MusicBrainz keeps closing itself?

I run Ubuntu and I have MusicBrainz PIcard installed. When I scan my tracks, it will close by itself. Any idea why? Also, I would like to know if it is possible to save a list of tracks that are misidentified or corrupt and export it as M3U or something similar so I can see it in Clementine or iTunes. PLEASE HELP! Thank you kindly.


Could you run Picard from a terminal with picard --debug, make it crash and then post the full log output here? That might help identify the reason for the crash.

1 Like

This is the last log on the terminal:
D: 08:07:41,057 /usr/lib/picard/picard/file.update:614: Updating file <MP3File ‘10 Rain Drops - Barry Brown.mp3’>

Please post the full log; not just the last line. Thanks.

I’m trying. There are about 148,000 entries (songs that are scanned) and the term “killed” is the last log I see.

This might be part of the problem, does it happen if you break down this in smaller batches (which is recommended anyway)?


Unfortunately this just indicates Picard is crashing hard, but gives no clue what is actually happening. It might be a bug, it might run out of memory or something else.

In general loading 149k entries at once is really not recommended. That will be barely manageable, both for Picard and for you as a user. Picard is an interactive tool, and you are supposed to check your files and save the changes you are satisfied with.

Also you might want to try the just released Picard 2.8 release candiate 2 and see if it works better for you, see the announcement at Picard 2.8 Release Candidate 2 – MetaBrainz Blog .


No, I’ll try that. Thank you.

I’m in exactly the same boat but put it down to importing a few thousand too many files at once. It’s been okay with around 1000 items at a time for me. Not a solution but might indicate it’s too simply many phat choons to digest in one go. Even Jean-Luc has his limits.

1 Like

I think Spock will tell you it is Illogical to load that many files at one go as you just can’t check them as a simple human. There will be errors…

Personally I think Picard should kick off a loud siren and warning box once the track count is over 200 with an “Are you mad?” warning… :rofl:


Great thing about installing Ubuntu on a Mac is it seems to sleep and wake without much energy loss so I leave mine on for weeks at a time and Picard is always a bit-by-bit save changes then remove them from the screen and check the rest while I’m waiting for another app to do something. Rotation flex. Like growing organic vegetables bro. The conveyor belt of vibes!

Zas, I tried your suggestion and it works. Now its only a matter of putting the flagged tracks (incomplete, corrupt or mislabeled) tracks to a Clementine Player playlist for Ubuntu. If someone who is familiar with Clementine, I would greatly appreciate some assistance. I’m also trying to import iTunes playlists to Clementine but to no avail.

1 Like

So, what is a good file count to add at a time? I have one music folder that has a little over 3K songs and I have trouble with Picard closing down in the middle of making changes to files after Picard has loaded and found the information. Is it possible to speed up Picard so it doesn’t take hours to find the data? I’ve wasted days trying to just have Picard to look up the information to close down without making changes because I haven’t made it through confirming the proposed changes.

I personally would do ~ten albums or so at a time.

Not just for Picard speed but because checking the results if you scan or lookup everything at once gets really messy.


I usually tag max 100 albums at a time, so loading somewhere between 1000 and 2000 files. But I save each album individually after validation.
If your files are already organized somehow it helps, if not, then I would do multiple iterations. Associating correct file to correct release isn’t always easy in this case.

Loading too much files at once is taking a risk (Picard can generate an uncatched exception, or lack of memory, you may make a mistake, etc…).

So, my process is usually the following:

  • select few hundreds of files
  • make backups of those files
  • load them in Picard
  • associated to correct tracks (with the help of Acoustid if needed)
  • save/move along
  • repeat

If your files are already properly tagged and you want to update them (using embedded MBIDs) the process is faster, but I would never load 100k files at once, nor save more than few hundreds of files at once.
Also it should be noted that audio file formats are not all equivalent, saving 5 minutes of 192khz/24 bits FLAC is much longer than 5 minutes of low quality mp3.

Adapt your workflow to your needs, but keep in mind Picard is a powerful human-driven tool, not a magic wand.


Thanks for the information!! I just started using rdswift script he shared with me. I’m still going through just one music folder and the script is rearranging my music which will help in the future. So far, I’ve been doing around 6 artist folders at a time and no problems. Once I apply his script to all three main folders, it will be easier to just select a folder (e.g. A, B, C, etc) to process! Thanks for the help!


My MusicBrainz issue is resolved. How do I put the mislabeled/corrupt tracks that MusicBrainz detects in a playlist in Clementine player?