I am curious to know if the Automatic Merge feature is no longer enabled. It seems like I’m using the proper naming convention (as I’ve done for a few years now), but the server is not merging the various versions together.
I can work-around by selecting the versions of the film and performing a manual merge, but this can be less than desirable.
Please also refresh my memory: was there a way to disable automatic merge within a given library? I feel like this was a setting somewhere, but I don’t see anything in the documentation (or reviewing the advanced library settings).
I thought I’d supplement this with an example of the naming convention that I’m following:
Judas Priest Live In London (2002) - H264 QSV 20QP.mkv
Judas Priest Live In London (2002) - H265 QSV 20QP.mkv
Judas Priest Live In London (2002) - x265 MKV 480p30 Preset.mkv
I’ve also tried it with underscores for the text past the dash:
Judas Priest Live In London (2002) - H264_QSV_20QP.mkv
Judas Priest Live In London (2002) - H265_QSV_20QP.mkv
Judas Priest Live In London (2002) - x265_MKV_480p30_Preset.mkv
You could interpret the naming convention instructions to mean the item following the hyphen should be a single text string, but that’s not clear. (This would be a good update to the documentation.)
Judas Priest Live in London (2002)/
Judas Priest Live in London (2002) [H264 QSV 20QP].mkv
Judas Priest Live in London (2002) [H265 QSV 20QP].mkv
Anything in square brackets is invisible to Plex.
In this example above, You see the two files by name you can curate with.
Plex sees two files, same name, different resolutions.
In addition to @ChuckPa’s suggestion, ensure that the items were matched using the same agent. If they were all added at the same time, they would have been. But if they were added in stages, and the library’s agent changed in the interim, then they won’t automatically merge and show as duplicates. In that case you’ll have to manually merge them.
I thought that use of the Plex Agent and the Plex database was a feature of the Plex Pass. I purchased mine years ago, so, admittedly, I’m not very familiar with what’s included and what’s not.
You can check to see what agent was used to match an item by examining its XML media information:
Have a look at the guid. If they were matched using the same agent, and detected as the same movie, they should be identical. If the first part is different (com.plexapp.agents.imdb in my example), they were scanned using different agents; if the back part is different, they were detected as different items:
It doesn’t require Plex Pass to use the Plex Movie agent. We don’t have any special database just for Plex Pass users.
We proxy metadata sites so the sites do not get so much traffic from our users that it takes down their site when a new episode of {insert really popular show here} or {insert really popular movie here} is released
Removed and re-added two of the three files. The GUID is different for each of them. Using that logic, they were detected as separate films. Thanks for giving me another troubleshooting tool.
Again, I have had no cause to change the Agent for the library, so I don’t suspect this is the issue.
I’m sure the marketing team had a hard time putting “proxy metadata” in the features list, but, for sure, that is a very useful feature. I’m not sure if that’s exclusive to Plex Pass, but I do appreciate it.
It’s likely down to naming then. Have you tried the suggestion above, placing the optional information in square brackets? I know that the document you linked above suggests that using a hyphen to separate the optional information should work; however, everything I’ve seen on these forums suggest that optional information should be placed in square brackets, causing it to be ignored. I can’t find any official documentation to support it right now, but it’s worth a shot.
Removed the two additional files, renamed the primary file (so I don’t flush my chapter thumbnails), re-scanned the library, emptied the trash, and re-added the two files with the new naming convention.
I tried the following naming conventions with the same results:
Judas Priest Live In London (2002)[H264 QSV 20QP].mkv
Judas Priest Live In London (2002)[H265_QSV_20QP].mkv
Judas Priest Live In London (2002)[x265 MKV 480p30 Preset].mkv
Judas Priest Live In London (2002) [H264 QSV 20QP].mkv
Judas Priest Live In London (2002) [H265_QSV_20QP].mkv
Judas Priest Live In London (2002) [x265 MKV 480p30 Preset].mkv
Judas Priest Live In London (2002) - [H264 QSV 20QP].mkv
Judas Priest Live In London (2002) - [H265_QSV_20QP].mkv
Judas Priest Live In London (2002) - [x265 MKV 480p30 Preset].mkv
In each case, two new GUID entries were created with the two films been individually added to the library.
The primary file has had a GUID of 13805 whereas the subsequent files have incremented with each removal and re-add (as documented above). They started somewhere around 13813, and we’re up to 13820 & 13821, respectively.
This should indicate the agent is, indeed, treating them like individual entries in the library (rather than the same).
I’ve looked through the Movie Files (Naming and Organizing) documentation on a regular basis, and I’ve never seen anything within the Support Articles that support that bracket convention with respect to the filename parser.
if it does not have a GUID like the one @pshanew mentioned above then it is possible they were not actually matched at all and are just getting a generic one like might happen in a personal media library that does not get online metadata
The key and guid fields seem to indicate the unique identifier for the entry in the Library. Thus, this would be the entry for Plex to identify the file itself.
The example you provided seems to use a Reverse Domain Name convention based on the IMDB entry. Since my Plex Agent is not set to use IMDB as a source, I speculate that convention was thrown out in favor of the numeric one I documented above.
