Over the course of multiple years I’ve been able to recompile around 100k songs (and counting). I’ve used iTunes for the longest time since I burn many of these albums into blank CDs. Many of these songs have been imported without adding their song names, artist and album names.
I tried using Picard to organize these music files but I’ve stumbled on multiple problems:
Picard lags like hell when reading all these files. I don’t even get a chance to cluster them and tag them, since the program keeps trying to read all the other files and just kinda lags (until it reads about 15k files then completely locks).
Folders are organized by iTunes (Music > Artist > Album > Song), combined with the fact that I use a small amount of these files daily for playback, I cannot remove the files from the folder hierarchy they’re in. So, Picard is forced to read the whole folder (?), I can’t pick the first 1,000 artists (or so I believe, I know this would be harder but it’s a remedy for it locking up).
So, in a nutshell, I have a bunch of files that have a folder hierarchy and Picard locks up after reading 10-15k songs (I have over 100k).
Other than that, I chose two or three untagged artists’ folders and it matched them perfectly, if I can square up the program for everything else I’ll be very happy.
I also have another problem and it’s that I have a lot of duplicated music (many times the same album three times, sometimes I have the same song but with different or similar tags that are really just the same thing.) Is there a way for me (after tagging everything) find a way to group up all the duplicate songs and delete them? iTunes shows duplicate songs but I’d have to go one by one and delete them (I can’t ctrl+a, del), and honestly it’s something I wouldn’t try. I guess Picard with AcoustID could do a faster/better job than me.
Last but not least, this is my first time using Picard. Consider I’m a complete noob.
I might’ve missed something. Thanks for the help in advance.
There is only one solution for this: Work in smaller batches by only loading a limited number of files at once. This should also work with the iTunes managed folder hierarchy.
IMHO this is also faster for you to work with the files. Loading 100k files into Picard, which maybe roughly translates to 10k albums, will make it hard for you to review the changes written by Picard. It is way easier to work with a limited number of files and verify them then to do this all in one go on this huge amount of files.
You likely will also end up with a couple of hard to identify files or stray files that get matched across different compilations. So best tactic usually is to tag the easy ones first (working in batches) and deal with the difficult ones separately. E.g. it is way easier to drag a file to a different album if there is only a limited number of albums loaded. If you have to scroll around to find some random album in a list of thousands it’s much more time consuming.
This is my main problem, I don’t want to split/move all the albums/artists from the main folder iTunes work them, since I sometimes burn/listen to random songs. There’s also the issue that this would likely take me a long time too (since I have to copy from a folder, to a folder). If I have no solution I might attempt this, but if there’s a better way I’d go for it.
EDIT: Is there a way for me to select say, 1,500 folders from iTunes and add them into Picard, tag that, then select another 1,500, so on so forth? Since iTunes divides as Music>Artist>Album>Song, I’m forced to select the Music folder and it automatically picks everything.
You will upset iTunes by updating the tags anyway… so expect to need to rescan that iTunes folder now and then.
But do look close at your computer. If it is a Windows PC then your iTunes folders ARE sorted by artist.
Something like C:\Users\Aeeshia\Music\iTunes\iTunes Media\ABBA\
Go to Your MUSIC folder, look into the iTunes folder, look into the iTunes Media folder… and eventually you will find all the artist folders.
Now you can select a few artists at a time and hand to Picard.
BUT iTunes is a mess when organising compilations and many albums as it can make some weird folders up.
It also has a habit of breaking albums apart. Very weird program.
But do dig into your folders. Look closer and you’ll find the level where iTunes has all the files. And this is where you can select a few folders at a time and pass into Picard.
Be patient with Picard. Make sure you are using a v2.1 release as this is SOOOOO much faster.
Also remember that Picard and MB is extremely over clever and will know of the dozens of different releases of albums. iTunes won’t care, it just sees some tracks that can be sold to you. So you are moving from a disorgansed shop world to an extremely organised and detailed world instead.
So give things time to get yoru head around this stuff. There is no magic “fix it all” button. but we are also here to help…
I cannot seem to find a way in which Picard allows me to select multiple artists in one go. I am forced to pick one artist at a time with the “choose folder” option, and the “choose files” option allows me to select multiple folders but does nothing other than opening the first folder.
I use the “File Browser” option in Picard. (View in the file menu, then select “File Browser”. Or “CTRL + B”)
I use the option to change the starting folder to keep track of where I left off if I’m re-tagging, or working through a large list. (Rt. click the folder, and select “set as starting directory” on the context menu.)
Something has happened in recent versions that makes dealing with large numbers of files almost impossible.
I used to be able to load my library of about 80K songs in under an hour and work with Picard as things were loading. Now it takes over 24 hours and the performance is so dragged down that I can hardly interact while the loading is going on.
Not sure if it went bad with 2.0 or one of the versions after that, but only solution now does seem to be to deal with small batches.
But, I think it is a bug, or someone changed something about how things are processed without taking scaling to large collections into account.
Picard devs are aware of performance issues with large batches, we already fixed some code that was scaling very bad (like https://tickets.metabrainz.org/browse/PICARD-1339).
We’ll do a complete audit of the code looking for this kind of issue, but it isn’t an easy task.
Meanwhile, and because Picard is not meant to be a magical one-click solution, it is always recommended to proceed in small batches, and carefully review changes.
You don’t need to retag your whole collection every day.
500-1000 tracks per batch is usually more than enough, Picard is doing a lot of queries to rate-limited web services.
It all depends on enabled options and plugins (you may disable plugins to see if one of them is causing problems).
When you report a problem, please include all required informations (Picard version, system, filesystem, format of audio files, etc.), Picard has many options and run on many platforms, we can’t really help without detailed information (see https://picard.musicbrainz.org/docs/troubleshooting/).
I typically load my collection every week or two, not so much for tagging, but to monitor edits that I may have missed, etc. So while I do use it for tagging, I actually use it as a helpful tool as an editor. I am using the latest dev build and still seeing the major performance issue, so perhaps I will try to assemble all the information together for a good bug report.
