Thanks for the clarification. I’ve changed the “yes” bullet point to:
Different cover art. Any differences beyond size/quality changes and minor color differences justify multiple releases.
and the “no” bullet point to:
Differences in cover art size/quality or minor color differences. Some online services impose cover art requirements or reprocess a release’s original cover art.
Does that sound okay to you?
Yes, thank you!
I still think you can remove the ‘no’ bullet point - the shorter all this text the more likely someone is to read it/take it in - but it’s not a deal breaker for me.
Best sources for dates if you can match barcodes is either Mora.jp or ISRCsearch.net. Not always correct, but a good source. Even they sometimes have original dates, but seems most often have the actual release date.
Speaking of cover art.
Every now and then I see photos of physical releases being presented as “the” cover on Bandcamp. It doesn’t feel to me these are meant (as in artist intent) to represent the digital release.
I suspect they are put there for other reasons, e.g. it was supposed to be a physical release so just a photo (or scan) was created/added and the artist didn’t care enough to make a “proper” one for the digital version.
Recently scans of a physical release (apparently copy pasted by the artist from Discogs) were added to a digital release arguing they were part of the download and should therefore be added.
This seems very wrong to me as now two different releases have almost the same set of images. This can also get in the way of future clean-up efforts of e.g. incorrectly set cover art types and could give the impression of bad quality control. (there are likely far more users - who won’t see edit notes, etc - than editors)
A digital release can’t have a physical medium
For the above case it seems more reasonable to add an annotation the download came with scans of an older physical release.
Similar case where the digipak was cropped and added as “the” cover on Bandcamp but other platforms a proper one is used.
Old case where an erroneous cover was fixed shortly after. The old version has a “misprint” comment.
Relatedly netlabel releases (and apparently bootlegs too) will sometimes include images resembling a physical release (often jewel case CD). If no physical release counterpart exists uploading those feels less wrong but still the cover art type problem remains.
Another thing I find problematic are placeholders that don’t seem representative for the release in question. Or generic images that are only tangentially related as the actual download has no image at all.
Maybe I’d add to the guidelines that linking spotify and/or youtube music for digital releases is encouraged since both have a free tier available and both are supported by BrainzPlayer and ListenBrainz app/website.
Does this include making an unique release for no barcode set on platforms that allow for this (ex. Bandcamp)? That’s how I’ve been interpreting the guidelines for unique releases, anyways.
I personally wouldn’t block people from adding images from a download - if the artist intent is to share crap images or photos of something with a download, that’s unfortunate, but their prerogative.
I’m basically fine with anything that came with an actual release, or denoted the digital release on its store/download/streaming page, anything else shouldn’t be there.
Whether they should be tagged with a physical type tag… that’s less clear to me. Quite often ‘print your own’ images are included, for instance to print and make a Jewel Case release, and I reflexively would set ‘medium’ and ‘back + spine’ on those. Quite often a back cover is included. I guess I personally don’t mind those tags being there.
Hmm. I suspect that there are cases where it’s impossible to tell whether the omission of a barcode on one platform (while present on other platforms) is deliberate or accidental. The Bandcamp userscript doesn’t seed the barcode field unless the download is the album’s only medium:
// UPCs generally apply to physical releases so set the barcode when
// digital download is the only available medium
let mediums = bandcampAlbumData.packages;
let upc = bandcampAlbumData.current.upc;
if ((mediums === null || mediums.length === 0) && upc !== null) {
release.barcode = upc;
}
If I’m remembering correctly, Bandcamp often (usually?) doesn’t even display the barcode, choosing to bury it inside a JS object instead.
I personally wouldn’t bother adding a release for no-barcode cases, but I probably also wouldn’t downvote someone’s edits if they did so.
From the guideline proposal itself:
This makes sense to me, but we should probably also specify more clearly then that intentional separate releases should be kept separate. I get that most often they involve different labels, but this might not be obvious since we only have (P) and (C) credits in several stores. Sometimes this also involves different track titles (Latin vs non-Latin) which is also different releases, but might be worth clarifying?
Does this mean different main labels, or also different distributors and whatnot?
I generally agree with this idea, but we should probably also specify that in cases where some countries are explicitly separated (such as if there’s a release in one region and one rest of the world release), the countries should be listed.
Other comments:
I think I agree with this view.
Agreed - I wouldn’t be surprised if some albums were released in two versions, cheaper without a PDF and more expensive with one, tbh. But in any case, I’d certainly call a physical release different if it came with vs without a booklet and I’m not convinced this is not the case for digital, so I’d suggest skipping this entirely for now and seeing what common sense leads us to.
The release date guidance should be relatively easy: “The date the release was added to the first of the digital platforms it’s available in”. You could specify that often the date is not given or is replaced by the release date of a physical version of the release, so in case of doubt it’s better to not enter a date at all unless there’s certainty it’s the date for this specific digital release.
One thing I’m missing here is evolving releases (I’d want to specify that a release that has tracks added to it should be entered as a new, second release rather than edited).
Any suggestions for how to word this? Given the (sometimes breathtakingly low) quality of the data provided by the services, I’m having trouble thinking of any hard-and-fast rules. I’m wondering if it’d be best to just let editors use their best judgment here.
Hmm, I’m not sure that there’s typically enough information provided by the platforms to make these distinctions. Are there any easy-to-find examples of online releases that otherwise seem identical but were released by different labels?
