A few people commenting that some of their collection "doesn't exist in any DB", the best way to fix it is to add it to Musicbrainz[0] yourself!
I have found that adding things to Musicbrainz is actually pretty easy (and if you are so inclined like me, pretty rewarding and fun).
Streaming releases (and Bandcamp) you simply drop the release URL into Harmony[1] and it does most of the work for you.
Musicbrainz can represent nearly everything musically related and its all freely licensed, a very cool thing to exist.
Most (non-destructive) edits are auto-applied, whilst the rest go through a 7 day voting period (they are still applied by default unless someone votes against). The barrier to entry is very low.
You're still assuming that all of a person's music metadata _belongs_ in a global database. Some of my collection includes:
-"frankenstein" musical soundtracks where I've assembled my favorite version of each song from all the various published recordings. Sometimes I've even edited pieces of different recordings together into a hybrid track.
-a soundboard recording of my friend's high school talent show performance
-Music I've personally recorded from video games or other random sources
-Songs where I've edited out parts I don't like (such as overly-long drum solos) leaving just the parts I do.
I've organized these things in a way that makes sense for me, not for the world.
This is like a bookshelf that can only store books with an ISBN that can be classified in the Dewey decimal system. Too bad about your family photo scrapbooks or your personal sketchbooks!
What I think they're doing (and what I've done) is add music that wasn't already in the global database to it. For example, a promotional CD, or releases from a small publisher, etc.
In these cases, you're not adding personal metadata or mixtapes.
That said, definitely do self organization if it works for you. Most of your examples seem like they're suitable for custom playlists.
I have similar reservations about custom tags. My collection makes heavy use of non-standard tags, such as performer:<instrument>, opus, key, subtitle, or style. I can't find anything in the documentation about which tags are supported, or for which formats (I keep A/V registrations in Matroska containers, using the same tagging convention -- like using performer:<role> to save actor credits).
There is value for me in having a central database for this data, for example to find misspellings of the same name. But the fear of having 20 years of custom data entry destroyed by an overzealous tool makes me very hesitant to even try a solution like this.
Wow I've needed harmony for years, thanks for sharing! My dumb ass was filling out the Musicbrainz by hand for like two months before I just gave up on beets.
You'll want to bookmark Harmony even if you're using Picard since it's driven by the same database. Happily, once you add an item using Harmony, Picard can find your just-entered release almost immediately via the Release ID of the release you just created.
I find myself needing to create releases for ~10% of the albums I tag, and Harmony is a game-changer for that.
I spent a truly obnoxious amount of time importing my music library into beets. It took a couple of weeks to get to 95% imported, and got so bogged down in the last 5% that I never completed the import and never switched over.
This isn't necessarily a fault with beets, really, but a model mismatch. The model of beets is very, very strongly tied to associating each imported item to one well-known, commercial release. While it's possible to stray from that, it takes tons of time and experimentation to cram some things into its model.
Purchased, popular albums are a breeze; they import nicely and make sense. I struggled differing amounts with:
* brand new indie label releases (bandcamp)
* commercial albums variants missing from musicbrainz/discogs
* non-commercial albums (self-released CDRs)
* fan-recorded concerts
* fan-recorded festivals (a special case, a true nightmare)
* fan edits/remixes of commercial releases
* playlists & mix tapes
* mixed media releases
Each was eventually possible, but sometimes it took hours to figure out how to import a specific folder. Worse, after doing one festival it didn't necessarily make it easier to do the next festival. Even if I get to 100% imported, additional imports will still take thought.
This isn't an argument against it, I still think it's a fantastic tool. Just understand that the farther you stray from collecting commercial releases, the more of a struggle it is.
> * commercial albums variants missing from musicbrainz/discogs
I fixed those two by adding the missing releases to the database beets uses as its data source (musicbrainz.org), and importing the album in Beets afterwards. I still get notifications for edits to entries I contributed over a decade ago!
Everyone else is telling you to go update musicbrainz, and that's a sensible course of action for the first two bullets, but the much easier path is to import all these files with whatever metadata they have.
There is no canonical metadata for a fan recording of a concert or DIY CD-R, so you lose out on nothing by importing the files as they are today.
Once you're over the hump of the first import, beets is a fabulous tool for ingesting new music. It's well worth it.
This is the answer I arrived at myself a few hours into importing my own music library. Cleaning up metadata for commercial releases is a nice feature beets offers, but the real value (IMO) of beets is it's a powerful toolbox for managing a large music library, and that's true even if you were to eschew musicbrainz integration for all your music.
I do think it's a pretty fair complaint that it really feels like the software is fighting against you when you first encounter something absent from the musicbrainz database (especially if it's something fundamentally unsuited to be added to the database), but I'm not sure if there's an easy solution other than telling people "just hit the `import with existing metadata' button, it's totally fine" when they complain about it.
That's really surprising! I would think they'd want some kind of "official" source for such things, otherwise you'd end up with numerous entries for random live recordings of varying quality.
I think the only thing I've seen is that there should be at least some notoriety to the bootlegs. Like that it's a bootleg that's been circulating around in fan communities or a batch of fake CDs sold in some shady market or a leaked version of an upcoming album shared on P2P. Something that makes it beyond just a thing between a couple of people.
I think if you read through the MusicBrainz about page, it makes a lot more sense:
>As an encyclopedia and as a community, MusicBrainz exists only to collect as much information about music as we can. We do not discriminate or prefer one "type" of music over another, and we try to collect information about as many different types of music as possible. Whether it is published or unpublished, popular or fringe, western or non-western, human or non-human — we want it all in MusicBrainz.
Can I ask what you choose to manage your library with today? I feel like streaming has made me stray so far from the joy (and pain) of library curation, and I’d really like to get back to it, I just don’t know what folks are using these days.
I have all of my mp3s on a NAS and point plex at it to index. I’ve also set it up as a source for Sonos (which I’m growing less fond of as each day passes).
I guess my point is that I like having a directory of music (organized by artist/album) and make the discovery applications I use do the work of finding and playing the music I want.
I've always struggled to get my classical CDs tagged in a way that makes sense to me. Apple has figured it out in their classical music app and I should probably see if I can copy what they did in my personal library.
> fan-recorded festivals (a special case, a true nightmare)
I've long enjoyed extracted audio from eg Glastonbury sets. I've only got a few that I particularly enjoyed and where the specific track was on youtube - is it that sort of thing? Is there a community of such reprobates?
It provides a nice interface for managing and editing tags that integrates well with other command-line tools and keeps your file system well-organized/up-to-date as you edit the tag contents. The batteries-included integration with datasources like musicbrains and discogs is nice, but, at least for me, beets is mainly a better tool for accomplishing the same tag and file janitoring I was doing with things like foobar2000 and eyeD3 a decade ago.
If you're using Navidrome or similar to stream your music then check out beets-alternatives [0]. It lets you sync (and optionally convert) your library or a subset of it to another location, in my case my music storage mounted with Rclone. It's especially useful if you need to have a different naming structure in your target directory for whatever reason. I like to keep each disc of a multi-disc album in in its own subdirectory but most streaming servers seem to prefer all tracks of an album to be in the same directory. With Beets-alternatives I can have a different naming structure for each collection vs. having to rename my primary collection to suit whatever streaming server I happen to be using.
I've tried many times to find a nice UI for beets and somehow never come across this. It is exactly what I've been searching for all these years... Thanks for sharing!
I spent a couple of nights working out how to configure beets to my liking and have loved it ever since. My “workflow” is now:
- buy album on Bandcamp
- download zip
- beet import {zip file name}
And beets extracts the zip, matches the album to musicbrainz, updates any metadata, and drops the files into the directory structure that I like (naming the files how I like them, too).
Very rarely an album will need some more attention, in which case I use Picard to fix it before using beets to import it.
A very similar workflow on my end, both beets as the main tagger/organizer and Picard to pick up whatever can't be processed through beets. Beets is amazing!
Sometimes there’s just weirdness though - for example recently I bought an album and the band included all the tracks from their previous album as bonus tracks. So I used Picard to split them into the 2 “proper” albums.
