I noticed a slight oddity, and am wondering how to best handle it.
Please forgive me if this has been discussed; I tried a few searches, but didn't see anything obvious.
I rip albums to single-FLAC-file plus CUE file. I also use DISCNUMBER & DISCTOTAL/TOTALDISCS tags, so that multi-disc albums show up as a single album in LMS, with tracks 1-1 to n-m, where n is the number of discs, and m the last track on the last disc.
I end up with a file structure like:
album/CD1/file.flac
album/CD1/file.cue
album/CD2/file.flac
album/CD2/file.cue
[I don't do this manually; I use XLD on Mac]
So there are multiple CUE files for each album, one for each CD. Each CUE file has a single REPLAYGAIN_ALBUM_GAIN value. Generally the value for this is different for each CD in the album (as much as 1dB).
When LMS scans the files, it seems to use the REPLAYGAIN_ALBUM_GAIN from the last CUE file read for that album, as the single gain for the whole album, regardless of which value is greater.
[There are no embedded gain tags in the flac file's VORBIS_COMMENT section; all gain values are in the CUE files]
I'm not saying this is a bug, but I'm not sure it's ideal, either. One annoyance is that if the last CD in the album is quieter, that gain value is applied to all the CDs in the album, and you get a fright when the first is too loud.
It might be nicer if LMS would take the largest of the REPLAYGAIN_ALBUM_GAIN values it obtains, for a multi-file set, and use that as the album gain? That would at least avoid the first CD being played louder than I'd like, when the second CD has a lower gain value.
Or, perhaps even better, use a different "album" gain value depending on which track in the album is being played? Although that wouldn't make sense in random-order play, but could perhaps be done only when playing the album in normal track order. e.g. the applied gain would be the same for tracks 1-1 to 1-n, but then change to the next value for tracks 2-1 to 2-n, etc.
Thoughts?
thanks very much
Please forgive me if this has been discussed; I tried a few searches, but didn't see anything obvious.
I rip albums to single-FLAC-file plus CUE file. I also use DISCNUMBER & DISCTOTAL/TOTALDISCS tags, so that multi-disc albums show up as a single album in LMS, with tracks 1-1 to n-m, where n is the number of discs, and m the last track on the last disc.
I end up with a file structure like:
album/CD1/file.flac
album/CD1/file.cue
album/CD2/file.flac
album/CD2/file.cue
[I don't do this manually; I use XLD on Mac]
So there are multiple CUE files for each album, one for each CD. Each CUE file has a single REPLAYGAIN_ALBUM_GAIN value. Generally the value for this is different for each CD in the album (as much as 1dB).
When LMS scans the files, it seems to use the REPLAYGAIN_ALBUM_GAIN from the last CUE file read for that album, as the single gain for the whole album, regardless of which value is greater.
[There are no embedded gain tags in the flac file's VORBIS_COMMENT section; all gain values are in the CUE files]
I'm not saying this is a bug, but I'm not sure it's ideal, either. One annoyance is that if the last CD in the album is quieter, that gain value is applied to all the CDs in the album, and you get a fright when the first is too loud.
It might be nicer if LMS would take the largest of the REPLAYGAIN_ALBUM_GAIN values it obtains, for a multi-file set, and use that as the album gain? That would at least avoid the first CD being played louder than I'd like, when the second CD has a lower gain value.
Or, perhaps even better, use a different "album" gain value depending on which track in the album is being played? Although that wouldn't make sense in random-order play, but could perhaps be done only when playing the album in normal track order. e.g. the applied gain would be the same for tracks 1-1 to 1-n, but then change to the next value for tracks 2-1 to 2-n, etc.
Thoughts?
thanks very much