If your host does that, then yes, it's a possible working method. However...I don't really grok the concept of normalizing to an average conceptually (I have no idea how other hosts handle it, mine doesn't do any RMS/average normalizing). The whole concept of normalizing involves taking the highest peak and setting it to the normalization value, adjusting everything else by the same amount. When you're talking about average, you're talking about the highest average value and matching that to -18? What kind of RMS window are you using (because that's going to vastly change your RMS reading)? Seems like an awful lot of "engineering" being automated in a way that's kind of out of control and arbitrary.scooter wrote:Wouldn't it be as simple as Normalizing the WAV file to -18dbfs "average" and not peak? Making sure to keep them at 32bit or 64bit float
Gain staging used to be one of the main things that audio engineers worried about in a session. It's part of how you get nice sounding mixes (I realize this is a debated matter; suffice it to say I'm firmly in the "gain staging matters a lot even in digital audio" camp). It shouldn't be looked upon as an extra step (although, all too often, it is). Having 1000+dB of dynamic range doesn't render gain staging useless; there's a lot more to it than just making sure you don't clip. But even if you're in the "gain staging isn't necessary in digital" camp, using Nebula to replicate analog devices does require some attention to this detail.
Just my curmudgeony opinion. I also had to walk to school uphill both ways =).