I mean, you can just do that in the browser too. "Enter ID" allows you to enter the MusicBrainz UUID (or just full URL). You can even do in the command itself.
(Only because I have an axe to grind) I really dislike "genre" and work hard to strip that metadata tag. It is so reductive. And flawed as well.
Do we assign a "genre" to an artist? Album? Song?
The band R.E.M.: "College Rock"? "New Wave"? "Alternative"?
(Wow, don't even get me started on all the ways you can further slice up "alternative" into "shoe-gaze", "twee", etc. It's, ha ha, naval gazing for the music intelligentsia.)
R.E.M.'s "Don't Go Back to Rockville"? "Country"?
I think "live" and "soundtrack" might be the only two interesting "genres".
More seriously, there’s no reason you can’t assign more than one genre to a song, or say that “it’s a mix of A and B”. You can extend the concept to albums and artists just as easily.
I’m not familiar with R.E.M., but on rateyourmusic.com they’re marked as alternative rock, jangle pop, pop rock, indie rock, neo-psychedelia, and folk rock. “Don’t go back to Rockville” is marked as Jangle pop, and the “Reckoning” album overall is mostly Jangle pop with some influences from post-punk and Paisley Underground.
You may not agree with these categorizations, but it’s a crowd-sourced website so you can go vote for different genres. You may also disagree with the existence of all of these genres, some of which I’ve never heard of myself (but as I said, I haven’t listened much to R.E.M. in general), but my point is more that if you’re familiar with them then you kind of know what to expect from the song.
I’d never heard “don’t go back to rockville”, but just from these genres you named to make your point, I already knew what it would approximately sound like. And indeed listening to it confirmed that.
RYM genres would be extremely useful to have as a beets autotag plugin and I've been waiting for years for their API to be opened up specifically for them.
Genres are fun to play with too. If something has breakdown and maybe some screams? Put 'core' at the end. Its all up to interpretation and it gives us some easy short hand to discuss music.
I don't think theres anything wrong with labelling genres, its good for describing things. If I say "I love post hardcore, does anyone have more of that?" its a lot easier than going "hear me I out, I like punk sounds but maybe not necessarily punk songs?" What about movies? Can books have a genre? seems like a silly thing to take issue with. Alternative is a huge umbrella and it means something very different now than it did in the 90's when it meant pavement or sonic youth.
'naval gazing for music intelligentsia' did a pitchfork writer run over your mom?
I Find it stupid for artists but helpful for albums. NB: I use RYM as source (via https://git.sr.ht/~q3cpma/rymscrap) for my extended metadata, gotta exploit that autism.
There are probably broad categories that I might find myself using, but for my own core collection of music (deeper than it is wide, mostly), it barely matters.
I've loved beets (and MusicBrainz Picard) for years. This bit from the beets docs has always stuck with me when organizing my library.
"An Apology and a Brief Interlude
I would like to sincerely apologize that the autotagger in beets is so fussy. It asks you a lot of complicated questions, insecurely asking that you verify nearly every assumption it makes. This means importing and correcting the tags for a large library can be an endless, tedious process. I’m sorry for this.
Maybe it will help to think of it as a tradeoff. By carefully examining every album you own, you get to become more familiar with your library, its extent, its variation, and its quirks. People used to spend hours lovingly sorting and resorting their shelves of LPs. In the iTunes age, many of us toss our music into a heap and forget about it. This is great for some people. But there’s value in intimate, complete familiarity with your collection. So instead of a chore, try thinking of correcting tags as quality time with your music collection. That’s what I do."
Beets is amazing. The fact that it exists is a blessing for those like us who maintain our own music library.
I've been wanting to build my own Plex alternative for a while now. I've tried all the other tools out there, but Plex is definitely the least bad tool that let me enjoy my music without subscribing to Spotify and others.
I've already spent hours trying to figure out all the things that are required to make this new system. And it's unbelievably complicated. From parsing metadata to converting music, to understanding how deep and complex tagging work for music (yeah, it doesn't sound like it at first, but it truly is infinitely complex), I'm overwhelmed by how hard it would be to build a tool that compete with a combination of beets/iTunes.
There are a lot of alternatives for music servers in the world. Can I ask if you've explored Plex alternatives (Jellyfin, Emby, Lyrion, Navidrome, various Subsonic/OpenSubsonic-compatible servers, etc.), and if so, what you need that they don't do?
I ask because I'm working on a new music app for iOS which has pluggable sources, and so my bias is that most of the issues preventing me from easily enjoying my music library are actually on the client side.
I can install Plex, all my movies, tv-shows, anime and music is there.
Then I can drop Plexamp on my phone, log in with my Plex account and all my music is right there and if I rate a song on my phone, it's updated on Plex and vice versa.
Plexamp is what I settled on too, before deciding that I was only going to be happy if I built something better.
Plexamp is quite good as a standalone app, but as close to greatness as it is, it's a React Native app (better than a Flutter app, but still) that doesn't play well with the iOS ecosystem — no widgets, virtually no Siri/App Intents support, no Apple Music integration, etc.
I also decided that it shouldn't matter where my music "lives", and so it supports any local, self-hosted, or cloud-based source. (These are working today: https://imgur.com/szrkeIJ)
Tried Beets once because it pairs well with Navidrome, it's pretty feature packed and exhaustive in what it does. Though after fiddling with it and realizing that it's more effort for my use case than I'd like I ditched it and now I don't really bother tagging. I hope to do the same thing for Navidrome too and find something to replace it with similar to KDE Elisa which makes it easy to make playlists on the fly and organize music around folders instead of tags.
Different strokes for different folks, that's not to say that Navidrome and Beets aren't amazing pieces of kit.
I have this problem too. I find it completely useless to have tags like genre=Post Rock Jazz Fusion" or whatever with basically one band in each genre.
The other thing I've never quite got right is how to deal with classical music and popular music with multiple pressings. A lot of the tagging structure seems oriented around popular music with just one pressing. But I have like 10 different recordings of The Planets and several versions of Red Hot Chili Pepper's Californication, for example.
Set whitelist to a file path and then add a file with each genre that you want to keep on a new line. You can use the top-level genres in this file to base it off (except the whitelist isn't YAML, just a genre per line): https://raw.githubusercontent.com/beetbox/beets/master/beets...
Be sure to enable canonical so it converts the specific genres into their parent genre.
I recently looked into beets, but it seems it’s focused on automation, is that correct? I often buy stuff that has just been released, and that’s not yet in any database, so no automation works.
My current workflow is download, unzip and manually tag (mainly genres, sometimes badly named artists/albums) with MusicBee (which gives me autocomplete for genres so I can reuse existing ones), and then copy the files to the server for Navidrome to pick them up.
Throwing this out here to see if beets would actually work well with what I want after all.
If you add a lot of new stuff to your library, you'll often need to add them to MusicBrainz yourself. I use beets and I currently have added 2697 releases (2543 release groups) to MusicBrainz.
Beets is what got me into MusicBrainz. It's an incredible resource. To be fair it's not the easiest of things to get started with and the usability could be better (release drafts, anyone?) but it's efficient once you get the hang of it.
First: Thank you! Secondly: I wish more people did this. If you end up needing to manually enter the data, at least if you add it to a public database more people can benefit, which also makes it feel less frustrating.
I'm in the midst of a major music library overhaul and would not know how I'd ever get it done without Beets. For example, it's clearing out embedded images, fetching new hi-res artwork and verifying FLAC integrity, as I go through artist by artist.
I use MusicBee for everything too and sounds like my workflow is the same as yours. I basically take whatever is tagged on the file straight out of Bandcamp and then manually follow Discogs-like rules for normalization of the artist name so the directory structure works out, plus adding iTunes compilation metadata to deal with the "various artists" issue, adding BPM, customizing genre and comments etc.
