Skip to main content

Notice

Please note that most of the software linked on this forum is likely to be safe to use. If you are unsure, feel free to ask in the relevant topics, or send a private message to an administrator or moderator. To help curb the problems of false positives, or in the event that you do find actual malware, you can contribute through the article linked here.
Topic: Foobar2000 Wrapped (Read 5405 times) previous topic - next topic
0 Members and 1 Guest are viewing this topic.

Re: Foobar2000 Wrapped

Reply #25
That's probably the problem. The version I put at the readme works. But it's 64 bit.

Have you tried this? 21.6 should be more similar to 22.7 than the one you got xd
https://www.npackd.org/p/org.miktex.MiKTeX/21.6

EDIT: you can also try any other LaTeX suite, is just that I recommend that one because I have tested it and works from win 11 to win7 in 64 bits oses. But feel free to try other latex installations (and adjust the CMD at the settings if needed). Can not test LaTeX on Wine 32 bits, sorry. And again, nothing stops you to manually compile it with any LaTeX editor (or wait for my html version).

Great sucess!!!!

the version you pointed out works

[attach type=thumb]29057[/attach]

Various artists have a huge advantage, maybe could be interesting more focus in the artist tag.....
That's what testing is for hahahaha

I don't use a "various artists" value as an artist, but it's just a matter of skipping it. I can add a filter for specific artist like this then.

Will add that version to the readme for 32 bit builds too.

EDIT: going to update all repositories now (with the fixes from this thread), I would appreciate if you update them and report problems you found on other scripts in their threads. In case they persist.

Re: Foobar2000 Wrapped

Reply #26
That's what testing is for hahahaha

I don't use a "various artists" value as an artist, but it's just a matter of skipping it. I can add a filter for specific artist like this then.

Will add that version to the readme for 32 bit builds too.

EDIT: going to update all repositories now (with the fixes from this thread), I would appreciate if you update them and report problems you found on other scripts in their threads. In case they persist.

Various artists are from the album artist tag, in my tags there is no single various in artist. Is very noticeable because this little time of 2024 I listen to a lot of BSO and label compilation and the artists was very varied, but the album artists are always various artists and soundtracks or label name as grouping. Only Angelo Badalamenti can enter the top, thanks to David Lynch's films.

I'll give it a try tomorrow at the new builds. Hope that it will include the artist instead of album artists to see the difference

Re: Foobar2000 Wrapped

Reply #27
That's what testing is for hahahaha

I don't use a "various artists" value as an artist, but it's just a matter of skipping it. I can add a filter for specific artist like this then.

Will add that version to the readme for 32 bit builds too.

EDIT: going to update all repositories now (with the fixes from this thread), I would appreciate if you update them and report problems you found on other scripts in their threads. In case they persist.

Various artists are from the album artist tag, in my tags there is no single various in artist. Is very noticeable because this little time of 2024 I listen to a lot of BSO and label compilation and the artists was very varied, but the album artists are always various artists and soundtracks or label name as grouping. Only Angelo Badalamenti can enter the top, thanks to David Lynch's films.

I'll give it a try tomorrow at the new builds. Hope that it will include the artist instead of album artists to see the difference
No, but there will be the possibility to remap artist to %ARTIST% instead of %ALBUM ARTIST%. I can not offer a single setting for everybody, I don't use album artist tag to group various artists albums, and just substituting as you said would break the scrip for others.

(I prefer to use TF to group various artists albums dynamically on the album list)

Re: Foobar2000 Wrapped

Reply #28

No, but there will be the possibility to remap artist to %ARTIST% instead of %ALBUM ARTIST%. I can not offer a single setting for everybody, I don't use album artist tag to group various artists albums, and just substituting as you said would break the scrip for others.

