Thursday, 02 May 2019
Is there a «standard level» for digital recording?
Well it’s a tricky question. In one sense, a standard might be to adjust the recording process so that the highest peak in each song uses 100% of the sample capacity – and it’s quite common to find songs recorded like that. But of course, this will result in the *average* amplitude of songs varying enormously, and it’s the average amplitude that mostly determines how «loud» the song sounds to you (although the frequency distribution has quite a bit to do with it too – perceptual audio is complex). If you make the songs all the same average amplitude, you will probably end up «clipping» the peaks of some of them, unless you aim for a pretty low average amplitude. This is why I prefer to adjust each song manually in a good wave editor. I might add that I don’t think a little clipping is necessarily the end of the world when you’re dealing with rock… ultimately perception is everything, and most automated «normalisation» doesn’t account for perception very well.