I think I started all this fussiness exactly because back in the old days when you synced to CDDB, you'd get somebody else's useless record label classification for a reissue or international release that obscured the original label, which in certain genres of music is quite important to be able to better categorize the music for auto playlists. Original year is also critical for auto playlists where you're looking for tunes around a certain era. I also have a vague memory of old CDJs or Serato or some damn thing not displaying the name I wanted so having to shuffle stuff around to Composer or Album Artist or something to make it work, and now here I am with custom-tagged all the things.
My problem is I was considering replacing all my 320k MP3s with FLACs, which I have for all my Bandcamp purchases and most of my CD rips (although not from Beatport and other online stores)... but I realized it was going to be absolute hell to try and match the untagged files that I downloaded and put straight into cloud storage with the tagged files I already have in the library. It would need some kind of heuristics to fuzzy match the files with what's already there, then retag them with my preferred tags from the MP3, then move the file into the same location the MP3 is, then move the MusicBee proprietary metadata (play count, add date etc) across, then delete the MP3. I keep thinking there has to be a program or a plugin for this, but seems like not, so for now I've just given up and decided that 320k is good enough, actually, which makes me wonder why I'm keeping all these FLACs around in the first place. Not like I'm starting the Library of Alexandria here, I just like music.
> actually, which makes me wonder why I'm keeping all these FLACs around
Sometime last year, I actually redownloaded all my BC files, and reripped all my CDs (had to re-do genres, but that was fine, I wanted cleaner genres anyway). It’s simply archival. Lossless means you can convert it to whatever file you might want in the future without loss of quality.
This made extra sense for my CD rips, because many of those were from the early 2000s and I made questionable quality descisions ;)
I'm wondering about diminishing returns, because up until switching to FLAC about 10 years ago I ripped all my CDs to 320k and thus only have about 20 files in my "low bitrate" auto playlist that are weird indie or bootleg things that I couldn't bring myself to delete for nostalgia reasons. Is FLAC really all that much better? Can anybody hear the difference? I don't want to spend months awkwardly moving thousands of files around just because it's more useful to a hypothetical future archaeologist version of myself with hearing that got better with age instead of worse...
That said, I wouldn't mind going back and retagging my genres in particular because one of the annoying things is that over the years my perception of the genre of certain tracks has changed, so it might be easiest to just roll back to ID3v1 "we got both kinds: house and techno" genres for the broad category and then cram subgenre keywords into the comment field instead... but it's been so long since I used CDJs that maybe all these hacks to maintain compatibility with different players are moot now. I saw that the latest and greatest DJ equipment just hooks up direct to online streaming services so I wouldn't be surprised if it also had a feature to sync everything to MusicBrainz as soon as you plug in a USB, which makes all the fussing around with legacy tags kind of pointless for a casual who mostly just listens or mixes at home.
Pioneer DJ, I mean AlphaTheta, probably don't even know MusicBrainz exists. They're too busy selling subscriptions for RekordBox. And they do nothing to help you with the metadata on your files, besides filtering in browser mode.
I can relate to the problem of revising genre or energy ratings over time. I've gone with custom genre tags for ages, ie "dub/house/techno" or "funk/disco/edits" with a sprinkling of extra qualifiers in the comment field and do bulk updates from MP3Tag/Foobar2k. The extras only really help when preparing "crates" for export to USB for outside use, or when just playing off the entire collection at home. I'm fast, but still not much time to read the comment fields when browsing on the players, much less input any words with the scroll wheel.
I keep every purchase around in FLAC, and the part I might realistically play out stays in AIFF, for minimum fiddling of tags (ie stars map a bit differently between Traktor and Rekordbox) - because of course Rekordbox will warn you you're exporting files you can't play anywhere, but won't do anything to transcode them.
Lossless whenever possible because I just want to give the sound quality as much of a chance as I can when recording sets, especially if they might get posted online and getting lossy-transcoded multiple times. I've tried the mp3 of mp3 thing, and you do hear it at home (out at a gig, most of the time, probably not).
I don't suffer from track bulimia, so the numbers work out - and disk space has gotten a lot cheaper in the last 20 years.
It seems like many people online say that there is no perceptible difference between e.g, 320kb/s mp3 and "flac" (which usually means CD-quality 44.1kHz 16-bit). Often sources say something about there being no way it sounds different and then talk about the Nyquist theorem [0].
Personally I don't get it as for me, there is a clear difference not only between mp3 vs. CD, but even between different bitrates beyond that. Maybe I'm not typical as I've been usually listening to stuff through studio monitors and also usually through a recording interface which handles 192kHz and >24-bit.
Definitely I noticed on certain systems you aren't going to notice a difference as the system itself is the bottleneck (i.e, bluetooth). In my experience though if you use the right driver, so ASIO or WASAPI in Windows (or anything in Mac and Linux nowadays), I can tell the difference instantly on recordings I know well.
Most music did not get released in ultra HD but some things are available in 96kHz and beyond. I recommend checking out Radiohead, Bob Marley, or Pink Floyd in ultra-HD ( >= 48kHz, >= 24-bit) as there have been releases. I have found Bob Marley - Legend in 192kHz 24-bit and it sounds incredible. You can hear each individual member of the percussion section.
The core benefit to me of FLAC is that it's lossless. I can convert the same file to an MP3 for the car and OPUS on my phone and don't have to worry about (compounding) losses in quality.
I stream FLAC at home because I already have it there but I can't say it's "better" than 320 MP3s.
I did a re-rip project during Covid because so many of my rips were 128 bit MP3 from back when storage costs drove all CD ripping decisions.
I started ripping before FLAC was a thing, and my first albums are all in 256k CBR mp3. I've never seen a big reason to re-do them. Yes, the files have gone through LAME's acoustic filter so they don't match the original bit-for-bit, but I mostly can't tell the difference. There have been a few times when I got annoyed by audio quality issues, but it's been 50/50 that it turned out those were also on the cd, so either a deliberate choice or a mistake during mastering.
As others say, there is an issue with compounding losses when you transcode mp3 to vorbis or opus, but I've never needed to do that -- and when I export, it's to move the files to lower-quality devices anyway.
320 vs FLAC: unless you are both a mutant and audiophile with the right equipment, you won't hear any difference at all. It's purely for archival purposes for me, that I have the "originals" or as close as possible.
Legacy tags: luckily not an issue for me, I only care about multiple (usually metal) subgenres and that my players show/filter them.
You can use the current tags in the media files and not use any metadata database. I have a simple scheme for organizing my collection, and can correct my metadata rather quickly, but these days, so I don’t even use beets.
Looks like a general case. I do the same a lot with Picard and Foobar, but beats can also auto tag based on file name and populate musicbrainz. Set library directory to that of navidrome so it will be coppied there on import.
I like beets' functionality but don't as much like "babysitting" it for long sessions when I have a bunch of new music to import. Especially because crashes are frequent and sometimes make it lose all incremental progress. Lack of progress bar on large imports is painful also.
It would be nice if a beets background worker could be kicked off when new music is detected to avoid the sequential nature of the commandline and see all the waiting tagging actions at once. This[0] is supposed to do that but it appears deprecated in favor of something else[1].
That being said, the "beets way" works pretty well past the import stage, so I haven't felt the need to move off it for local music.
It's more for adding tags before you add them to your collection. You can setup shortcuts to often used tags. Of course it can also rename and move files based on tags.
I then can auto create playlists based on those tags (with a different program)
Well yes, it's not that hard to use but then even using the terminal for an average user is complicated.
There is a webpage on beets website where they list all the functions and explain what it does. Very useful when you go back after half a year and can't remember how it works.
If anyone is looking at options in this space, I used to use Media Monkey for organising and curating my large offline music collection. It did things like metadata and album art management, album-level volume levelling, and would even reorganise the file structure of your music to whatever pattern you set.
I also loved that you could set up separate "libraries", so I could have metal and classical arranged by album, and EDM as a bunch of single tracks that are organised into playlists. Only downside is that it's Windows software.
Correct me if I'm wrong, but different target users here :) Beets seems to be for consumers who want to organize their music collection, and while Mixxx can help with that too, it's mainly for music connoisseurs who play for others (AKA DJs). I'm not sure how well someone looking for the first could use Mixxx, the interface isn't exactly the epitome of user-friendliness if you just wanna organize then press play :)
Mixxx is best used with a DJ console, however for me it is also easily used as a player that automatically picks a song from some selected crates in Auto DJ mode.
