First Digit Being Removed on "Load Selected Files" losing Disc Numbers from Tracks

I haven’t seen this particular issue in the forum, but after organizing thousands of tracks already, some of my older 90s releases on CD, DVD, & HDD archives that are “organized,” but when I “Load Selected Files,” it strips the leading digits that are the disc numbers.

eg: 1998 2CD Italian techno album called “Dance Attack” has filenames like:

“2 - 13 - Synthex - Come In To My World (Eta Beta J Mix).mp3”

It loads unclustered as:

“- 13 - Synthex - Come In To My World (Eta Beta J Mix).mp3”

Not sure where the leading digit went, and left a leading “ - “ nothing shows in the “Disc No.” column, and without a clean track number and title, adding this release’s 28 tracks to the MBDB will be a pain.

The default naming script has worked well for me so far, so I’m hoping there’s a fix to keep future tagging, saving, and submissions working well.

The workflow is as follows:

  1. Load the files - Picard will group them into clusters based on directories and any other file/tags info it can use as metadata.
  2. Click lookup - Picard will try to match the cluster against a MusicBrainz Release. It may get it right first time - it may not find a match and you have to help - it may find the right MusicBrainz Release Group but choose the wrong Release from within it - or (rarely) it may get it completely wrong.

What your screen shots show is that you have only done the first step, and that the clustering has used the disk number as a guess at the track number, and the track number is still part of the title.

However, whilst this may reduce slightly the chance of Picard getting its MusicBrainz matching correct, in most cases it should still get it right.

So, my advice is to select the Cluster node(s) and click Lookup and see what happens…

1 Like

Your files have no existing tags, hence Picard tries to guess track number and title from the filenames. In the automated ways it tries to find certain numbers. You see how it used the initial number and assumed it to be a track number, then used the rest as the title.

As sophist pointed out above, this might have little relevance to the lookup. You can still try to cluster the files and lookup matches, and then tag your files with the data found on MusicBrainz.

But there are also some things you can do about the guessing:

  1. You can turn off the track number / title guessing in Options > Metadata > Guess track number and title from filename if empty. Then Picard won’t extract the initial number and use the full filename as the fallback title. It won’t change the tagging situation much, though.
  2. Use Tools > Tags From File Names to fill in disc number, track number, artist and title from your file names.

The second option is interesting in your case, as you have very well named files always having the same patter of disc number, track number, artist and title.

Load the files into Picard, select them and open Tools > Tags From File Names. You have to enter a naming pattern, in your case use:

%discnumber% - %tracknumber% - %artist% - %title%

Click “Preview” to see the results, it should show something like the below screenshot:

If you are satisfied, click Ok. This will set the tags for disc and track number, artist and title for the files inside Picard. It will not yet save the tags to the files, for that you have to actually save the files. If you only want to save those initial tags and not move / rename the files at this stage, make sure you only enable “Save Tags” and not “Move Files” and “Rename Files”.

You don’t need to save the tags at this stage, though. Having those tags properly set can also help with clustering and lookup and help you find the proper matching releases on MB.

See also the documentation:

2 Likes

Thanks Philipp @outsidecontext - I had a feeling that there was a flexible file name to tag function, but I don’t use it myself and couldn’t find it when I was writing the original answer and went to look.

Loading files = “unclustered files”; Drag & Drop, no matter if I do all or a group of tracks by disc at a time = “unclustered files”; loading separately still puts everything into the same cluster. which is what the screenshot is showing, the parent folder PFALSE, which became the cluster name.

Regardless, it always strips the leading digits.

Lookup and scan don’t return any results because the release isn’t on MB but is in Discogs. I started creating the release, including adding missing artists’ info during submission, but when I added the 2nd disc/medium, the site glitched, and all the data disappeared, including all 28 manual entries. I’ll try again soon after I give myself time to recover. :sob:

Sweet, I guess “Tag From Filename” doesn’t mess with the default naming script? Is there a way to put this somewhere in the future or just a per selection tag? Either way it’ll come in handy the next time. I was able to tweaking the combinations of spaces dashes and slashes to get things lined up and displayed correctly using he code below which seems to work pretty well:

%album%/%discnumber% - %tracknumber% - %artist%-%title%

Adding a screenshot of the results, should make it much easier for submissions.

Still fails to lookup/scan because of the missing info, but It’ll be much easier to submit as a release now. THx :ok_hand:

1 Like