I could be wrong, too, but I thought MC did its own analysis and wrote the info to the tags, similarly to how MP3Gain does it. When I used to run MP3Gain, I only ran it on the tracks after I copied them to the CF card for my portable, so none of the music on my server has MP3Gain tags, and my volume is even when I listen on the PC and on my Squeezeboxes. The Replay Gain field is filled in, in MC.
It must be saved in the tags, also, or Squeezebox Server would not be able to read it. Although, maybe it depends on the MC settings you have for that library field--I suppose I may have changed it from the default if it was not to save a tag.
It makes me wonder if you are really using MP3Gain's tags or if it's using MC's tags and that's why you weren't hearing it on the iPod.
Sue, I'm glad you brought this up as I had a few misconceptions about how MC's replay gain worked vs. MP3Gain and I think I've sorted it out now after a little research and reading.
From what I understand, both MP3Gain and MC do replay gain adjustment but in different ways. MP3Gain does it by altering the data stream in an MP3, in a lossless and reversible way, whereas MC does it by simply writing a value to a tag. (The tag writing MP3Gain does is just for undo purposes.) But writing a value to a tag -- in this case "Replay Gain" in ID3v2.3 -- means that file is now dependent on a player's ability to read and act on that value. Some do and some don't. For example, Windows Media Player and of course MC do, but iTunes does not. Nor do iPods.
In my case, I was running files through MP3Gain with an 89 db target, and then through MC's audio analyzer which sets the volume level at 83 db automatically (value can't be changed). The audio analyzer read my MP3GainAdjusted files and then made an adjustment
on top of that to bring the level down to 83db. So the file was set at 89 db in the data stream, but the tag for Replay Gain brought it down to 83 db. This meant sometimes I was getting playback at 89 db (like on my Touch or Creative Zen player which don't recognize the replay gain tag) and sometimes I was getting 83 db (ie, MC and Windows Media Player because they read and interpret that tag).
I still prefer the MP3Gain method, since it means my MP3s will be played at the desired volume levels no matter what device I play the files back on. Also, many ipod users claim SoundCheck adds distortion on playback, though most of the complaints I saw were regarding older ipods. So I don't ever plan on using that.
The thing I plan to do now is stop running my MP3Gained files through MC's audio analyzer. And for all my music files, I've already deleted the fields that MC adds to a file when run through audio analyzer.
I thought I needed to run MC's audio analyzer to get MP3Gain values read into the database, but now that I know this is not the case, I don't need to use audio analyzer anymore. Saves me a step in my workflow.