So I'm in the mood for Salsa, I put that crate as the AutoDJ source.
Later I can change it to progressive rock, or 90's disco, and it will randomly pick songs from the selected crate or crates.
I can also do DJ things, like looping some sections or mashup some songs.
Having hit those first two pain points, I changed my transcode config to 1) run every time, and 2) use a custom bash script as the transcode command. The bash script keeps a plaintext list of files + their modtime and disk size, then only transcodes files it hasn’t seen before. Because it’s a plaintext list and relatively small, there’s not a terrible performance hit since it’s paged into memory.
Not at all saying this to dismiss your criticism though; absolutely would be great to have better OOTB options. Just putting it out for anyone who wants a workaround :).
Consider a case where a DJ's entire show can be downloaded as a single MP3, AND there is a complete playlist for the show.
Is there any software that can use this to break up the single mp3 into the numerous individual song tracks ? And insert metadata ? And perhaps also fingerprint each track to match it to a specific version of the song ?
If the playlist is something like a CUE file then yes, certainly. CUETools/XLD/foobar2000 can split by cue file. And Picard can do audio fingerprints to match tracks and get metadata for them.
In the case I describe there's no CUE files. There's a DJ writing the playlist as the show proceeds, and the typical entry is Artist and Song title. So it's enough information to do a search, but maybe not enough to identify the particular version, and then you have to find the boundary between tracks in the mp3.
So this might be a task for A.I. - to automate it with sufficient precision.
Don't most DJ sets have elaborate cross-fades between tracks? It may not be possible to find a clean cut, and the best you can hope for is good judgement rather than precision.
Point taken! So give me an A.I. that knows when to begin/end a track and when to keep two or three tracks glued together cos the segue is sooo smooooth.....
What I really want is a plugin that starts an icecast server with a given song as a root, and continues to stream songs whose albums/artists are within a certain reference distance of the root on discogs/wikipedia/etc.
So if I start on Guzzlemug I'll get some Blood Incantation, but maybe also some Pink Floyd.
How do you all handle FLAC/CUE releases? A problem I have is that I want to have processed/tagged tracks, but I also want to keep FLAC/CUEs. I end up have redundant and separate organized library and original data.
I, personally, split the flac. It sounds like you don't want to do that but though but it does remain lossless amd I prefer to see the tracks representing individually in the filesystem.
I just split everything into tracks like they are on the album. Is there a tangible benefit to a single large FLAC with a CUE to tell the player where the songs are?
- The CUE maps the audio in a "CD-native" way, like each track can have 1 or more indices, which enables things like pregaps. When you split FLACs into discrete tracks you need to figure out how to project that onto files, do you put the pregaps at the end of previous tracks? At the beginning of next tracks? In short, with a CUE you can emulate the CD experience more accurately.
- If you are seeding a torrent, you need to keep the original data intact.
I've been thinking about this recently too, mostly to figure out how to preserve indexes in tracks of "Murmurs of Earth"[0]. I think the FLAC embedded cuesheets could be used for this. While intended for CD-ready cuesheets embedded into one big FLAC, there's no reason they couldn't be used in individual files, with index 00 for pregaps.
I'm not sure it's well supported anywhere however. Maybe foobar2000?
Edit: I pulled up the RFC[1] for FLAC, and it seems like this is a use case they had in mind for the embedded cuesheet:
> A cuesheet metadata block can be used either [...] or to provide a mechanism to store locations of interest within a FLAC file.
Auto-tagging is nice to have but it will screw up verification checksums. I would prefer having metadata in a separate DB and keep the files as is, unless I explicitly request exporting the metadata to the original files.
The -C option to avoid copying files and the -W option to prevent writing tags (or set it in the config file)
It will still attempt to match and store the results in its DB for later writing to files, or moving, or querying.
For the data to be most useful, you will need a client that connect to Beets to get the data though, Im not sure how prevelant those are. You can use the web-ui plugin though.
It's a good tool. However, last I checked, it was not possible to run it in a one-shot stateless fashion, like, passing it a list of music files so it auto-fetches album art, lyrics and updates the very same input files.
> Relatedly, the -q (quiet) option can help with large imports by autotagging without ever bothering to ask for user input. Whenever the normal autotagger mode would ask for confirmation, the quiet mode performs a fallback action that can be configured using the quiet_fallback configuration or --quiet-fallback CLI option. By default it pessimistically skips the file. Alternatively, it can be used as is, by configuring asis.
Beets with this flag works relatively well for me when wrapped in a shell script, but I have spent a lot of effort trying to make it actually non-interactive.
I'm still not totally sure I got all the certainty thresholds right, I still sometimes get unexpected behavior when using Beets this way, so I agree that the non-interactive experience is lacking.
Wow, I remember beets from about 10 years ago. I never got into meticulously organizing my music, but I'm glad to see it's still around. Also the site loaded absolutely ridiculously fast.
Does this have a bulk command to look at the metadata of files in a folder and use those to rename the files? Assume the current filenames are just random strings
A few people commenting that some of their collection "doesn't exist in any DB", the best way to fix it is to add it to Musicbrainz[0] yourself!
I have found that adding things to Musicbrainz is actually pretty easy (and if you are so inclined like me, pretty rewarding and fun).
Streaming releases (and Bandcamp) you simply drop the release URL into Harmony[1] and it does most of the work for you.
Musicbrainz can represent nearly everything musically related and its all freely licensed, a very cool thing to exist.
Most (non-destructive) edits are auto-applied, whilst the rest go through a 7 day voting period (they are still applied by default unless someone votes against). The barrier to entry is very low.
0. https://musicbrainz.org/
1. https://harmony.pulsewidth.org.uk/
You're still assuming that all of a person's music metadata _belongs_ in a global database. Some of my collection includes:
-"frankenstein" musical soundtracks where I've assembled my favorite version of each song from all the various published recordings. Sometimes I've even edited pieces of different recordings together into a hybrid track. -a soundboard recording of my friend's high school talent show performance -Music I've personally recorded from video games or other random sources -Songs where I've edited out parts I don't like (such as overly-long drum solos) leaving just the parts I do.
I've organized these things in a way that makes sense for me, not for the world.
This is like a bookshelf that can only store books with an ISBN that can be classified in the Dewey decimal system. Too bad about your family photo scrapbooks or your personal sketchbooks!
I'm not sure that's what the earlier poster ment.
What I think they're doing (and what I've done) is add music that wasn't already in the global database to it. For example, a promotional CD, or releases from a small publisher, etc.
In these cases, you're not adding personal metadata or mixtapes.
That said, definitely do self organization if it works for you. Most of your examples seem like they're suitable for custom playlists.
I have similar reservations about custom tags. My collection makes heavy use of non-standard tags, such as performer:<instrument>, opus, key, subtitle, or style. I can't find anything in the documentation about which tags are supported, or for which formats (I keep A/V registrations in Matroska containers, using the same tagging convention -- like using performer:<role> to save actor credits).
There is value for me in having a central database for this data, for example to find misspellings of the same name. But the fear of having 20 years of custom data entry destroyed by an overzealous tool makes me very hesitant to even try a solution like this.
Wow I've needed harmony for years, thanks for sharing! My dumb ass was filling out the Musicbrainz by hand for like two months before I just gave up on beets.
If you have the files downloaded, picard is also useful - https://picard.musicbrainz.org/
You'll want to bookmark Harmony even if you're using Picard since it's driven by the same database. Happily, once you add an item using Harmony, Picard can find your just-entered release almost immediately via the Release ID of the release you just created.
I find myself needing to create releases for ~10% of the albums I tag, and Harmony is a game-changer for that.
I spent a truly obnoxious amount of time importing my music library into beets. It took a couple of weeks to get to 95% imported, and got so bogged down in the last 5% that I never completed the import and never switched over.
