OK the “duplicate algorithm” needs major improvements. I have a lot of videos which are named very similar e.g. a word plus a counter example: Flex - 001 Flex - 002 up to Flex - 999. So the duplicate algorithm makes random pairs that do not make sense. Why does it pair up Flex - 009 with Flex - 015 (why not with Flex - 010)
Needless to say that these are not duplicates one is 12sec long 1080p resolution the other is 45 minutes long and 480p resolution. they are different in almost all the attributes and not even one of the found duplicates is a real duplicate.
Please either have a switch to turn this feature off or at least take in account more properties than just the name and some attributes form the meta data.
I have over 5000 of these type of titles so don’t tell me to change the meta data in order to have the “duplicate algorithm” not match it. with 5000 titels that is not feasible.
I did once the enormous task to split them apart but as soon as you run “refresh metadata” it will make duplicates again! And guess what - the Flex - 009 hat was previously paired with Flex - 015 is now paired with Flex - 254 so not even the algorithm is consistent.
Are we talking personal videos here or TV Shows/movies?
If it is about Tv Shows or movies it seems like a naming issue to me.
these are all personal videos
Is it in a Personal Videos library? I kind of thought those would never be matched or made “duplicates” or any other file. I’ve only just recently been dipping my toe into it, with some similarly-named generic file names from my drone footage.
have you checked with duplicates filter yet?
