I think i'm using clusters wrong. Can you check this video and tell me what i'm doing wrong?

Tags: #<Tag:0x00007fd5eced6998>


I’ve been using musicbrainz to organize my files for use on a plex server, but I think i’m spending too much time correcting cluster errors. There must be something i’m not understanding.

I couldn’t think of how to phrase what I was doing and what I expected, so I made a short video showing it all. Hope I don’t get flagged since my first post has a link.

EDIT: Looks like you’l have to make it full screen to see what i’m doing in the video. Embed is pretty small.

I tried to keep it brief. Hopefully i’m missing something simple.



SCAN will do an AcoustiID calculation. And then attempts an AcoustiID based match - which is a headache on a compilation album like your example.

What I do next step is DRAG a box round everything on the RIGHT and drag them back to the LEFT again. So everything is back in the cluster on the left hand side again. And on the right everything is now 0/15 etc in the track counts.

(Or at least drag each of those missing SEVEN files you mention back to the left hand side)

Now you press the LOOKUP button and this time Picard looks up files based on the folder and tag names. This should then pick the correct release. AND it will have the AcoustiIDs in the files still.

If it still fails a match, drag the SEVEN tracks back again. And this time drag and drop the cluster and drop the cluster on the name of the correct Release. This should have more chance of lining up those lost tracks.

Notice that the “Submit AcousticID” button is often lit up after this has happened.

(I’m no expert. This is just how I do it based on methods I have seen before. And this is what works for me)

Oh - and Welcome to the Forum where all the odd puzzles are solved. :slight_smile:


Another trick is when you drag everything back to the left and put it back in the cluster, drag the cluster folder across and drop it onto the correct release on the right. This will also line the tracks up correctly whilst keeping those freshly calculate AcousticIDs.

Having a CD in the drive and doing LOOKUP CD will also help put the correct release on the right hand side to act as a drop target for your cluster.

In Picard OPTIONS… \ Metadata \ Preferred Releases set Picard as to what it favours when doing a match. When matching compilations it is worth tweaking the Compilation slider to the right.


Yes, use Cluster > Lookup instead of Scan