(I prefer to use TF to group various artists albums dynamically on the album list)

I use this scheme for albums with a lot of artists:

Grouping ----> soundtracks, label of compilation, Remixer, artist that choose the tracks, etc
Album artists ----> Various Artists or whatever comes from discogs or musicbrainz tagger. Very popular, all "internet" album comes with it :-D
Artist ----> single value tag with full artist name, for example, XXX vs. YYY feat. ZZZ. Useful for internet searches, last.fm scrobbling, etc. To avoid scripts and just use %artist%
Artists ----> multi value of artists, for example XXX; YYY.
Featuring ----> multivalue of featuring artists, with artists, presenting, etc of single track
Remixed by----> multivalue of remixer artists of single track

Grouping only is present when necessary and set manually. The rest are as come from taggers or processed from. I trend to avoid manual edit as much as I can. :-P

This way I can make grouping in playlist for all album and compilation remixes from one artist, although some of them have a lot of different artists. Or grouping under one set artists that use different artists names along his career, etc. It is basically a last resource tag.

With this it is easy to make playlists well organized with very little scripting and have a file system structure easy to maintain clean for other programs like kodi or UPNP servers.

Re: Foobar2000 Wrapped

Reply #29
Tried new build, no idea if it is because of external tools, but it is a lot faster.

Besides the various artists Imperator :-P, I found a problem with title tags like "untitled".

I made one report from 2023 and the most listened track was "untitled" with 29, that seems to integrate all untitled songs from every artist.

This happens as well with typical title like "movement",  "part", etc and in general with all songs with the same title from different artists.

All my 5 favourites songs are like that, tracks from different artists with the same title. All of them of course are not the most played or loved song from the year.

untitled, movement, burn and voices was the chosen ones :-P

Also, will be cool to integrate the %STYLE% tag. It could give more interesting results than the more general %GENRE%. Electronic and rock are first, but it is said not too much about the real habbit, it is more relevant to know if there are more IDM, techno, house, ambient, noise, industrial, alternative rock, blues rock, post-rock, hard rock etc.

In summary, petition to make the search camps configurable, like the other scripts that read from a .json in js_data, and a link between artist and song to avoid duplicates in common title names

Edit: the playlists that create are very accurate, In the top favourite songs 2024 only 19/100 comes from Various and the number of reproductions is more similar to a playlist made with most paled iof 2024 than the results in the resume

Re: Foobar2000 Wrapped

Reply #30
The suggested artists have a lot of duplicates. Removing duplicates with playlist tools only leaves 38 songs of 100

Re: Foobar2000 Wrapped

Reply #31
Genre can already be remapped to style, so I don't understand your request.

I want to be able to merge multiple tags (genre + style) but that requires a rewrite of the statistics framework which was initially thought for one tag per axis (timeline script).
EDIT: will be possible as long as you separate tags by ', ': '%GENRE%, %STYLE%'

Already noticed the problem with title duplication for different artists, which again requires a rewrite.

About your last concern, what do you mean by duplicates? Show an example. The online version just uses whatever ListenBrainz reports.

Re: Foobar2000 Wrapped

Reply #32
Genre can already be remapped to style, so I don't understand your request.

I want to be able to merge multiple tags (genre + style) but that requires a rewrite of the statistics framework which was initially thought for one tag per axis (timeline script).
EDIT: will be possible as long as you separate tags by ', ': '%GENRE%, %STYLE%'

Already noticed the problem with title duplication for different artists, which again requires a rewrite.

About your last concern, what do you mean by duplicates? Show an example. The online version just uses whatever ListenBrainz reports.

Shift+left click, I didn't notice, sorry..

I changed to ARTISTS (my multivalue artist tag) and STYLE in the console there is a message that said "getEntitiesByTag: 400 {"code":400,"error":"multiple tags are specified but the condition param is missing"}" . Not sure if it was there before but the pdf is produced and the playlist too.

By duplicates I mean 4 times the exact same song.

Like that

X

X

X

The playlist have 90 items, after removing the duplicates 31. Is only in suggested artists, the rest of playlist seems fine, in fact very good.

The results are much better changing the tags.

