I listen to podcasts.

I edit podcasts.

I love that this medium has seen the tremendous growth that it has over the last few years. No matter how esoteric the subject, there is probably a podcast about it. The chances are also good that the audio quality is, shall we say, lacking. Let's see if we can do something about it. I'd like to propose a small step in the right direction.

First, let's get some caveats and assumptions on the table. For the purpose of this discussion, I am going to assume that GarageBand is the editor that most folks use for their burgeoning podcast empires. The underlying technique is useful regardless of your choice of editor. The steps will just be different. I am also going to assume that folks know GarageBand well enough to add effect plugins to tracks, as well as the master output.

What I am proposing are two simple adjustments that will help raise the bar for consistent audio across varied listening environments. It will not magically make your podcast sound like NPR. It might even reveal other shortcomings in your recordings (EQ, mic technique, reverberation, etc). This is OK. That's how we improve.

OK. Let's talk 'audio levels'. When I listen to podcasts, differences in audio levels are more frustrating than virtually any other issue. This is most evident in podcasts with two or more participants. Quite often one person's voice will be louder than the other's. If I'm listening with headphones or earbuds, I can usually find a volume setting that's tolerable. All bets are off when I listen in the car. The inherent background noise in a moving vehicle makes those quieter voices very hard to hear. I find myself fiddling with the volume constantly.

Let's see if we can improve this.

In GarageBand, we are going to add a compression plugin to individual audio tracks, as well as the master output, and make a few simple tweaks. Let me introduce you to AUDynamicsProcessor. This is a basic AudioUnit compression plugin that's available in GarageBand. Start by adding this to each voice track in your project. We are going to focus on making two simple adjustments. Take a look the screenshot and notice the two controls highlighted by the circles.

Make these adjustments while listening to a representative section of the voice track your working on. When I say 'representative', I mean a section where the voice sounds 'conversational'; no yelling, no whispering, etc. Grab the point on the graph and drag it down the line until you see a tiny bit of red where the two lines diverge (see screenshot). Next, grab the 'Master Gain' control and slowly drag it to the right until the blue level meter dances around the 0dB mark.

That's it.

Now use your ears. Adjust the control on the graph to find what sounds best to you. Make very small adjustments. You should be able to hear the difference. Keep an eye on that level meter to ensure that the overall level remains near 0dB.

Repeat this same process on all voice tracks. Once all of those are done, add this plugin to the master track and repeat the process. This should give you a podcast with much more consistent levels.

Give this a shot. I think you'll find that it makes a difference.