This isn't necessarily a fault with beets, really, but a model mismatch. The model of beets is very, very strongly tied to associating each imported item to one well-known, commercial release. While it's possible to stray from that, it takes tons of time and experimentation to cram some things into its model.
Purchased, popular albums are a breeze; they import nicely and make sense. I struggled differing amounts with:
* brand new indie label releases (bandcamp)
* commercial albums variants missing from musicbrainz/discogs
* non-commercial albums (self-released CDRs)
* fan-recorded concerts
* fan-recorded festivals (a special case, a true nightmare)
* fan edits/remixes of commercial releases
* playlists & mix tapes
* mixed media releases
Each was eventually possible, but sometimes it took hours to figure out how to import a specific folder. Worse, after doing one festival it didn't necessarily make it easier to do the next festival. Even if I get to 100% imported, additional imports will still take thought.
This isn't an argument against it, I still think it's a fantastic tool. Just understand that the farther you stray from collecting commercial releases, the more of a struggle it is.
> * brand new indie label releases (bandcamp)
> * commercial albums variants missing from musicbrainz/discogs
I fixed those two by adding the missing releases to the database beets uses as its data source (musicbrainz.org), and importing the album in Beets afterwards. I still get notifications for edits to entries I contributed over a decade ago!
Everyone else is telling you to go update musicbrainz, and that's a sensible course of action for the first two bullets, but the much easier path is to import all these files with whatever metadata they have.
There is no canonical metadata for a fan recording of a concert or DIY CD-R, so you lose out on nothing by importing the files as they are today.
Once you're over the hump of the first import, beets is a fabulous tool for ingesting new music. It's well worth it.
This is the answer I arrived at myself a few hours into importing my own music library. Cleaning up metadata for commercial releases is a nice feature beets offers, but the real value (IMO) of beets is it's a powerful toolbox for managing a large music library, and that's true even if you were to eschew musicbrainz integration for all your music.
I do think it's a pretty fair complaint that it really feels like the software is fighting against you when you first encounter something absent from the musicbrainz database (especially if it's something fundamentally unsuited to be added to the database), but I'm not sure if there's an easy solution other than telling people "just hit the `import with existing metadata' button, it's totally fine" when they complain about it.
Fan recordings/edits/anything that shouldn't be on Musicbrainz just gets imported as-is, with maybe some metadata additions/tweaks.
> * brand new indie label releases (bandcamp)
> * commercial albums variants missing from musicbrainz/discogs
This is a great opportunity to fill in those blanks for those services :) I didn't have much to contribute to MB but did have a few albums to add.
>Fan recordings/edits/anything that shouldn't be on Musicbrainz
Those should be on MusicBrainz. There's even a bootleg release type for fan recordings/illegal copies, and official style guides for live bootlegs.
https://musicbrainz.org/doc/Style/Specific_types_of_releases...
That's really surprising! I would think they'd want some kind of "official" source for such things, otherwise you'd end up with numerous entries for random live recordings of varying quality.
I think the only thing I've seen is that there should be at least some notoriety to the bootlegs. Like that it's a bootleg that's been circulating around in fan communities or a batch of fake CDs sold in some shady market or a leaked version of an upcoming album shared on P2P. Something that makes it beyond just a thing between a couple of people.
I think if you read through the MusicBrainz about page, it makes a lot more sense:
>As an encyclopedia and as a community, MusicBrainz exists only to collect as much information about music as we can. We do not discriminate or prefer one "type" of music over another, and we try to collect information about as many different types of music as possible. Whether it is published or unpublished, popular or fringe, western or non-western, human or non-human — we want it all in MusicBrainz.
Can I ask what you choose to manage your library with today? I feel like streaming has made me stray so far from the joy (and pain) of library curation, and I’d really like to get back to it, I just don’t know what folks are using these days.
I have all of my mp3s on a NAS and point plex at it to index. I’ve also set it up as a source for Sonos (which I’m growing less fond of as each day passes).
I guess my point is that I like having a directory of music (organized by artist/album) and make the discovery applications I use do the work of finding and playing the music I want.
There is a Bandcamp auto-tagging plugin for beets which should at least help with the first point: https://github.com/snejus/beetcamp
I've always struggled to get my classical CDs tagged in a way that makes sense to me. Apple has figured it out in their classical music app and I should probably see if I can copy what they did in my personal library.
Maybe worth looking at Roon for that? It’s not free but they handle classical music very well.
> fan-recorded festivals (a special case, a true nightmare)
I've long enjoyed extracted audio from eg Glastonbury sets. I've only got a few that I particularly enjoyed and where the specific track was on youtube - is it that sort of thing? Is there a community of such reprobates?
Musicbrainz can do events
https://musicbrainz.org/search?query=glastonbury&type=event&...
> Each was eventually possible, but sometimes it took hours to figure out how to import a specific folder.
Thanks for saving me time, I guess? I just maintain very spartan id3 tags on my music, artist, album, song name, track id, and that's about it.
What more would beets give me? How would it improve my experience?
It provides a nice interface for managing and editing tags that integrates well with other command-line tools and keeps your file system well-organized/up-to-date as you edit the tag contents. The batteries-included integration with datasources like musicbrains and discogs is nice, but, at least for me, beets is mainly a better tool for accomplishing the same tag and file janitoring I was doing with things like foobar2000 and eyeD3 a decade ago.
If you're using Navidrome or similar to stream your music then check out beets-alternatives [0]. It lets you sync (and optionally convert) your library or a subset of it to another location, in my case my music storage mounted with Rclone. It's especially useful if you need to have a different naming structure in your target directory for whatever reason. I like to keep each disc of a multi-disc album in in its own subdirectory but most streaming servers seem to prefer all tracks of an album to be in the same directory. With Beets-alternatives I can have a different naming structure for each collection vs. having to rename my primary collection to suit whatever streaming server I happen to be using.
[0]: https://github.com/geigerzaehler/beets-alternatives
One of my favorite beets projects is beets-flask.
It lets you set up fully or partially automated import pipelines with a nice web UI to manage any manual steps needed.
Importing is usually as simple as dropping a zip in a folder and the rest is managed automatically.
https://github.com/pSpitzner/beets-flask
I've tried many times to find a nice UI for beets and somehow never come across this. It is exactly what I've been searching for all these years... Thanks for sharing!
Nice! Have you figured out how to manage album art with beets-alternatives?
I spent a couple of nights working out how to configure beets to my liking and have loved it ever since. My “workflow” is now:
- buy album on Bandcamp
- download zip
- beet import {zip file name}
And beets extracts the zip, matches the album to musicbrainz, updates any metadata, and drops the files into the directory structure that I like (naming the files how I like them, too).
Very rarely an album will need some more attention, in which case I use Picard to fix it before using beets to import it.
A very similar workflow on my end, both beets as the main tagger/organizer and Picard to pick up whatever can't be processed through beets. Beets is amazing!
Isn't beets going to just overwrite whatever you did in Picard?
No, you can tell it to use the metadata as-is.
Sometimes there’s just weirdness though - for example recently I bought an album and the band included all the tracks from their previous album as bonus tracks. So I used Picard to split them into the 2 “proper” albums.
Well, one might use picard to find a musicbrainz release id, so that beetz has something to grab on to when importing.
I mean, you can just do that in the browser too. "Enter ID" allows you to enter the MusicBrainz UUID (or just full URL). You can even do in the command itself.
(Only because I have an axe to grind) I really dislike "genre" and work hard to strip that metadata tag. It is so reductive. And flawed as well.
Do we assign a "genre" to an artist? Album? Song?
The band R.E.M.: "College Rock"? "New Wave"? "Alternative"?
(Wow, don't even get me started on all the ways you can further slice up "alternative" into "shoe-gaze", "twee", etc. It's, ha ha, naval gazing for the music intelligentsia.)
R.E.M.'s "Don't Go Back to Rockville"? "Country"?
I think "live" and "soundtrack" might be the only two interesting "genres".
A genre is a volume in song space.