X

This is more representative of what I listened the last month and a half

Edit: Out of context question, how do you put the description when you insert a thumbnail in the forum?

Edit 2: Is so good that it predict the song of March XD. I didn't notice that, in the artist graph, some have highlighted March instead of February


Re: Foobar2000 Wrapped

Reply #33
Quote
Edit 2: Is so good that it predict the song of March XD. I didn't notice that, in the artist graph, some have highlighted March instead of February
xd ok, just noticed it too. It's one month less, so it's January and February.

Quote
Edit: Out of context question, how do you put the description when you insert a thumbnail in the forum?
I just attach them to the post, but I don't insert it into the text though

Quote
The playlist have 90 items, after removing the duplicates 31. Is only in suggested artists, the rest of playlist seems fine, in fact very good.
Have to investigate it, since that playlist is generated with ListenBrainz suggestions using the data I retrieve, and never had so many duplicates (I will also deduplicate it anyway on next updates).
Fixed.

Quote
"getEntitiesByTag: 400 {"code":400,"error":"multiple tags are specified but the condition param is missing"}" . Not sure if it was there before but the pdf is produced and the playlist too.
That doesn't make any sense on the current code an looks to me an API error on the server, since the param is always present.

Re: Foobar2000 Wrapped

Reply #34
Quote
Edit 2: Is so good that it predict the song of March XD. I didn't notice that, in the artist graph, some have highlighted March instead of February
xd ok, just noticed it too. It's one month less, so it's January and February.

No singularity.... :-(

Quote
Edit: Out of context question, how do you put the description when you insert a thumbnail in the forum?
I just attach them to the post, but I don't insert it into the text though

Thanks

Quote
The playlist have 90 items, after removing the duplicates 31. Is only in suggested artists, the rest of playlist seems fine, in fact very good.
Have to investigate it, since that playlist is generated with ListenBrainz suggestions using the data I retrieve, and never had so many duplicates (I will also deduplicate it anyway on next updates).
Fixed.
Quote
"getEntitiesByTag: 400 {"code":400,"error":"multiple tags are specified but the condition param is missing"}" . Not sure if it was there before but the pdf is produced and the playlist too.
That doesn't make any sense on the current code an looks to me an API error on the server, since the param is always present.

No worries then

Re: Foobar2000 Wrapped

Reply #35
All points should now be fixed on the latest commit.

Quote
I made one report from 2023 and the most listened track was "untitled" with 29, that seems to integrate all untitled songs from every artist.

This happens as well with typical title like "movement",  "part", etc and in general with all songs with the same title from different artists.
Including this one. With duplicated titles being assigned to different artists internally (please check this one, since I have not been able to test it properly).

Quote
EDIT: will be possible as long as you separate tags by ', ': '%GENRE%, %STYLE%'
And this is now possible.

Re: Foobar2000 Wrapped

Reply #36
All points should now be fixed on the latest commit.

Quote
I made one report from 2023 and the most listened track was "untitled" with 29, that seems to integrate all untitled songs from every artist.

This happens as well with typical title like "movement",  "part", etc and in general with all songs with the same title from different artists.
Including this one. With duplicated titles being assigned to different artists internally (please check this one, since I have not been able to test it properly).

Quote
EDIT: will be possible as long as you separate tags by ', ': '%GENRE%, %STYLE%'
And this is now possible.

The duplicates in the suggested seems to be fixed, only one song duplicated.

The title-artist thing situation is better than before. One untitled is the first with 10 listens. The album have 8 songs, all with 8 plays in last.fm. cleaning the last.fm data, the internal playcount show one with 2 plays and the others with 1 each So seems to be a last.fm problem with an album with the same title in various tags. Others albums with the same structure show the same problem in last.fm. Mental note, put some aditional info in title in album with a lot of songs with the same title

The rest are ok

But in the graph with the month the nº1, the untitled, have no artist name. And is missing in the top 5 too.

In your favorite no artist name and 1415 listens with 805 tracks. this is a bit weird.

