Print

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

[GHammer]

I thought something was odd before but didn't have time to track it.
Here is a smartlist that I have used for most of the 11.1 run. It no longer works. I get frequent plays (supposed to be excluded if played within 10 days) and multiple entries of the same selection.

Also, I can refresh the list and get the same selections in a different order.

Here's the code and an example result:

Code: [Select]

[Media Type]=[Audio] -rating=1 ~sort=[Random] ~mix=30,80%,{-[Rating]=> [3] [Last Played]=>10d ~limit=-1,1,[Artist]},20%,{[Rating]=> [4] ~limit=-1,1,[Artist]}

Code: [Select]

U2 - 03 - I Will Follow.mp3
Badly Drawn Boy - 03 - Something to Talk About.mp3
Diana Krall - 03 - I Love Being Here With You.mp3
Tears for Fears - 03 - Everybody Wants to Rule the World.ape
Sting - 03 - Russians.ape
Dido - 07 - Who Makes You Feel.ape
A Hard Day's Night [UK]\The Beatles - 03 - If I Fell.mp3
ACDC - 03 - Big Balls.mp3
Pink Floyd - 03 - Time.ape
Diana Krall - 03 - I Love Being Here With You.mp3
Pink Floyd - 03 - Time.ape
Badly Drawn Boy - 03 - Something to Talk About.mp3
Pink Floyd - 03 - Time.ape
Yes - 04 - America (Bonus Track - Single Version).ape
ACDC - 03 - Big Balls.mp3
The Beatles - 03 - If I Fell.mp3
Sting - 03 - Russians.ape
U2 - 03 - I Will Follow.mp3
ACDC - 03 - Big Balls.mp3
Diana Krall - 03 - I Love Being Here With You.mp3
The Beatles - 03 - If I Fell.mp3
U2 - 03 - I Will Follow.mp3
Dillard-Hartford-Dillard - 03 - Two Hits and the Joint Turned Brown.mp3
Badly Drawn Boy - 03 - Something to Talk About.mp3
Dire Straits - 01 - Telegraph Road.ape
Sting - 03 - Russians.ape
Dillard-Hartford-Dillard - 03 - Two Hits and the Joint Turned Brown.mp3
The Beatles - 01 - Back in the U.S.S.R..ape
Paul McCartney & Wings - 10 - Nineteen Hundred and Eighty Five.ape
Dillard-Hartford-Dillard - 03 - Two Hits and the Joint Turned Brown.mp3

[marko]

reading your string through, the first thing I notice is spaces in the [rating] rules.

not rating=1 is fine.

these are not:
-[Rating]=> [3]
[Rating]=> [4]

try replacing those with
-[Rating]=>[3]
[Rating]=>[4]

tip:
-r=>3
r=>4

do the same thing but are quicker to type. does that fix things?

[marko]

also, it's more complicated than it needs to be:

-rating=1 followed later by -rating=>3 is the same as r=2,3
[rating]=>4 is the same as r=5
~sort=random is not needed because ~mix takes care of that by default.

Simplifying it here, I noticed that you also don't have recent plays excluded from the 20% of five rated tracks, which could explain the recent plays you've been getting. The following string should return desired lists:

Code: [Select]

[Media Type]=[Audio] ~mix=30,80%,{r=2,3 [Last Played]=>10d ~limit=-1,1,[Artist]},20%,{r=5 [Last Played]=>10d ~limit=-1,1,[Artist]}
Is that any better now?

Re. dupes
Another thing to remember when using ~mix:

What does MC do if, say, the r=5 part of the mix can't be satisfied?
In your example, MC needs to find 6 tracks rated 5, not played for more than 10 days, and all by different artists. If you just put those requirements into the search bar, ie:

r=5 [Last Played]=>10d ~limit=-1,1,[Artist]

and MC only returned, say, 4 tracks, then that will not be enough tracks to satisfy the ~mix rule.
I've found that in cases like that, ~mix takes priority and MC will pad the list with duplicates to satisfy the ~mix,total_number_of_tracks, rule.

this can result in some peculiar things on show in the file lists...



possible this is where your dupes are coming from?

[GHammer]

Hi Marko!

I have no idea where the spaces came from, but removnig them does work. I recently restored a library but have never had smartlists change. I'll check that later.

As for the ratings, well...

Because these are in a mix and in brackets, r does not work. Or it may work and the entire keyname is working differently.

If I use [rating]=>4 I get selections rated both 4 and 5. And that is what I want 20% rated equal or greater than 4. [r]=>4 results in no tracks rated 4 or 5

Actually, there are no tracks rated 2 or 3. I have 1s which are things I don't want to hear right now, maybe later. Or I'd delete them.

I have tracks rated 4 which means I like it but it is not an all time favorite. Those are 5s.

The dupes were only on the unrated tracks. Also, the last played does not affect 4s and 5s.

Hmmm, sounds complicated doesn't it?

In English:
Get 30 tracks
No tracks should have a rating of 1
20% should be rated 4 or 5
80% should not be rated 4 or 5 (Remember I have no 2 or 3 ratings) and should not have been played in the past 10 days.
Randomize the list

The last code you posted results in only a list of rated 5 tracks here.
This works as described above:

Code: [Select]

[Media Type]=[Audio] -r=1 ~sort=[Random] ~mix=30,80%,{-[Rating]=>[3] [Last Played]=>10d ~limit=-1,1,[Artist]},20%,{[Rating]=>[4] ~limit=-1,1,[Artist]}
Good catch on the spaces, I had looked and looked and the library restore had the same so wasn't helpful.

Cheers!

[marko]

thanks for the rating description. now I know that, this should work for you...

Code: [Select]

[Media Type]=[Audio] ~mix=30,80%,{r=[] [Last Played]=>10d ~limit=-1,1,[Artist]},20%,{r=4,5 ~limit=-1,1,[Artist]}
interesting thing with the rating syntax. (I have tracks rated 1-5)

using the string above works fine for me so I did some tests using a simple rating search.

[rating]=[5]
[rating]=5
r=5
r=[5]
[rating]=>4
r=>4
all return the same 242 5 rated tracks I have here.

[rating]=>[4]
r=>[4]
-r=[]
-[rating]=[]
all return the same tracks, ie, all tracks with a rating.

this means that these parts of your string, while doing what you want, are not doing so in the way you expect:
-[Rating]=>[3] is returning every track without a rating
[Rating]=>[4] is returning every track that has a rating, bearing in mind you excluded the 1 rated tracks earlier in the search.

I would suggest using plain numbers when searching by rating. I always have done and have never had any problem with it.
It's not like [track #], where a search for [track #]=1 would return tracks 01, 10, 11, 21, etc. etc. as there can only be values of 1 to 5.

try my revised search, you should find it returns the desired results, and for all the right reasons!!

[GHammer]

Ah, I see now.
Helps to have the entire range of rating possibilities available.
I wondered why the ratings worked at all in my version. But it is clear that one was selecting non-rated and the other since it found a valid rating would select from them.

Yes, your posted code creates the desired list. I will use the entire number in the future. Not only does it work, but is easier to determine what is going on.

Thanks again Marko!