I'm working on a video game at the moment with 1500+ WAV files which represent all the bits of speech in our game. The problem is that they're not all consistent in average RMS output, varying anywhere from 3-9dB depending on which recording session they came from. Ideally, I'd love to find a way to batch process them all so that their average RMS output is consistent, as it's my understanding that this is the best measure of "perceived loudness". Anyone know a way? I've thought about using Cool Edit Pro to set up a batch script, but as far as I know you can only do this for peaks (normalising), not average RMS output. I'm on Windows XP, btw. Cheers.