The server message now have more information

Code: [Select]
Wrapped: retrieving images...
lookupArtistMBIDsByName: Retrying request for  to server on 500 ms...
search: 400 {
  "error" : {
    "status" : 400,
    "message" : "No search query"
  }
}
search: 400 {
  "error" : {
    "status" : 400,
    "message" : "No search query"
  }
}
lookupArtistMBIDsByName: 400 {"help":"For usage, please see: https://musicbrainz.org/development/mmd","error":"The given parameters do not match any available query type for the artist resource."}
lookupArtistMBIDsByName: 1 artists (1 failed)
Wrapped: processing track images with nconvert.exe




Re: Foobar2000 Wrapped

Reply #37
Correcting the data in last.fm and locally fix the first issue.

the rest the same

Other weird thing

The report says that 20 jun 1627 min (21 h), wierd too. this day I don't have any song played.

 X

Re: Foobar2000 Wrapped

Reply #38
Not going to fix user errors while tagging their files but glad you finally sorted it out.

Also if an entire album has all their tracks tagged with the same title... then... xd just use the global filter to remove it. Can not account for it unless I force people to tag all their tracks with picard to get unique musicbrainz ids. I could offer a setting to switch between tittle or ids though.

Do you have tracks without artist? That's weird.

.\main\spotify\wrapped.js
Go to line 49 and set bDebug to true.

And paste the log here.

Re: Foobar2000 Wrapped

Reply #39
Not going to fix user errors while tagging their files but glad you finally sorted it out.

Also if an entire album has all their tracks tagged with the same title... then... xd just use the global filter to remove it. Can not account for it unless I force people to tag all their tracks with picard to get unique musicbrainz ids. I could offer a setting to switch between tittle or ids though.

Do you have tracks without artist? That's weird.

.\main\spotify\wrapped.js
Go to line 49 and set bDebug to true.

And paste the log here.

No, all have artist. And generally musicbrainz tags too

I fix it already and add %album% #%discnumber%.%tracknumber% to every song untitled XD.

Re: Foobar2000 Wrapped

Reply #40
You say all have artists.. but the log clearly states there are tracks without %ARTIST% tag.

Check the track Through Abraxas III

Re: Foobar2000 Wrapped

Reply #41
You say all have artists.. but the log clearly states there are tracks without %ARTIST% tag.

Check the track Through Abraxas III

All have artist

X

but not all have artists, my bad.

Trying again with correct tag.

No server error, report more similar to the last.fm data

Only strange thing is the weird data of 20 jun and no data at all in last.fm.

Ah, and the photo of dominican ilegales instead of spanisn ones :-D

Can I check in foobar2000 the data of this day?




Re: Foobar2000 Wrapped

Reply #42
The photo is whatever Spotify has.

The report does not necessarily has to match last.fm or listenbrainz stats, since you can also filter the listens by query. By default it filters tracks rated with 1 or 2 stars, since they don't give any interesting info. But all that is up to you and your settings. People can change the filter or tags to whatever they want.

About the 20 jun data, that's something you will have to investigate yourself xd using %played_times% I suppose and some TF expression to check for 2023-06-20- Maybe you played a track multiple times but not fully, so the playcount was increased...

Re: Foobar2000 Wrapped

Reply #43
The photo is whatever Spotify has.

The report does not necessarily has to match last.fm or listenbrainz stats, since you can also filter the listens by query. By default it filters tracks rated with 1 or 2 stars, since they don't give any interesting info. But all that is up to you and your settings. People can change the filter or tags to whatever they want.

About the 20 jun data, that's something you will have to investigate yourself xd using %played_times% I suppose and some TF expression to check for 20023-06-20- Maybe you played a track multiple times but not fully, so the playcount was increased...

Yes, it's something to investigate, I'm intrigued by where 21 hours of music in one day can come from.
The rating couldn't be the problem, I never used it, but I removed anyway.
Maybe some error in the change to foobar v2, I remember some problem with the data import in the first versions.