More seriously, there’s no reason you can’t assign more than one genre to a song, or say that “it’s a mix of A and B”. You can extend the concept to albums and artists just as easily.
I’m not familiar with R.E.M., but on rateyourmusic.com they’re marked as alternative rock, jangle pop, pop rock, indie rock, neo-psychedelia, and folk rock. “Don’t go back to Rockville” is marked as Jangle pop, and the “Reckoning” album overall is mostly Jangle pop with some influences from post-punk and Paisley Underground.
You may not agree with these categorizations, but it’s a crowd-sourced website so you can go vote for different genres. You may also disagree with the existence of all of these genres, some of which I’ve never heard of myself (but as I said, I haven’t listened much to R.E.M. in general), but my point is more that if you’re familiar with them then you kind of know what to expect from the song.
I’d never heard “don’t go back to rockville”, but just from these genres you named to make your point, I already knew what it would approximately sound like. And indeed listening to it confirmed that.
RYM genres would be extremely useful to have as a beets autotag plugin and I've been waiting for years for their API to be opened up specifically for them.
Genres are fun to play with too. If something has breakdown and maybe some screams? Put 'core' at the end. Its all up to interpretation and it gives us some easy short hand to discuss music.
"…alternative rock, jangle pop, pop rock, indie rock, neo-psychedelia, and folk rock…"
Yeah, that is kind of making my point. Others perhaps enjoy creative pigeon-holing.
I don't think theres anything wrong with labelling genres, its good for describing things. If I say "I love post hardcore, does anyone have more of that?" its a lot easier than going "hear me I out, I like punk sounds but maybe not necessarily punk songs?" What about movies? Can books have a genre? seems like a silly thing to take issue with. Alternative is a huge umbrella and it means something very different now than it did in the 90's when it meant pavement or sonic youth.
'naval gazing for music intelligentsia' did a pitchfork writer run over your mom?
[dead]
The classic line more or less sums it up: "Writing about music is like dancing about architecture."
King Gizzard and the Lizard Wizard
Genre: Yes
I Find it stupid for artists but helpful for albums. NB: I use RYM as source (via https://git.sr.ht/~q3cpma/rymscrap) for my extended metadata, gotta exploit that autism.
Same, glad to know I'm not alone!
Agreed!
There are probably broad categories that I might find myself using, but for my own core collection of music (deeper than it is wide, mostly), it barely matters.
I've loved beets (and MusicBrainz Picard) for years. This bit from the beets docs has always stuck with me when organizing my library.
"An Apology and a Brief Interlude
I would like to sincerely apologize that the autotagger in beets is so fussy. It asks you a lot of complicated questions, insecurely asking that you verify nearly every assumption it makes. This means importing and correcting the tags for a large library can be an endless, tedious process. I’m sorry for this.
Maybe it will help to think of it as a tradeoff. By carefully examining every album you own, you get to become more familiar with your library, its extent, its variation, and its quirks. People used to spend hours lovingly sorting and resorting their shelves of LPs. In the iTunes age, many of us toss our music into a heap and forget about it. This is great for some people. But there’s value in intimate, complete familiarity with your collection. So instead of a chore, try thinking of correcting tags as quality time with your music collection. That’s what I do."
Beets is amazing. The fact that it exists is a blessing for those like us who maintain our own music library.
I've been wanting to build my own Plex alternative for a while now. I've tried all the other tools out there, but Plex is definitely the least bad tool that let me enjoy my music without subscribing to Spotify and others.
I've already spent hours trying to figure out all the things that are required to make this new system. And it's unbelievably complicated. From parsing metadata to converting music, to understanding how deep and complex tagging work for music (yeah, it doesn't sound like it at first, but it truly is infinitely complex), I'm overwhelmed by how hard it would be to build a tool that compete with a combination of beets/iTunes.
There are a lot of alternatives for music servers in the world. Can I ask if you've explored Plex alternatives (Jellyfin, Emby, Lyrion, Navidrome, various Subsonic/OpenSubsonic-compatible servers, etc.), and if so, what you need that they don't do?
I ask because I'm working on a new music app for iOS which has pluggable sources, and so my bias is that most of the issues preventing me from easily enjoying my music library are actually on the client side.
Frictionless use is the biggest one.
I can install Plex, all my movies, tv-shows, anime and music is there.
Then I can drop Plexamp on my phone, log in with my Plex account and all my music is right there and if I rate a song on my phone, it's updated on Plex and vice versa.
Plexamp is what I settled on too, before deciding that I was only going to be happy if I built something better.
Plexamp is quite good as a standalone app, but as close to greatness as it is, it's a React Native app (better than a Flutter app, but still) that doesn't play well with the iOS ecosystem — no widgets, virtually no Siri/App Intents support, no Apple Music integration, etc.
I also decided that it shouldn't matter where my music "lives", and so it supports any local, self-hosted, or cloud-based source. (These are working today: https://imgur.com/szrkeIJ)
Tried Beets once because it pairs well with Navidrome, it's pretty feature packed and exhaustive in what it does. Though after fiddling with it and realizing that it's more effort for my use case than I'd like I ditched it and now I don't really bother tagging. I hope to do the same thing for Navidrome too and find something to replace it with similar to KDE Elisa which makes it easy to make playlists on the fly and organize music around folders instead of tags.
Different strokes for different folks, that's not to say that Navidrome and Beets aren't amazing pieces of kit.
I love beets - the one thing I can't figure out is how to set the genre very wide.
I like having a small number of broad genres - Rock, Hip-Hop, Jazz etc - but the tagging comes up with hundreds of distinct genres :(
I have this problem too. I find it completely useless to have tags like genre=Post Rock Jazz Fusion" or whatever with basically one band in each genre.
The other thing I've never quite got right is how to deal with classical music and popular music with multiple pressings. A lot of the tagging structure seems oriented around popular music with just one pressing. But I have like 10 different recordings of The Planets and several versions of Red Hot Chili Pepper's Californication, for example.
I assume lastgenre with canonicalization[0] enabled is what you need, along with count=1 if multiple genres are an issue.
[0] https://beets.readthedocs.io/en/stable/plugins/lastgenre.htm...
I think I'm using that already.
Limiting to one genre I don't think will fix it - what I want is to define the allowable values for genre, and then limit to one.
It's probably better with the limit one though!
Set whitelist to a file path and then add a file with each genre that you want to keep on a new line. You can use the top-level genres in this file to base it off (except the whitelist isn't YAML, just a genre per line): https://raw.githubusercontent.com/beetbox/beets/master/beets...
Be sure to enable canonical so it converts the specific genres into their parent genre.
There's a whitelist option. Just read the docs, really, lastgenre is very configurable.
I use lastgenre plugin with a short whitelist.
I recently looked into beets, but it seems it’s focused on automation, is that correct? I often buy stuff that has just been released, and that’s not yet in any database, so no automation works.
My current workflow is download, unzip and manually tag (mainly genres, sometimes badly named artists/albums) with MusicBee (which gives me autocomplete for genres so I can reuse existing ones), and then copy the files to the server for Navidrome to pick them up.
Throwing this out here to see if beets would actually work well with what I want after all.
If you add a lot of new stuff to your library, you'll often need to add them to MusicBrainz yourself. I use beets and I currently have added 2697 releases (2543 release groups) to MusicBrainz.
Beets is what got me into MusicBrainz. It's an incredible resource. To be fair it's not the easiest of things to get started with and the usability could be better (release drafts, anyone?) but it's efficient once you get the hang of it.
First: Thank you! Secondly: I wish more people did this. If you end up needing to manually enter the data, at least if you add it to a public database more people can benefit, which also makes it feel less frustrating.
I'm in the midst of a major music library overhaul and would not know how I'd ever get it done without Beets. For example, it's clearing out embedded images, fetching new hi-res artwork and verifying FLAC integrity, as I go through artist by artist.
If you are into that, that’s good, but that’s the exact opposite of making my life easier ;)
I use MusicBee for everything too and sounds like my workflow is the same as yours. I basically take whatever is tagged on the file straight out of Bandcamp and then manually follow Discogs-like rules for normalization of the artist name so the directory structure works out, plus adding iTunes compilation metadata to deal with the "various artists" issue, adding BPM, customizing genre and comments etc.
