Pros and Cons.  It's not more meaningful either way.

What you get is an average per disc, and not an average per album.  I believe 
that for albums distributed on two disks, it should be album replaygain should 
be calculated as an average of all tracks on an album.  Double albums used to 
be distributed as 4 sides on vinyl - I wouldn't normalise across for sets of 
tracks on the album.

It's irrelevant whether double-disk albums are ripped with disc numbers or as 
one album with contiguous track numbers if "treat multi-disk albums as one 
album" is enabled.  The effect will be that there will only be one album 
replaygain value.

If you rip with the disc number in the album name, or treat multi-disk albums 
as separate albums, you will have one album replaygain value per disk from that 
album.

>imagine if you had a 10 disc album, which is kinda what moonbase was
>saying...  you could have a LOT of varience disc to disc that you are
>trying to compensate for with one single RG value for all discs.
>
I rip those with a volume number in the album title, because they are generally 
not single albums, but collections of albums.  eg. I have a multi-disk 
collection by Klaus Schulze called "Jubilee Edition" that is 25 disks.  I 
haven't ripped that one yet, but when I do, I will rip as 25 albums called 
"Jubilee Edition Volume 01 - Tradition & Vision", "Jubilee Edition Volume 02 - 
Avec Arthur", etc.

However, "Pink Floyd - The Wall" is ripped as one album.  If mastered 
correctly, such albums don't have much of a varience in average volume between 
two disks.  If they did, who's to say that that is not intentional.  I don't 
want a jump in volume between the track that happens to be at the end of disk 1 
and the start of disk 2 if it's not meant to be there.

Albums are albums, and tracks are tracks.  Disks are largely parts of an album, 
and irrelevant for album normalisation.  Volumes are collections of albums - 
each album would have album volume normalisation.

Taking the averaging one step further in your way, you could forget disk album 
averages and just use track replaygain.  One step further would be to split 
each track into 10ths and have an average volume per 10th of a track...  that 
is of course nonsense, I'm just pointing out that it doesn't make more accurate 
volume normalisation.
_______________________________________________
ripping mailing list
[email protected]
http://lists.slimdevices.com/lists/listinfo/ripping

Reply via email to