Hi there! I’m fairly new here, but have been using Picard for a while now to organize my digital collection. Occasionally I’ll come across a release that’s not in the database yet, and I wanted to double-check the etiquette around adding new releases.
I think I’m in a fairly common situation where most of my digital files come from solidly tagged, but “unofficial” sources such as soulseek. The days of dodgy limewire mp3s are long gone, and most of the stuff you find nowadays is in good shape. I’m not interested in bootlegs or otherwise unofficial releases, so when I find myself in this situation, the release I’m thinking of adding almost always has a Discogs entry and is usually even available on Spotify.
But, since I don’t know the original source, I can’t claim to know exact details.
From scanning through the forums, and given the existence of Discogs-importer userscripts, I feel like this is OK, especially for missing releases. My interpretation of the vibes is that you’d rather have the release (given that other sources confirm it actually exists) than wait for someone with perfect information (a.k.a they actually bought the cd).
IMO it’s fine to enter data based on Discogs or other listings like that, trying to double-check it with other sources if possible. Though from the official release is ideal.
If there’s information that only came from an unofficial copy I’d suggest to be cautious and try to find corroboration elsewhere for it, since CD rips don’t always retain significant distinctions like barcode numbers that would determine which Discogs release you’re looking at. Checking that your sources match is important since otherwise it’s easy to end up making Frankenstein releases, like entering a barcode for a CD with no bonus track on a track listing with a bonus track.
If the metadata seems good and there’s a Composer in the ID3 tags, for instance, I might look up the release on iTunes store and see if they have the same composer listed before entering it. If you can’t find a corroborating source, I’d definitely be sure to mention in the edit note and maybe also the annotation that you’re not 100% sure about the data.
One problem with Soulseek \ Torrents is not being able to be certain if the artwork actually matches the release. I have had torrents that look good, but on close inspection I realised the artwork has a different track count to the rip. Or a country is changed. Or even weirder when you spot that the artwork has been “censored” to block out the country.
Trusting Discogs as a source is more reliable than those dubious files. I usually don’t submit AcoustIDs or DiscIDs from stuff sourced from a torrent. Especially a common album that will have different editions worldwide.
Sometimes I have had albums with such dubious history that you notice the tags in the files makes it clear that it was manually assembelled from tracks from multiple CDs.
if you have a release, and then bring over the Discogs details about it to Musicbrainz then this is a trustable source. And once you have the release in the MB database you can now tag your files. This is ideal as you have improved the database and your tags at the same time.
Partial information is better than no information.
Most ‘hardcore’ editors here edit/add thousands of releases that they don’t have!
But you are correct that care should be taken. There’s really no reason not to trust a Discogs release with scans/photos (or any source with scans). Even without scans, MB operates on a ‘information can always be improved’, so as long as you’re not adding really derped info, someone with the in-hand cd can always check it later.
Digital is even easier, since you and everyone else (usually) can see the ‘source’ directly, in their browser. Plug the link into harmony.pulsewidth.org.uk and you’re off!
Just go slow at first, so that you don’t have to go back and fix a ton on things, if you get feedback on your edits.
welp, this was something I needed to read. A lot of my stuff is sourced from these places and I have been submitting fingerprints with everything I pipe through Picard. { - _ - } I’m sorry if I’ve caused extra work for those of you with the experience to know better. I spend enough time here to know that y’all are also the ones running around fixing all the wonky releases.
I recognize this is silly, but could this not be considered a bootleg release? Just a thought, not sure if that’d even be useful, but maybe if it’s something that gets lots of circulation. I’m struggling to see much difference between something cobbled together by a community member (or members) vs something cobbled together by a bootlegger.
Don’t take my word for it, but I imagine bootlegs operate on a bit of a sliding scale.
Think about me just throwing some files into a folder and calling it an album. Clearly, this is not something particularly interesting to the rest of the world.
However, imagine the same thing happening in the 60s with some bootleg Beatles recordings at small gigs. Sure, it’s essentially the same thing, but due to time and interest, something like that will have gained way more significance than my botchjob “bootleg”, and is now worth capturing in something like MB.
To me it feels a bit similar to the same reason your neighbor probably doesn’t have their own Wikipedia page, and if you’d attempt to make one for them, they’d reject it. There’s a sort of soft cutoff where information is just too niche to be relevant for a global database such as this.
So yeah, maybe if there was some legendary compilation torrent that gained some cultural significance, but I’d imagine in that case you probably know ahead of time what you’re getting into.
I feel like maybe I didn’t make myself clear enough about why I found the idea interesting. I wasn’t so much saying it’d be the same as a bootleg in terms of interest or even really in definition, though I’d argue it’s not far from it practically speaking. I was referring to using it as a category of release for MusicBrainz. That way the metadata is there for folks to catalog it for their personal libraries without infecting official releases by submitting a bundle of AcousticID fingerprints from recordings taken from multiple different releases, for example.
Lots of people get their music from these sources. I’d wager the Venn diagram of people who also use Picard to organize their library, but aren’t in here to be warned not to bork the database with Frankenstein’s Black Album is close to looking like a single circle.
I’m going to refrain from hashing this out any further in this thread though since it’s pretty off topic from OP’s post. I may make a post about it later.
