8:48am UTC](https://community.metabrainz.org/t/an-example-of-why-digital-platform-metadata-should-not-be-treated-as-canonical/644927/9 "2023-07-13T08:48:30Z")
Thanks for providing some great examples @chaban, I figured you’d probably have an extensive list of weird and wonderful edge cases!
The vast majority of these issues seem to stem from one of two things:
1. Simple human error on data entry either by the label, artist or digital distributor.
2. Attempts to work around the limitations of the overly simplistic metadata schema used by the various platforms.
---
### Author:  [@reosarevok](https://community.metabrainz.org/u/reosarevok)
#### Post date: [July 13, 2023, 9:23am UTC](https://community.metabrainz.org/t/an-example-of-why-digital-platform-metadata-should-not-be-treated-as-canonical/644927/10 "2023-07-13T09:23:31Z")
> [@chaban](#):
>
> [Another example](https://musicbrainz.org/release/8de625ad-c9f5-4668-a05c-1f34d28c9425) with a cover performance When it was released it also listed Ryuichi Sakamoto as secondary artist. He is written on the cover art too but since it’s a cover performance I’ve dropped him and only linked the work.
That’s a classical guitarist, I expect it should be treated as a classical release 🙂
---
### Author:  [@chaban](https://community.metabrainz.org/u/chaban)
#### Post date: [July 24, 2023, 2:05pm UTC](https://community.metabrainz.org/t/an-example-of-why-digital-platform-metadata-should-not-be-treated-as-canonical/644927/11 "2023-07-24T14:05:02Z")
Sometimes group members are “credited” alongside the band:
- [Edit #98869982 - MusicBrainz](https://musicbrainz.org/edit/98869982)
Or the “reverse” where a (solo) album additionally lists the associated group/label/crew:
> **[Edit #101881571 - MusicBrainz](https://musicbrainz.org/edit/101881571)**
Sometimes I’m not sure what to make of it:
> **[Saftboys](https://musicbrainz.org/artist/70bf0f8c-f3ca-47ac-b26b-dd42c48f514b)**
>
> Group, Germany
* * *
Since [atj asked](https://community.metabrainz.org/t/a-multi-source-seeder-for-digital-releases/444000/173) about examples in the past. Have a list of minor or “well-known” examples from my bookmarks:
- [Bandcamp dates](https://community.metabrainz.org/t/bandcamp-release-date-discrepancies/401016)
- [Bandcamp dates (future)](https://community.metabrainz.org/t/label-using-fictitious-release-dates/581756)
- [ISRC formatting](https://community.metabrainz.org/t/fetching-isrcs-from-spotify/20910/79)
- [Original dates shown](https://community.metabrainz.org/t/spotify2musicbrainz-py-pre-alpha-testing-release/250504/12)
- [Deezer web player shows different lengths](https://community.metabrainz.org/t/a-multi-source-seeder-for-digital-releases/444000/172)
- [iTunes gives “Instrumetal” as label and “℗ 2021 Instrumetal”
artist is also given as "Instrumental](https://musicbrainz.org/edit/89873523) (unfortunately not archived)
Lately I’ve noticed for releases I add that Deezer often lists an earlier release date. E.g.
- [Hypercube Records](https://musicbrainz.org/label/32b7c28f-3d90-479b-b03b-a1281c42cb4f)
- [https://atisket.pulsewidth.org.uk/?cached=197902420378-d\_458177865-s\_37ftrq8MYmXiRrznhBOXF9-i\_1694431407](https://atisket.pulsewidth.org.uk/?cached=197902420378-d_458177865-s_37ftrq8MYmXiRrznhBOXF9-i_1694431407)
- [Blaue Augen by T69 | a‐tisket: A multi‐source seeder for MusicBrainz](https://atisket.pulsewidth.org.uk/?cached=197902858577-d_455337655-s_7x5Te8PIyJw0MBVK3LVo30-i_1693369063)
- [https://atisket.pulsewidth.org.uk/?cached=197901258989-d\_459598835-s\_4afi02oZMLnjo5tGctWtdC-i\_0](https://atisket.pulsewidth.org.uk/?cached=197901258989-d_459598835-s_4afi02oZMLnjo5tGctWtdC-i_0)
In this case however it was a later date:
- [https://atisket.pulsewidth.org.uk/?spf\_id=6TGSIc0WrZ8FiKyBSY4FKd&upc=197745927430&preferred\_vendor=spf](https://atisket.pulsewidth.org.uk/?spf_id=6TGSIc0WrZ8FiKyBSY4FKd&upc=197745927430&preferred_vendor=spf)
- [Hypercube, Chapter 1 by Hypercube Records - DistroKid](https://distrokid.com/hyperfollow/hypercuberecords/hypercube-chapter-1)
- [https://twitter.com/HyperCubeLabel/status/1649286202061918208](https://twitter.com/HyperCubeLabel/status/1649286202061918208)
- [https://twitter.com/HyperCubeLabel/status/1644770636584357888](https://twitter.com/HyperCubeLabel/status/1644770636584357888)
- [https://twitter.com/HyperCubeLabel/status/1646940541370920960](https://twitter.com/HyperCubeLabel/status/1646940541370920960)