The loudness war has inched up the levels over the years to the point where our perceptions are becoming tainted. If you compare the levels between this and 80's CDs (generally considered to be better sounding by a lot of audiophiles and collectors) or all classical CDs, you'll find they are all about the same. This is because RG analyses the perceived loudness not the peaks. If you try to adjust this simply with peak normalizing, it will sound a lot less loud than the same material procssed with replaygain. Yes, some material will come across with a peak of -5.57dB, but then again I've seen other very dynamic material with peaks of -1dB. Of course this all a matter of one's taste, which is why the Replaygain settings in Foobar provide an adjustment for the target level (by default -89dB). This might seem low compared to a lot of modern commercial recordings, but believe me, once you start using Replaygain and targeting a more reasonable level like this, you get used to using the volume control, and the sound quality seems better. I recently read some writing by Bob Katz in which he mentions that some electronic equipment actually starts distorting or clipping before the signal reaches 0dB. If you go to close to 0dB, some programs will actually calculate 'overshoots' that go beyond 0dB. A lot of programs and people will tell you anything under 0dB is fine, but this depends on a lot of factors. I find it works well for me.ДЄs for level, definitely no more than -1dB.

With properly implemented Replaygain, everything will be adjusted and if the peaks would go over clipping, they are limited to avoid distortion. The peaks will be adjusted to the target you set but the rest will remain very low. This won't help if, for example, you have a file with a few really loud peaks. A lot of audio programs do something called 'normalizing' but all this does is find the loudest points of the file and adjust the whole file so that the loudest part hits the target. This uses proven scientific algorithms to get all files sounding roughly the same volume. I strongly support using Replaygain or similar (Adobe Audition's Group Waveform Normalize function with 'equal loudness' ).

Best replay gain settings