I use ssk 100% for my collection. The problem is that even if release is tagged on MB/discogs, and some are, you can’t know that info is correct as I find most people care only about track list and will take whatever release is having it the same.
My method:
If it has ISRCs it probably originates from streaming service.
If it has cd rip log, and many have, the probability is higher it is a cd format and I look into label/catalog number if it has it in tags (often do).
If it has very low amount of tags, like, only label and catalog number, i trust it more that it is that particular release as it was definitely manually added.
If its mp3, I skip it. FLAC users have higher standards regarding entire process, be it tagging or quality or whatever.
I download several different looking releases (file sizes, other folder data etc) and compare.
In any case, I add all releases I find in MB release group, not just the one I have at hand (I use Platform Check userscript for this) - original date version, remastered, all formats. This is pretty fast with good userscript toolset. Its valuable info for all, and if I was mistaken about my particular release, I can be sure its probably some other one in RG. I don’t listen commercial music so this is easier as some of those can have a huge number of releases in RG. For those, I just pick according to the best info I have. Regarding covers, Discogs definitely have them best, but not above 600x600. I replace some of those if I find exactly looking one around in higher res (almost always only possible with front cover).
It is a case of trust. If you feel that is really the real album you are linked to, then submit. But if you are picking one out of a dozen different variants and not sure if you have a US or UK edition, then be more fussy.
Most important part is to check track lengths. If the lengths are off from the real album, then it is likely made up from different recordings to what you think.
If in doubt, don’t submit.
No, I just call it a mess and don’t submit it in any way. Sometimes people think they can build an “album” by pulling tracks from different places to match the original album tracklist. But from many years on MB I realise that not all albums use the same tracks as each other.
I am thinking especially of “soundtracks” or where someone is trying to make something look like a real album. I’ll tag it in some way, but certainly won’t submit any audio data from it.
Discogs is quite trustworthy and I use it often, with some care.
Be aware, that sometimes there is the wrong artist (of those with the same name) selected on compilations. Since Discogs has no track/recording based editing, such mistakes are not as easily caught there as on Musicbrainz in the list of recordings of an artist.
I wouldn’t enter any release with hidden tracks, if I don’t own exactly that release, because these are hard to verify with cover scans alone.
I would work it out on a quality idea. If it feels scrappy, I don’t trust it. Not worth adding.
If it is a real actual interesting new combination of bootleg hard to find tracks that is passing around due to it being hard to find tracks, then I’d consider a bootleg release. We have CDs like this, but the CD is something of actual effort needed to create.
If the album is just someone’s playlist \ mixtape \ greatest hits of the 80s then I would not add it. If it is just a selection of easy to find tracks in a different order, I would not add it. This is more like getting a copy of your mate’s mix tape and that has guidelines about not adding those.
I agree with that as a quality check. Personally I have dumped most of my MP3 downloads of the past as my audio system cries when playing 128Kbps MP3s. Would never add any data to MB from a MP3 rip unless we are back to that really rare bootleg. If I know this 128Kbps rip is of a really unusual bootleg of hard to find tracks then I will document it.
Honestly, I hadn’t even given it much thought which was stupid. I really didn’t even consider Picard as a potential avenue of bad data in that direction. The AcousticID Fingerprints are the only thing that can be submitted from Picard right? At least without some plugin, but none have been interesting enough for me to install apparently because I think fingerprints are all I can submit. I kinda figured that there’d be some safety mechanism against them being linked to the wrong tracks. They are unique so if it doesn’t fit and there’s already one there, what happens?
I do have more to say on this, but like I said in my last reply I don’t wanna take away from the post with something off topic.
Bad AcoustID data is a common one… but if you are submitting it from an album that at least has the correct length tracks then you are probably not causing too much havoc.
The trouble comes when you have a USA edition of a UK release that changed a few mixes on the release. Or where someone swapped a 12" version of a track on to SSK version instead of the album version. Now we can cause hiccups in the AcoustID data.
But to be fair… the odd entry ain’t too bad. I also used to do it, but the more I learnt about AcoustID the fussier I got as to what I submit. The biggest damage to AcoustID is from those people who throw 10,000 tracks at Picard. I don’t think this is what you are doing if you are identifying an album first.
Start a new thread and link it here then the OT Police can’t tell you off.
nvm that pretty much sums it up. To be clear, I wasn’t saying to drop some random monstrosity into the database and call it a new release. But essentially what your describing here with the wide spread compilation bootleg.
I think it is summed up with “use a bit of common sense”.
Soulseek \ Torrents are a way to share things - but not an actual release channel. If something being shared is ripped from something real that exists, then it is a thing that can be added to MB. If something being shared is Bob’s favourite tracks, then it don’t qualify.
I remember the real early years of eDonkey when people would accidentally share their whole documents folder… There used to be some amazing piles of low quality junk on that system…
I agree with that as a quality check. Personally I have dumped most of my MP3 downloads of the past as my audio system cries when playing 128Kbps MP3s. Would never add any data to MB from a MP3 rip unless we are back to that really rare bootleg. If I know this 128Kbps rip is of a really unusual bootleg of hard to find tracks then I will document it.
There used to be some amazing piles of low quality junk on that system…