Why not just raise the Amplification or Volume of the entire recording
Sorry for the confusion, that's what I meant by normalizing the entire fileset - perform the operation against the master recording all at once, not individual tracks or files.
I think as long as you dont Raise and lower each track by huge amounts the dynamic range will still be intact.. (Such as dont normalize one track to -20 and another to -5) . I usually dont have but a db or two difference in each track . which is about where it was to start with.
I think your comments are a little mixed up: in order to NOT raise or lower each track by huge amounts, we must utilize different RMS values for each track in order to maintain relative dynamic range, i.e. do exactly as you suggest we should not: normalize one so it's RMS is -20dB and another to -5dB.
Let's take an example: We have two tracks, [1] with an RMS of -21dB, [2] with RMS of -9dB. That's an RMS difference of 12dB across the two files. Now, let's normalize each track independently to the same RMS value, -8dB. That entails raising the RMS on [1] 13dB, and raising the RMS on [2] only 1dB. So now, both tracks have an RMS of -8dB. The RMS
difference across the two files is now 0dB. Dynamic range: gone!
In this case, in order to preserve relative dynamic range - but still increase amplitude - we would want to normalize each track to very different RMS values, say increase [1] so it's RMS value becomes -16dB, and [2] so it's RMS value becomes -8dB. This brings the amplitude of track [1] up to a better level for listening (without having to adjust volume during playback). The dynamic range is reduced, but not removed. However, I believe there's a better way to "amp up" the recording but still maintain relative dynamic range...
I use a general rule of thumb that you should never have to turn your volume knob on your stereo over Half way or you simply need more power or in this case more db .. Seems like the same argument here.
The way to accomplish this, IMO, is to apply compression to the entire master WAV and then raise the dB across the entire master WAV. That way, you maintain the relative dynamic range, if not the actual dynamic range proper.