Interview Audio Levels

Since Jon Udell left a comment on my earlier post, I thought I’d do a quick write up on my current interview recording and processing technique. I adapted my current process from many other smarter people like Doug Kaye, after having multiple instances of having to laboriously fix the sound on interviews I had conducted. I’ll keep it as generic as possible since people might be using a number of possible setups that are equally applicable. In order to make this work, you do need to be able to get yourself and your guests on different channels. With a mixer that’s trivial. If you are doing this straight on a computer without a mixer in the middle, you might have tougher sledding. Once you cross the conceptual hurdle of having each person in one ear, the rest falls into place and becomes trivial.

  1. Get yourself and your guests in different channels. This is crucial. Not only does it enable all the subsequent stuff, it also lets you fix it when you and the guest talk over each other or if Skype lag screws things up.
  2. In Audacity or other sound editor, do a pass of compressing and normalizing the left channel separately, then go back and do the right one. Normalize each of them to the same point, preferably -4 dB or so. If either you or your guests have enough variability that some sections are still significantly lower than the point you normalized to, go back and amplify those sections. Any reasonable sound editor will have this function.
  3. Absolutely without fail, before you publish it move the voices back to the center. Don’t publish this with you and your guests in the opposite ears. If you want to do that stereo effect of having a pan, don’t do it by any more than 10% from center for either ear. Me personally, I just convert the whole thing to mono when I’m done.
  4. That’s it. Pretty simple, no?

One Reply to “Interview Audio Levels”

Comments are closed.