I added the following: “If some countries are explicitly separated (e.g. a release in one region and a separate rest-of-world release), then the countries should be explicitly listed.”
Thanks! I added a brief “Release Dates” section based on the text you gave; please let me know if it aligns with your thinking.
Does this sentence (added to the end of the “Different song ordering” bullet point) seem clear? “If additional tracks are added to an already-released album, a new release should be created with the new tracklist.” I’m hopeful that “already-released album” will be enough to keep people from creating separate releases for pre-order tracks that are available on Bandcamp.
Pink Floyd. ( Ask @tigerman325 ). There are different distributors for Europe vs USA so everything is entered twice. https://musicbrainz.org/release-group/38b4c840-16ce-40fa-b7c0-a354b7c87178 I think the barcode usually changes as well, but the Tiger will know for certain.
I’ve been assuming that different relationships (distributor, phonographic copyright, etc.) means a different release, so I’d say yes to different distributors.
I know you can often tell when an artist used DistroKid, for example on this release, or DistroKid also has some special features which an artist might link on their socials, I think it’s called “SuperFollow” or something? this isn’t the case for all DistroKid distributed releases, but many of them. it is also the only example I know to look for
Yes, if a Bandcamp release has no barcode, then it’s not the same as one with a barcode. Many times it has the same barcode as those on Spotify or Deezer. There is a script for Bandcamp that easily finds the barcode and displays it on the page for you. The only place I’ve seen that it’s impossible to know a barcode for is Amazon. I usually just double check the labels, etc and if nothing else is different I link it to the Deezer or Spotify release.
They have different barcodes on the Sony vs. Warner distributed releases. Sometimes slightly different artwork even. Pink Floyd has some of the most complex digital releases as they sometimes even have the same barcode, but different track lengths on the releases due to remasters, etc.
I disagree with this dogmatic approach.
Releases such as https://mariadieruhe.bandcamp.com/track/rejection are clearly meant to be the same across other platforms
As I mentioned above, a difference in barcode is typically for a reason. Just because Amazon doesn’t expose barcodes doesn’t mean that we can’t reasonably decide that the Amazon release is one of these and not the other.
I’d never use a link page to determine if it’s the same release. Many times the Apple Music release linked on them are different than the Spotify release linked on them. If NO Bandcamp releases had barcodes that were known, i.e. Amazon, maybe, than I can see this. But many Bandcamp releases have barcodes and are easily identifiable. So, I see no harm in adding a Bandcamp release as a different release, especially if the Apple Music & Spotify releases have different barcodes. I think if someone wants to add a separate release for a Bandcamp release without a barcode that’s ok. However, I agree, I’d probably not add a different release if the barcode was the same on all other platforms.
Spotify/Deezer/Apple Music hide it in their interface as well, so I don’t see why it’s different. From this help article, it seems like they use the UPC to report sales for both digital and physical releases, so it should be seeded all the time I think?
The database schema, and the intentional inability to manually define recording original dates and/or track release dates, requires this, but it’d be good to answer the question this leads to:
Should the pre-evolution release be marked Withdrawn? Does it matter whether it was a streaming-only release that people couldn’t officially download a copy of? For example, if a streaming-only release with a certain level of DRM (e.g., a proprietary app is required) is no longer available because it was replaced with a different tracklist, a reasonable person could conclude it should go from Official to Withdrawn.
I’ve seen such extreme examples where similarities can even include having the same barcode (digit-by-digit) yet due to differences in labels (distributors, rightholders, licenced from/to, etc.) we have multiple instances of otherwise the same release on one same digital platform, each with their own URL and geographic restrictions.
Take this single released by the Swedish artist Fever Ray for example:
All three seem the same, even having a barcode in common (5060236638540), if you enter the editing history, you can see that each one of them was added or edited using information from a-tisket and similar companions. From the RG view, the first difference one can spot are the different release areas (the first one is for the artist’s home country of Sweden, the other released for US+CA and the third one has all the rest of the world covered). The US+CA one even has a different ‘main’ label set from the other two (Mute vs. Rabid Records).
Entering each of the releases, you can see some of the subtler differences in label info (distributors, licences) and make sure that each one has it’s own distinct release ID on the 3 big vendors (Spotify, Apple, Deezer). If it wasn’t for the different labels and distributors, I could have easily merged the 3 into one single worldwide release.
There’s a similar case with Icelandic singer Björk’s digital releases (at least I kept an eye on her most recent album “Fossora” and the accompanying singles). Unlike with Fever Ray, all of these releases are credited to One Little Independent Records with no further info, which caused most editors to assume they are the same and merge them after they were being added separately.
You can see this one having two different Apple Music links, each one with it’s own ID number. Yet both of them have the same Barcode, the API returns no additional label info other than One Little Independent, they only differ by the countries where they are available (US+CA vs. the rest of the world).
Similar case with her singles. A-tisket treated them as distinct releases, yet on MB we have just one instance of it with all links covered.
https://atisket.pulsewidth.org.uk/?cached=5016958102098-d_0-s_4gsVIuKRUMJCdNhHTx5GrI-i_0
https://atisket.pulsewidth.org.uk/?cached=5016958102098-d_347331777-s_5OXgQKo4eciKADPZPlkdF1-i_0