I think I started all this fussiness exactly because back in the old days when you synced to CDDB, you'd get somebody else's useless record label classification for a reissue or international release that obscured the original label, which in certain genres of music is quite important to be able to better categorize the music for auto playlists. Original year is also critical for auto playlists where you're looking for tunes around a certain era. I also have a vague memory of old CDJs or Serato or some damn thing not displaying the name I wanted so having to shuffle stuff around to Composer or Album Artist or something to make it work, and now here I am with custom-tagged all the things.
My problem is I was considering replacing all my 320k MP3s with FLACs, which I have for all my Bandcamp purchases and most of my CD rips (although not from Beatport and other online stores)... but I realized it was going to be absolute hell to try and match the untagged files that I downloaded and put straight into cloud storage with the tagged files I already have in the library. It would need some kind of heuristics to fuzzy match the files with what's already there, then retag them with my preferred tags from the MP3, then move the file into the same location the MP3 is, then move the MusicBee proprietary metadata (play count, add date etc) across, then delete the MP3. I keep thinking there has to be a program or a plugin for this, but seems like not, so for now I've just given up and decided that 320k is good enough, actually, which makes me wonder why I'm keeping all these FLACs around in the first place. Not like I'm starting the Library of Alexandria here, I just like music.
> actually, which makes me wonder why I'm keeping all these FLACs around
Sometime last year, I actually redownloaded all my BC files, and reripped all my CDs (had to re-do genres, but that was fine, I wanted cleaner genres anyway). It’s simply archival. Lossless means you can convert it to whatever file you might want in the future without loss of quality.
This made extra sense for my CD rips, because many of those were from the early 2000s and I made questionable quality descisions ;)
I'm wondering about diminishing returns, because up until switching to FLAC about 10 years ago I ripped all my CDs to 320k and thus only have about 20 files in my "low bitrate" auto playlist that are weird indie or bootleg things that I couldn't bring myself to delete for nostalgia reasons. Is FLAC really all that much better? Can anybody hear the difference? I don't want to spend months awkwardly moving thousands of files around just because it's more useful to a hypothetical future archaeologist version of myself with hearing that got better with age instead of worse...
That said, I wouldn't mind going back and retagging my genres in particular because one of the annoying things is that over the years my perception of the genre of certain tracks has changed, so it might be easiest to just roll back to ID3v1 "we got both kinds: house and techno" genres for the broad category and then cram subgenre keywords into the comment field instead... but it's been so long since I used CDJs that maybe all these hacks to maintain compatibility with different players are moot now. I saw that the latest and greatest DJ equipment just hooks up direct to online streaming services so I wouldn't be surprised if it also had a feature to sync everything to MusicBrainz as soon as you plug in a USB, which makes all the fussing around with legacy tags kind of pointless for a casual who mostly just listens or mixes at home.
Pioneer DJ, I mean AlphaTheta, probably don't even know MusicBrainz exists. They're too busy selling subscriptions for RekordBox. And they do nothing to help you with the metadata on your files, besides filtering in browser mode.
I can relate to the problem of revising genre or energy ratings over time. I've gone with custom genre tags for ages, ie "dub/house/techno" or "funk/disco/edits" with a sprinkling of extra qualifiers in the comment field and do bulk updates from MP3Tag/Foobar2k. The extras only really help when preparing "crates" for export to USB for outside use, or when just playing off the entire collection at home. I'm fast, but still not much time to read the comment fields when browsing on the players, much less input any words with the scroll wheel.
I keep every purchase around in FLAC, and the part I might realistically play out stays in AIFF, for minimum fiddling of tags (ie stars map a bit differently between Traktor and Rekordbox) - because of course Rekordbox will warn you you're exporting files you can't play anywhere, but won't do anything to transcode them.
Lossless whenever possible because I just want to give the sound quality as much of a chance as I can when recording sets, especially if they might get posted online and getting lossy-transcoded multiple times. I've tried the mp3 of mp3 thing, and you do hear it at home (out at a gig, most of the time, probably not).
I don't suffer from track bulimia, so the numbers work out - and disk space has gotten a lot cheaper in the last 20 years.
It seems like many people online say that there is no perceptible difference between e.g, 320kb/s mp3 and "flac" (which usually means CD-quality 44.1kHz 16-bit). Often sources say something about there being no way it sounds different and then talk about the Nyquist theorem [0].
Personally I don't get it as for me, there is a clear difference not only between mp3 vs. CD, but even between different bitrates beyond that. Maybe I'm not typical as I've been usually listening to stuff through studio monitors and also usually through a recording interface which handles 192kHz and >24-bit.
Definitely I noticed on certain systems you aren't going to notice a difference as the system itself is the bottleneck (i.e, bluetooth). In my experience though if you use the right driver, so ASIO or WASAPI in Windows (or anything in Mac and Linux nowadays), I can tell the difference instantly on recordings I know well.
Most music did not get released in ultra HD but some things are available in 96kHz and beyond. I recommend checking out Radiohead, Bob Marley, or Pink Floyd in ultra-HD ( >= 48kHz, >= 24-bit) as there have been releases. I have found Bob Marley - Legend in 192kHz 24-bit and it sounds incredible. You can hear each individual member of the percussion section.
[0] https://en.wikipedia.org/wiki/Nyquist-Shannon_sampling_theor...
The core benefit to me of FLAC is that it's lossless. I can convert the same file to an MP3 for the car and OPUS on my phone and don't have to worry about (compounding) losses in quality.
I stream FLAC at home because I already have it there but I can't say it's "better" than 320 MP3s.
I did a re-rip project during Covid because so many of my rips were 128 bit MP3 from back when storage costs drove all CD ripping decisions.
Edit:typos
I started ripping before FLAC was a thing, and my first albums are all in 256k CBR mp3. I've never seen a big reason to re-do them. Yes, the files have gone through LAME's acoustic filter so they don't match the original bit-for-bit, but I mostly can't tell the difference. There have been a few times when I got annoyed by audio quality issues, but it's been 50/50 that it turned out those were also on the cd, so either a deliberate choice or a mistake during mastering.
As others say, there is an issue with compounding losses when you transcode mp3 to vorbis or opus, but I've never needed to do that -- and when I export, it's to move the files to lower-quality devices anyway.
320 vs FLAC: unless you are both a mutant and audiophile with the right equipment, you won't hear any difference at all. It's purely for archival purposes for me, that I have the "originals" or as close as possible.
Legacy tags: luckily not an issue for me, I only care about multiple (usually metal) subgenres and that my players show/filter them.
You can use the current tags in the media files and not use any metadata database. I have a simple scheme for organizing my collection, and can correct my metadata rather quickly, but these days, so I don’t even use beets.
Looks like a general case. I do the same a lot with Picard and Foobar, but beats can also auto tag based on file name and populate musicbrainz. Set library directory to that of navidrome so it will be coppied there on import.
[dead]
I like beets' functionality but don't as much like "babysitting" it for long sessions when I have a bunch of new music to import. Especially because crashes are frequent and sometimes make it lose all incremental progress. Lack of progress bar on large imports is painful also.
It would be nice if a beets background worker could be kicked off when new music is detected to avoid the sequential nature of the commandline and see all the waiting tagging actions at once. This[0] is supposed to do that but it appears deprecated in favor of something else[1].
That being said, the "beets way" works pretty well past the import stage, so I haven't felt the need to move off it for local music.
[0] https://github.com/sentriz/betanin
[1] https://github.com/sentriz/wrtag
I currently use OneTagger https://onetagger.github.io
It's more for adding tags before you add them to your collection. You can setup shortcuts to often used tags. Of course it can also rename and move files based on tags.
I then can auto create playlists based on those tags (with a different program)
Been using this for several years now. It's definitely not for the average user. But nothing is quite so flexible.
Well yes, it's not that hard to use but then even using the terminal for an average user is complicated.
