benzo8 wrote:I can't quite see why this should be the case for bitrate or track volume. Surely the data that MediaMonkey needs to refer to is that of the file that actually gets played, not of the file that's been deleted?
No, the main purpose of this script was at begin to replace some audio file with much smaller redirecting file which should have same metadata as replaced file. If this small file has same metadata as master track then you would lose its main advantage, so you could use simple .m3u or .pls files instead. If I have two same tracks from two different albums, I want to see different metadata for those two tracks, e.g. metadata of official album and metadata of compilation album. If I put metadata of master tracks into .asx files, then I would lose information about compilation (album name, release year, publisher...).
If I put in .asx files all metadata from replaced files excluding bitrate and track volume then I will get inconsistency with a source of metadata.
benzo8 wrote:In one instance, the Master Track had a ReplayGain track setting of -4.2, and the replaced track had a ReplayGain track setting of +2.1. As the .asx file contained the deleted track's metadata, MediaMonkey applied +2.1 gain to a file that actually needed 4.2 taken away, and as a result it played 6.3dB too loud, and distorted.
Your description seem to be very reasonable and maybe I could make an exception with the track volume, but what about the album volume? If the replaced track had 2.1 dB, when I change the track volume of .asx file to -4.2 dB the album volume of all tracks from that album should change as well. So, I need to calculate that new album volume and to put it into all tracks from that album. This is something that I don't like to do. One more thing, what if some user change mind and decide to apply Remove Redirection from the Track
option? The .asx file would be replaced with the originally replaced audio file which have 2.1 dB in your case, but database contains information of -4.2 dB about it taken from master file. So, if the user decide to apply the mentioned option I need to extract the original track volume from the audio file somehow and to put it back into database which is another thing that I don't like.