It is in the provider settings, but I just realized not for Subsonic-providers. Kodi and everything parsed directly by Symfonium supports it.
In the Now Playing screen, the cover for the current track is shown then. On the album level it is still one cover (the first) for the album though.
You cannot add per-track covers in Picard AFAIK, but it is possible with more direct editors like kid3, where you operate on the file level.
@Tolriq, are there any plans for track covers for Open Subsonic?
A subsonic server can already set a cover art for each track. If the server supports it, it should return an image id that points to the embedded cover (thatâs the case for LMS)
I tried that when I started ripping my fathers CD collection but it did not play nice with using multiple drives at the same time. Loved to lock up the entire PC while trying to re-read one bloody sector. Thatâs why I ripped with dbpoweramp CD-Ripper (in the end with 9 drives at the same time).
When I started ripping CDs, our internet was utter garbage (2MBit/s shared with 3 others), so looking for covers online was no option. Scanning them was painfully slow so I set up my dslr, 2 studio flashes, a remote trigger and mounted the dslr upside down on a tripod perpendicular to the covers (used a mirror on the table to get that right easily). I even used a color checker to get white balance, color and contrast right and consistent and adjust for the quirks of the sensor.
I even went through the madness of using the perspective crop feature in photoshop to crop the covers exactly right at 400% zoom. Thinking about how AI could probably do this these days in a couple of minutes hurts more than Iâd like. As a result I often have way better covers than the streaming services or any other source available online. I probably shouldâve uploaded them to musicbrainz but since I named them sequentially that train has pretty much sailed unless I want to check each album I have to see if itâs one I shot or one I downloaded.
Never heard of it. For quick 1-offs I use TeraCopy (with verfication on). TeraCopy is sadly still single-threaded. If I want maximum speed I use FastCopy, which is multi-threaded and goes full throttle (also with verification on).
When I have a copy action thatâs repeated a lot I use SyncBack.
Iâm sadly also a very big movie and tv show fan, which eat even more space but at least donât have as complicated metadata. Unless youâre an autistic fudge like me who wants the tracks of each file in the same order, with the same names and correct flagsâŚ
My main covers average 1.707MB (mainly because of the ones I shot with the dslr), so for your 60k tracks this would mean 102,4GB of embedded covers. Compared to (conservatively assuming each album has 10 tracks) 10,24GB when not embedding them. If I had all my (main) covers embedded Iâd be âwastingâ well over a TB so the decision was a nobrainer for me.
Iâd like that. Finding someone who doesnât recoil when you start talking about correct metadata, file structures or regex is so incredibly rare.
*tripled
OK, you are outright crazy. I will be officially worried if we share the same cell in hell.
I am limited by the RPi 4 USB I/O and after that by a not-more-than-ok wifi, so if I reach 300 kbps I am more than happy and that can be done single-threaded.
I like movies and series, too, but with a different dedication. I have a moderately big shelf for DVDs/BDs, but I never felt the urge to catalog those more than in a spreadsheet; and even that is mostly obsolete, as I add everything to Kodi, even the hardware discs as media stubs, and I am mostly fine with the scraped data.
PS: I thought for about five minutes about cataloging my books and created accounts at goodreads and thestorygraph. But it is not likely to happen⌠if I ever find spare time, I might actually read them.
Thanks again guys, iâm about to build my own workflow based on your recommandations 
One question though, i donât think that you mentioned the âGet Lyricsâ part, right?
I would like to get the Synced Lyrics when available, if not basic Lyrics (but .lrc files anyway)
I know that Media Human Lyrics Finder is often advised, so my questions for you:
I use tinymediamanager to manage films and series and am happy with the free version of it.
I do not use lyrics so I cannot help here.
I use this tool for searching covers:
I has an scripting engine to create your own searches.
Never heard of that. I use Lyrics Reloaded in MusicBee to fetch synced/unsynced lyrics during playback. If they are correct, I can just save them to a .lrc/.txt file. If there are no synced lyrics, I sometimes sync them myself (itâs possible directly in MusicBee).
However there are also solutions like lrcget which has a massive database and can yield good results. But I dislike their chosen formatting [mm:ss.xx] text so I donât use it to sync lyrics. Only to fetch them (then I run my own script lyrict on the resulting .lrc files with the --m tag_external and --standardize force.xx arguments to at least get rid of the space and store them as [mm:ss.xx]Text while also adding related metadata from mp3/flac files of the same name as tags like described here.
There are other solutions for fetching lyrics as well. Foobar 2000 for example has Lyric Show Panel 3 and Open Lyrics.
All yielding different results from different sources, using different formatting.
Same, but I use the paid version. Having my movies and tv shows look the same (with minor differences) in Kodi, Plex and Jellyfin with full control over every single image while only managing it in one piece of software was worth it for me. I just wish their bloody UI supported more hotkeys. I literally created a âclickingâ macro when I rescraped the data for my entire movie library last year.
