I've lurked in these forums off and on for the better part of a year, and thanks to a lot of good advice from people that obviously know a little of what they're talking about, my MP3 colllection (And growing OGG one) sound better to me than my first days of using Real Jukebox (ugh!) and CDEX with the old LAME DLL (CBR, mind you) ever did.

In the alt.binaries.sounds.mp3.d group, some one is extolling the virtues of normalization, and how it doesn't adversely affect the mp3. I've always read not to use normalization, and I currently use MP3Gain to fix my files now. But it hit me - I really don't know the technical details. *why* is normalization not preferred? I've done some searches in google and on these forums and didn't come up with anything regarding a listening test / technical information - although I probably just didn't know what terms to search for.

So my question - why is normalizing bad? Links to articles / forums is fine, I don't mind having to find it out for myself, I just need pointers on where to start.

>>>'*why* is normalization not preferred?'<<<

because you get rounding errors in your file unless you re-dither. thats first. most of the people hanging around this newsgroup (i've read some of their "advice" in the past) don't bother to mention that. second, peak normalisation does not make the tracks off, say, different CDs sound equally loud. it just applies a gain so that the maximum sample value is reached. this has nothing to do with the "perceived" loudness of the audio file, and will only mess up the levels of recordings that are meant to sound quiet