RFE: External file for headers and tags

One thing I found is that, with mp4 files, getting some header substitution in place is as simple as modifying the id3 tags that can be found in the file. For instance, you might have The Vvitch, but in the id3 titlesort tag you can call it “Witch, The” to get it to sort in W. There doesn’t seem to be a…simple way to do this, so how about an extra text file that could be stuck in with the file that can be read? This can be any format, but simple flat text should be fine - so one here owuld have this:

title=“The Vvitch”
titlesort=“Witch, The”
album=“Horror” # and of course, you can stick comments in like this

I expect this is not a simple matter of programming; you would have to tell the media scanner how to check for an external text file if relevant headers cannot be found in the file, presuming that no such routine exists, and it would be a separate operation from scanning the actual file for those headers.

Can you clarify what you’re trying to accomplish, or what obstacles you’re encountering?

I would suggest ignoring MP4 tags. It’s usually simplest to let Plex match based on filename and pull metadata from the cloud.

Plex allows you to edit a movie’s Title and Sort Title. If you set the Title to The Vvitch and Sort Title to Witch I believe it will behave exactly how you describe.

What purpose would external files serve?

There is an existing request for this feature. Rather than splitting votes over several threads, it’s best to add your vote and comments there.

1 Like

I will absolutely clarify, Volts. =)

tl;dr: the web UI is a bit clunky and I would rather have at least some metadata ready to go as I load the file. Also, this allows the same functionality for .mkv files.

The usual method I tend to use is this - when I am uploading a .mp4 to my server, have metadata already on board; Plex picks up on tags within the file, as I’ve found that using ‘titlesort’ tag - and the title tag too. Yes, it can pick up the filename, but to this end it’s not really something I’m worried about. Likewise, I can use the ‘album’ tag if I want to add it to a collection - so f’rinstance, I could put all of the works of Studio Ghibli in such a collection really quickly. Since I batch upload overnight (I have my reasons), this is the method I’ve chosen.

(Of special note: Sort Title is analogous to the ‘titlesort’ tag. Likewise, one can add an ‘album’ tag to the file and this sticks it in that named Collection. If it’s just one or two files, or several files (say, DVD special media), then it’s easy enough to locate the files, select them, and then do a bulk edit - but I find this to be a bit faster.)

The problem here is that .mp4 only supports one ‘album’ tag - so in order to fix that, I can stick a new file in a single collection before uploading - and then add it to other collections after it’s uploaded.

With .mkv files (which I use for HD stuff, since compressing to .mp4 tends to wipe out the subtitles in BD pulls), the tags are…well, they’re not intuitive to me. I can conceivably add them in mkvtoolnixgui, but the tagging scheme may as well be in anglicized Urdu from my perspective - and my google skills may not be enough to find the document that better describes these tags, if such exists.

So why not make this such that we can just use a simple text file for the not-so-automatic metadata, following a universally created schema, that would do that same job insofar as metadata goes? That’s the idea I have. Stick the tags in a text file, and once Plex picks up the new video (or new metadata file) and parses that for additional data.

There are a couple reasons for this one. One, of course, is mentioned - I can’t tag .mkv files effectively, and there are a few limits in .mp4 files. Second is that this would give server operators the ability - should they so choose - to add their own metadata in a more universal fashion. Third is because without ‘titlesort’ usage being strictly held to, Plex does not always seem to sort correctly - so The Vvitch might wind up in V, W, or T depending on data available, and likewise The Blob could wind up under B or T. Third is to hopefully aid in tagging using cloud data - one upload I’ve had for SW Ep 4 wound up tagging as a coming of age film called “New Hope” rather than the space opera epic Star Wars Ep 4 A New Hope until I manually intervened; by getting this down, I don’t see so much misattribution (and the improvements to the search that Plex executes helps quite a bit, but there’s still some weirdness once in a blue moon).

The only downshot I can see is that it involves adding an extra file into the equation and one must use a text editor to do this. I don’t see much wrong with it - as a Linux guy my config file skills are at least reasonable to decent. Also, as you yourself point out, filenames tend to do the job pretty nicely. So the most this would be is an option for people like me who would like to see it and don’t mind editing text files such that they contain variables.

…today, I learned. Alright, we can close this request then - I’ll add it to the one you point out. Thanks, beckfield!