You can use âSearch & scrape selected movie(s) - force best matchâ instead of validating each scrape with a click.
@655321 do you have the same type of thing for Artistâs images by any chance??
You underestimate my autism. I not only select the correct match. I choose every.single.bloody.image manually. I used to let tmm scrape the art automatically, but having a wild mix of different languages (polish season banner, german poster, english fanart) as well as having duplicate extrafanart pissed me off enough to go the fully manual way for every one of my thousands of movies. Hence why I desperately craved hotkeys for selecting and confirming art.
Now each movie has the correct art for the language it was made in. French movie: french art. German movie: german art and so on (with exceptions for languages with non-latin script where I defaulted to english) and also a time-appropriate poster (1940 movie with a modernised poster also pissed me off).
It was a mind-numbing amount of work but Iâm happy with the result and from now on itâs easy to maintain.
Nope. I havenât bulk added artist images to my library yet. I tried getting the images from lastfm but noticed that they love to serve images with the wrong filetype, so if you decide to get images from lastfm, better check in a hex editor that they are what they claim to be.
ive installed it, thank you.
How do you launch the plugin exactly?
Can you also let me know how please?
You donât. If you have unpacked it in the âAppDataâ folder, it should be loaded into MusicBee when you launch it. You can check by going to Edit â Edit Preferences â Plugins
Then you can select the order of lyrics sources under Tags (2) â auto-tagging â lyrics.
When you have found correct lyrics, you simply double click the lyrics panel (or use right click â Edit Displayed Lyrics) and then press CTRL + T just before the current line is sung, which inserts the current timestamp at the start of the current line and jumps to the next line. Repeat this for the entire song and save the result with STRG + S or right click â Save Lyrics to External File. Under Tags (1) I have set save as an external file <music filename>.txt in the lyrics storage settings since I prefer external lyrics.
Thanks for this @655321 , I will definitely give it a try! 
Stumbling across this post in 2025, and, I must say its nice to meet fellow nerds like myself who take the time to tag and catalogue music correctly. Although, I must admit some users here are on an upper spectrum when it comes to this process, and, I say that without mal intentions but rather utmost respect 
Iâm curious, why has nobody use beets? I get it if the whole CLI thing might be to much for some to bear, though that can be simplified rather quickly with aliases etc. I find its a really great âall roundâ tool that grabs metadata from various sources, has multiple useful plugins, for example fetch lyrics which was mentioned here.
I did use it briefly years ago but found a few limitations. At the time, I wasnât able to get beets to properly copy/move related files (cover, .lrc, .cue, .log) when automatically moving the songs to a new folder structure based on the musicbrainz tags. There are a few plugins that allow to move such files alongside the tracks now (I was active in at least 2 threads if memory serves), but nothing that did what I usually use in Mp3tag.
In Mp3tag, by altering the _DIRECTORY field, I move all files and subfolders (cover, log, cue, artwork subfolder) of each source folder to each target folder. Mp3tag also automatically renames .lrc files alongside their corresponding songs, ensuring that the connection between lyrics and audio track is not broken.
The downside is that you have to ensure that each source folder only contains the files you wish to have in the target folder. For me this is still way quicker than only moving the songs to new folders and then cleaning up the other files manually.
It also depends on what kind of music you collect. If you only have popular artists and only collect complete albums, using beets is a no-brainer and probably the fastest solution. When you go for lesser known artists or specific editions (japan edition, deluxe edtition, huge collections, compilations etc.) the perfect results go down and manual labor is often needed. And when (like me) you also process a lot of bootlegs/concert recordings that are simply not present on MusicBrainz and require a lot of manual processing (they usually donât have any or only rudimentary tags while a text file holds the relevant information) I found it way simpler to edit such files in Mp3tag (extracting the info from the text file etc.).
All in all it comes down to precision and speed for me and while beets offers perfect results and speed for a good chunk of popular releases, for me there are too many outliers that I have to edit in Mp3tag anyhow. Which led to me simply using Mp3tag for everything as itâs marginally slower when used with plugins like the MusicBrainz expanded websource I linked to earlier in this thread for âeasyâ releases.
Im not sure about that either, because I let beets pull all info I need while doing the initial import. When I add new music I add it to an âuntaggedâ directory on my system, then I import the album/single/mix into my Main Directory and let beets grab all addtional meta in the process (cover, lyrics, add replaygain etc)
I get what you mean, beets will resort to using what is available from configured metadata sources (Musicbrainz, discogs, etc). I would always suggest adding all releases to the Musicbrainz Database if no match is found, as it benefits everybody
But like you say, it really does come down to personal preference.
Like you said though it comes down to personal preference, goals, and, how you prefer to achieve the result.
Personally, beets maks the process so flawless for me. If a realease cannot be found automatically, I can also add âsearch-id to the command to match manually. The list of plugins is also really great.