There is a webpage on beets website where they list all the functions and explain what it does. Very useful when you go back after half a year and can't remember how it works.
The average user likely just streams from Spotify. The users collecting mp3s are probably more tolerant to the learning curve.
If anyone is looking at options in this space, I used to use Media Monkey for organising and curating my large offline music collection. It did things like metadata and album art management, album-level volume levelling, and would even reorganise the file structure of your music to whatever pattern you set.
I also loved that you could set up separate "libraries", so I could have metal and classical arranged by album, and EDM as a bunch of single tracks that are organised into playlists. Only downside is that it's Windows software.
I organize everything into Mixxx.
It is tagged not only by genre, crates, some songs belong into more than one crate.
Also by mood, which I tag with colours.
And how much I like the song, which is by stars.
I tried Beets and it was very underwhelming in comparison =)
Correct me if I'm wrong, but different target users here :) Beets seems to be for consumers who want to organize their music collection, and while Mixxx can help with that too, it's mainly for music connoisseurs who play for others (AKA DJs). I'm not sure how well someone looking for the first could use Mixxx, the interface isn't exactly the epitome of user-friendliness if you just wanna organize then press play :)
You are correct in your appreciation.
Mixxx is best used with a DJ console, however for me it is also easily used as a player that automatically picks a song from some selected crates in Auto DJ mode.
So I'm in the mood for Salsa, I put that crate as the AutoDJ source.
Later I can change it to progressive rock, or 90's disco, and it will randomly pick songs from the selected crate or crates.
I can also do DJ things, like looping some sections or mashup some songs.
Anything like this for movies/TV shows/videos? I'm familiar with tinymediamanager, for example, but I'd love something fast on the CLI...
I've tried to use beets a number of times and given up each time.
- beets can't delete transcoded files which have been deleted at source.
- beets can't re transcode existing transcoded files when the source has been modified.
- It's impossible to preserve the original directory structure when transcoding because it strips all path separators for "security".
I use a Bash script I've been maintaining for a number of years instead.
Having hit those first two pain points, I changed my transcode config to 1) run every time, and 2) use a custom bash script as the transcode command. The bash script keeps a plaintext list of files + their modtime and disk size, then only transcodes files it hasn’t seen before. Because it’s a plaintext list and relatively small, there’s not a terrible performance hit since it’s paged into memory.
Not at all saying this to dismiss your criticism though; absolutely would be great to have better OOTB options. Just putting it out for anyone who wants a workaround :).
Consider a case where a DJ's entire show can be downloaded as a single MP3, AND there is a complete playlist for the show.
Is there any software that can use this to break up the single mp3 into the numerous individual song tracks ? And insert metadata ? And perhaps also fingerprint each track to match it to a specific version of the song ?
If the playlist is something like a CUE file then yes, certainly. CUETools/XLD/foobar2000 can split by cue file. And Picard can do audio fingerprints to match tracks and get metadata for them.
In the case I describe there's no CUE files. There's a DJ writing the playlist as the show proceeds, and the typical entry is Artist and Song title. So it's enough information to do a search, but maybe not enough to identify the particular version, and then you have to find the boundary between tracks in the mp3.
So this might be a task for A.I. - to automate it with sufficient precision.
Don't most DJ sets have elaborate cross-fades between tracks? It may not be possible to find a clean cut, and the best you can hope for is good judgement rather than precision.
Point taken! So give me an A.I. that knows when to begin/end a track and when to keep two or three tracks glued together cos the segue is sooo smooooth.....
What I really want is a plugin that starts an icecast server with a given song as a root, and continues to stream songs whose albums/artists are within a certain reference distance of the root on discogs/wikipedia/etc.
So if I start on Guzzlemug I'll get some Blood Incantation, but maybe also some Pink Floyd.
How do you all handle FLAC/CUE releases? A problem I have is that I want to have processed/tagged tracks, but I also want to keep FLAC/CUEs. I end up have redundant and separate organized library and original data.
I, personally, split the flac. It sounds like you don't want to do that but though but it does remain lossless amd I prefer to see the tracks representing individually in the filesystem.
I just split everything into tracks like they are on the album. Is there a tangible benefit to a single large FLAC with a CUE to tell the player where the songs are?
There are a few benefits in my opinion, mainly:
- The CUE maps the audio in a "CD-native" way, like each track can have 1 or more indices, which enables things like pregaps. When you split FLACs into discrete tracks you need to figure out how to project that onto files, do you put the pregaps at the end of previous tracks? At the beginning of next tracks? In short, with a CUE you can emulate the CD experience more accurately.
- If you are seeding a torrent, you need to keep the original data intact.
I've been thinking about this recently too, mostly to figure out how to preserve indexes in tracks of "Murmurs of Earth"[0]. I think the FLAC embedded cuesheets could be used for this. While intended for CD-ready cuesheets embedded into one big FLAC, there's no reason they couldn't be used in individual files, with index 00 for pregaps.
I'm not sure it's well supported anywhere however. Maybe foobar2000?
Edit: I pulled up the RFC[1] for FLAC, and it seems like this is a use case they had in mind for the embedded cuesheet:
> A cuesheet metadata block can be used either [...] or to provide a mechanism to store locations of interest within a FLAC file.
[0]: https://www.discogs.com/release/6057232
[1]: https://www.rfc-editor.org/rfc/rfc9639.html#name-cuesheet
Not sure, I could never get my mind around foobar2000. I've just been playing my single-file CUE/(FLAC|APE)s in VLC ;-(
Auto-tagging is nice to have but it will screw up verification checksums. I would prefer having metadata in a separate DB and keep the files as is, unless I explicitly request exporting the metadata to the original files.
Beets can do this too!
The -C option to avoid copying files and the -W option to prevent writing tags (or set it in the config file)
It will still attempt to match and store the results in its DB for later writing to files, or moving, or querying.
For the data to be most useful, you will need a client that connect to Beets to get the data though, Im not sure how prevelant those are. You can use the web-ui plugin though.
> Beets can do this too!
This is the best summary of beets. It's wildly flexible and configurable.
It's a good tool. However, last I checked, it was not possible to run it in a one-shot stateless fashion, like, passing it a list of music files so it auto-fetches album art, lyrics and updates the very same input files.
> Relatedly, the -q (quiet) option can help with large imports by autotagging without ever bothering to ask for user input. Whenever the normal autotagger mode would ask for confirmation, the quiet mode performs a fallback action that can be configured using the quiet_fallback configuration or --quiet-fallback CLI option. By default it pessimistically skips the file. Alternatively, it can be used as is, by configuring asis.
Beets with this flag works relatively well for me when wrapped in a shell script, but I have spent a lot of effort trying to make it actually non-interactive.
I'm still not totally sure I got all the certainty thresholds right, I still sometimes get unexpected behavior when using Beets this way, so I agree that the non-interactive experience is lacking.
Wow, I remember beets from about 10 years ago. I never got into meticulously organizing my music, but I'm glad to see it's still around. Also the site loaded absolutely ridiculously fast.
Does this have a bulk command to look at the metadata of files in a folder and use those to rename the files? Assume the current filenames are just random strings
Beets crashes with a Musicbrainz connection error more than half the time, on my machine. How does anyone use it productively?
MusicBrainz is currently being hammered by AI scrapers. There's a pull request on the Beets GitHub that seems to address the crashing.
Anyone have suggestions for automating CD rips into a Beets workflow?
Is there a beets LLM version as well, where I can RAG-style search through my music collection?
this is so cool. I am (slowly) working on a tool that uses an LLM to add tags to music based on vibes and it would be awesome to plug into this!
Thanks. I was looking for something like this.
Whatever happened to MusicBrainz Picard?
Nothing - it’s still getting regular updates.
> MusicBrainz Picard
I don't think it can be used as a CLI tool. It's mainly a GUI app. I've tested it in the past and it works fine too.
Been using with Claude via MCP to tag a bunch of old ripped CDs - fun with mixed results.
is Claude "listening" to the music in order to tag it?
No, just uses the mutagen or beets libraries via python
For physical media, Discogs is the way to go...