Confused about 'trackc' and 'discc' fields that are getting added

[chiefy]

I am new to beets, but have it setup and working - somewhat. I did notice that in MyMPD UI all my tracks started showing up with two track numbers as well as counts (for ex: 1,1 vs 1). That led me to see that my files are getting the track, tracknumber, tracktotal and trackc tags, which are the redundant. I don't see a way to strip this tag w/ zero etc.

tl;dr - how do you filter out tags w/ beets, I do not want to have to specify a whitelist

Here's my config...

directory: /mnt/Music
library: ~/.local/beet/musiclibrary.db

import:
  move: yes # Move files to new location
  # copy: no   # Use files in place
  # write: no  # Don't modify tags

plugins:
  - musicbrainz
  #- discogs
  - fetchart
  - importadded
  - lastgenre
  - listenbrainz
  - mpdupdate
  - smartplaylist
  - embedart
 # - zero
  - scrub
  - info

#zero:
#  omit_single_disc: true
#  update_database: true
  
mpd:
  host: xxx
  port: 6600

lastgenre:
  force: yes
  keep_existing: no

importadded:
  preserve_mtimes: yes

#discogs:
#  data_source_mismatch_penalty: 0.8

musicbrainz:
  data_source_mismatch_penalty: 0.3 # Lower penalty = preferred

listenbrainz:
  token: xxxx
  username: xxx

ui:
  color: yes

smartplaylist:
  relative_to: /mnt/Music
  playlist_dir: /mnt/NASConfig/mpd/playlists
  forward_slash: no
  playlists:
    - name: beatles.m3u
      query: "artist:Beatles"

paths:
  default: $albumartist/$album/$track - $title
  singleton: Singletons/$artist - $title
  comp: Various Artists/$album/$track - $title
  albumtype:soundtrack: Soundtracks/$album/$track - $title

[TanbirRamim]

Those extra keys are written on purpose, by the mediafile library beets uses for tagging, not by your config. For FLAC/Ogg it writes the track number as both TRACK and TRACKNUMBER, and the total as TRACKTOTAL, TRACKC and TOTALTRACKS (same for discs: DISCTOTAL, DISCC, TOTALDISCS). Its docs say it accepts any of the keys and writes all of them so players that disagree on the name still find one. That is also why scrub and zero do not help: they work per beets field, and all of those keys are the same field to beets.

There is no option to switch the duplicates off, but a tiny local plugin can drop them right after beets writes a file:

# ~/.config/beets/plugins/dedupetags.py
import mutagen
from beets.plugins import BeetsPlugin

DROP = ("track", "trackc", "totaltracks", "discc", "totaldiscs")

class DedupeTags(BeetsPlugin):
    def __init__(self):
        super().__init__()
        self.register_listener("after_write", self.after_write)

    def after_write(self, item, path):
        f = mutagen.File(path)
        if f is None or f.tags is None or not hasattr(f.tags, "keys"):
            return
        changed = False
        for key in DROP:
            if key in f.tags:  # Vorbis comment keys are case-insensitive
                del f.tags[key]
                changed = True
        if changed:
            f.save()

pluginpath: ~/.config/beets/plugins
plugins: ... dedupetags

That keeps TRACKNUMBER, TRACKTOTAL, DISCNUMBER and DISCTOTAL, which MPD reads. Existing files get cleaned the next time beets writes them, e.g. beet write -f (or beet modify on a query). Test it on a couple of files first.

[chiefy]

Thanks for the details, I did try to search around and didn't find anything. It's strange of all the MPD clients I've tried MyMPD is the only one who seems to have an issue w/ it. I might PR that project.