Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[Spike1000]

MC 24.0.41

In a nutshell despite me adding a Modifier in Play Doctor Options of 'Remove Duplicates' of 'Name' Play Doctor still lists many duplicates of the same track. In an example playlist of 159 tracks 1 track is listed 22 times, another 18 times and another 21 times. These 61 tracks are from all from the same artist and I used to 'seed' the Play Doctor playlist with a single track.

There are no other duplicated tracks from any other artist. . . I wonder if the 'Remove Duplicates' modifier is not being applied to the 'seed' Artist?

Are there any ways to fix this as it really is quite annoying 

Spike

[Spike1000]

Polite Bump.

Is this something somebody could take a look at? Could someone else try this as a test? Not sure if it's me or MC but it's spoiling a really neat feature of MC.

TL;DR
I'm getting a massive percentage of duplicate tracks when seeding Play Doctor with a single track despite having a filter to remove duplicate named tracks. The duplicates are all from the same artist and album used to seed Play Doctor.

Spike

[JimH]

Try using a seed with more variety, more tracks.

[Spike1000]

It sort of gets better when multiple tracks are used as a seed, but it's still not respecting the remove duplicates filter. And it still can go very screwy. . .

A little test:
1 track as a seed - up to 7 duplicates (of a given track) in the playing now list
2 tracks as a seed - up to 3 duplicates (of a given track) in the playing now list
3 tracks as a seed - up to 2 duplicates (of a given track) in the playing now list
4 tracks as a seed - 27 duplicates (of a given track) in the playing now list!!!  ?

All the tracks were from the same album in this case.

If i need to start dotting around and selecting multiple tracks by multiple artists from multiple albums just to seed Play Doctor I'm getting close to building they playlist myself  , I'd much rather Play Doctor did that for me and eliminate my cognitive bias The results as they stand are excellent; they are just spoilt by the duplicate entries.

The problem seems to be a Play Doctor pipeline issue in so much that the filtering is being ignored (or applied at the wrong stage) as it's allowing duplicates into the list. Any chance of this being investigated?

Neil

[SkGe]

@Spike1000
So i made a short test with Play Doctor (i don't use that often)
I can summarize here my end of testing.

Default items #100; less variety

Test 1
remove duplicates - filekey()
Seed Song - Remember When
Song Name x2 duplicated but different Artist
Artist Name x2 duplicated but different Artist
Album name x2 duplicated but same Artist

Test 2
remove duplicates -  name; artist
Seed Song -  High Horse by Kacey Musgrave
1 song x2 duplicated as the same name and artist from the seed
1 song x2 duplicated as the same name and artist from the seed
1 song x2 duplicated as the same name and artist from the seed
Artist Name x2 duplicated x10
Artist Name x3 duplicated x3
Artist Name x4 duplicated x2
Artist Name x5 duplicated x2
Artist Name as seed duplicated x17 times

Test 3
remove duplicates - name; artist
last played - <10mins
1 song used x5 times but on different artist
5 Artist x2 duplicated

Test 4
remove duplicates - filekey(); name
last played - <10mins
Seed Artist  - Echosmith
Artist duplicate
1 Artist x2
2 Aritst x2
3 Artist x3 (the seed artist)
4 Artist x3
Tracks -  no duplicates
As you can see the variable is changing all the time when you add different attributes. But the only Playlist test 2 suffer the most in therms of duplicates. I can't say how the continuity of the playlist stay after the 100 items added, but presumbly they follow the rules applied at the beginning.
Hope this can give you some help.

[Spike1000]

Thank you for taking the effort to do the experiment. Seems like there's not a workaround and there's an issue with Play Doctor itself  Hopefully this will be addressed soon.

By the way how did you get the filekey() modifier to work? I couldn't find the right syntax as it kept deleting that bit from my expression?

Cheers

Spike

[SkGe]

Filekey() is an internal function where it can give for each track a unique identity so you can use that as a standard of your playlist in Play Doctor.
For using that function you can add a new field and put the filekey() inside and you gone have that identifier as key per track. Because it has a unique key it can avoid mostly duplicates on playlist, but now depend on your source library how it is tagged and how many variable you can have. Just have some test with this method and see the result.

[Spike1000]

Ah ha. The key thing for me was adding a 'new field'; I assumed it already existed somewhere. Many thanks for taking the time to reply.

Spike