Re: Foobar2000 Wrapped

Reply #44
Will also add support for https://marc2k3.github.io/foo_playcount_2003/ plugin, now that it also offers the played times (and they can be edited).


Re: Foobar2000 Wrapped

Reply #46
I make a backup and after exporting the statistics I reset all files and wipe out the last.fm scrobbles.

20 Jun now have half the time listened. It is not the 20 Jun but is the data of 20 May.

The problem seems to be my database.

Re: Foobar2000 Wrapped

Reply #47
@regor

I think I found the problem.

Like I said before I reset the statistics and last.fm play.count and did a new report, that, beside the date offset, showed a more accurate data. I sippossed it takes the data from last.fm or listenbrainz, because in local there were nothing

well, after that I imported the last.fm scrobbles and make a new report.

In this one the tracks have double listens and the 20 jun have another time 21 hours, the total of minutes also almost double the initial.

In local db the playcount are correct, but the report said double, or almost double


Edit: the backup of the internal statistics is not imported yet, so the problem in the transition to v2 is discarded

Re: Foobar2000 Wrapped

Reply #48
The script currently takes times from:
- LASTFM_PLAYED_TIMES
- PLAYED_TIMES

And merges them (it may well be that I forgot to deduplicate this step, and therefore you have double times for everything). Since I don't use last.fm I never experienced this problem. Please post here the tag values for such track so I can check.

If those 2 return nothing then..
- LAST_PLAYED_ENHANCED
- FIRST_PLAYED_ENHANCED

And checks if the first and last played fall in the checked range, in such case it adds fake listens starting with the first listen date and then repeats the last played date until it reaches the total playcount. This was done to get something for tracks which were logged with foo_playcount but not with the enhanced version.

Now... it could be that case, and therefore it added all plays to the 20 jun date.

The algorithm was coded before I created wrapped, and the aim was just playlist creation so It was irrelevant... but now that dates are needed, I could fake those listens and space out the dates in a random way.

Re: Foobar2000 Wrapped

Reply #49
The script currently takes times from:
- LASTFM_PLAYED_TIMES
- PLAYED_TIMES

And merges them (it may well be that I forgot to deduplicate this step, and therefore you have double times for everything). Since I don't use last.fm I never experienced this problem. Please post here the tag values for such track so I can check.

If those 2 return nothing then..
- LAST_PLAYED_ENHANCED
- FIRST_PLAYED_ENHANCED

And checks if the first and last played fall in the checked range, in such case it adds fake listens starting with the first listen date and then repeats the last played date until it reaches the total playcount. This was done to get something for tracks which were logged with foo_playcount but not with the enhanced version.

Now... it could be that case, and therefore it added all plays to the 20 jun date.

The algorithm was coded before I created wrapped, and the aim was just playlist creation so It was irrelevant... but now that dates are needed, I could fake those listens and space out the dates in a random way.

Witxes - Through Abraxas III First report said 3, after import 6

%LASTFM_PLAYED_TIMES%  |||  %PLAYED_TIMES%

["2014-05-30 15:10:15", "2023-02-22 14:28:15", "2023-02-22 16:13:11", "2023-02-22 16:24:39"]  |||  ["2023-02-22 14:28:16", "2023-02-22 16:13:12", "2023-02-22 16:24:40"]

The Yardbirds - I Ain't Got You  Have 2 version, one with 1 and the other with 2 plays. In the first report said 3 times. After import 6

%LASTFM_PLAYED_TIMES%  |||  %PLAYED_TIMES%
["2023-02-16 17:05:54"]  |||  ["2023-02-16 17:05:55"]

%LASTFM_PLAYED_TIMES%  |||  %PLAYED_TIMES%
["2023-02-19 12:26:19", "2023-02-19 12:28:30"]  |||  ["2023-02-19 12:26:20", "2023-02-19 12:28:31"]


Seems the values are not identical, there is a second of difference