Thanks, I hope I got it now. Here should be the final code:

The System Shock MP3 files are 320kbit MP3s, so each frame is either 1044 or 1045 bytes in size. Stepping through the code with GDB, I realised that the problem is not the Xing/LAME tag itself, but rather the frame after it. mpg123 and Foobar both expect the next MP3 frame to start at the correct place, exactly 1044 or 1045 bytes after the first one. But it's not present: instead, that part of the file is zero. mpg123 and Foobar both assume that the Xing/LAME tag is invalid and ignore it.

drop table if exists albumSummary;
create table albumSummary as
  select album,
         [album artist],
         date,
         round(sum(rating*length_seconds)/sum(length_seconds), 2) as albumrating
  from Playlist
  where playlist_index = active_playlist()
  group by 1, 2, 3;

update PlaylistUpdatable as a
  set albumrating = b.albumrating
  from albumSummary as b
  where       playlist_index = active_playlist()
        and a.album          = b.album
        and a.[album artist] = b.[album artist]
        and a.date           = b.date;
drop table if exists albumSummary;

hi @forart.eu
will be released a gui to use it?
thanks

Duration : 1:00.000 (2 646 000 samples)
Sample rate : 44100 Hz
Channels : 1
Bits per sample : 16
Bitrate : 706 kbps
Codec : PCM
Encoding : lossless

Duration : 0:14.956 (659 567 samples)
Sample rate : 44100 Hz
Channels : 1
Bitrate : 128 kbps
Codec : MP3
Codec profile : MP3 CBR
Encoding : lossy
Tag type : id3v2.4
Embedded cuesheet : no

WARNING: "C:\Music\TEST.mp3" (offset 0x2e4): MPEG stream error, resynchronized successfully
WARNING: "C:\Music\TEST.mp3": Wrong number of MPEG frames specified in Xing header (2298 instead of 2301)
WARNING: "C:\Music\TEST.mp3": Wrong number of MPEG data bytes specified in Xing header (239409 instead of 240034)

Quote from: rc55 on 2023-09-25 17:53:01

... it's made a 40MB exe ...
This does not include the weights file which is ~300MB. The weights file contains the model data used to encode and decode the audio and is a necessity, and the model is not interchangeable. You have to use the same model to decode any encoded file.

Thanks for the analysis! Out of curiosity: could you 7zip (preset Ultra) that 40MB exe and 300MB weight file and let us know what file size comes out? That would be a rough estimate of how much room for reduction there is.

Thanks,

Chris

I've been trying to find matching crash reports in Google Play, but they've done something weird with the system, probably to "protect privacy", and no useful info can be extracted whatsoever (no stack traces etc).

1. i open the CUE file.