I also use MusicBrainz Picard so I took your query a little further JustinChase.
I have quite a few albums where there is one main Album Artist, with a few tracks that are either by other Artists or include other Artists as well as the main Album Artist. The [Artist] field being a list type field. In those cases I want the [Album Artist (auto)] field to stay the same as the main Artist for all tracks in the Album. If I delete the [Album Artist] tag on such Albums, the [Album Artist (auto)] correctly changes to (Multiple Artists).
So when I find a track that has duplicate [Artist] and [Album Artist] values, I can't decide whether to delete the [Album Artist] tag or not unless I can see all the tracks for the Album. Hence I made two changes to the Smartlist.
1. I add the "~a" modifier to the rules to append all tracks from any Album where one track has been selected.
2. In the resulting view I set the "Group By" setting to "Album".
I can then see the whole Album with all tracks and decide if I want to change what MusicBrainz Picard has done or not.
The new code is;
[Media Type]=[Audio] -[Album Artist]=[] ~sort=[Album Artist],[Media Type],[Album Artist (auto)],[Album],[Disc #],[Track #],[Artist] ~matches=[Album Artist],[Artist] ~a
Note that in this and the previous version I left the sort parameters set at the default for a new audio Smartlist.
EDIT: I had a bit of a go with the ~dup command as well, and just tried your new version Marko. I'm still getting some strange results. Sorting by Artist helps to see what MC is doing, but I am getting some results that are showing partial matches between the [Artist] and [Album Artist] tags. i.e.
[Artist] [Album Artist]
2Pac feat. Danny Boy 2Pac
801 Live 801
Also seeing
2 Sinners (Multiple Artists)
50 Cent Original Soundtrack
AFI A.F.I.
and so on. That was before I added the ~a function to expand to whole Albums, or Grouped By [Album]. It seems that ~dup is doing a bit of a fuzzy match, which I think it is supposed to, while ~matches does an exact match, which is probably better in this case.
Thoughts?