Hire Greg at iZotope

How I Use iZotope to Clean Podcast Audio - Part 1 

An amazing thing that has happened over the last 5-10 years is the democratization of audio production. The cost of gear has come down, and now more than ever it is easy for you and your friends to spend a couple hundred bucks and you are off to the races recording your own podcast! Exciting, right?! 

Well... let's not get too caught up in it all because a podcast can also take a bit of work to get it to sound *nice*. Luckily for us, iZotope has created a number of tools that will help clean up our recordings and have things sounding professional before you know it.

The issues that arise are actually not all generated from the quality of the mic. In fact, if you won the lottery tomorrow and went out and dropped $10,000 on a Telefunken mic, got it all setup, hit record and then listened to the playback, you would be appalled. What you would be hearing is the room, your central heating, the sounds your mouth makes as it speaks, the fan from your computer, you get the point. 

Noise pollution. It is our enemy. And iZotope is our saviour.

Let's have a listen to this raw audio that was captured using Blue's Snowball iCE microphone in an untreated room. 

Great content, great take, but there are lots of artifacts present that degrade the quality of the recording. Let's analyze this a bit:

  • We can hear the room. That is to say that the microphone picked up a lot of ambient noise along with the beautiful voice.
  • There are lots of mouth sounds. It is a natural part of speaking, that the mouth will make clicks and other textures that are not wanted on a recording.
  • The microphone and the room are also both contributing a low-level hum to the recording.
  • As always, there is EQing to do in order to decrease the frequencies we don't like and increase the ones we do. 

Let's tackle this one issue at a time and start with the room noise. I use iZotope's RX 9 De-reverb as an insert on the dialogue track. This allows me to very finely tune in how much room reverb is coming through. Beyond that, it goes even further and allows me to set the threshold in four frequency bands to compress the reverb amount in specific tones of her voice. The toolset is especially nice as it also lets you solo each band to hear exactly what is going on. Remember, the greatest tool we have as people in audio is our ears, so listen carefully and dial in exactly what sounds best to you.

Also, if you are looking for a hand in this process you can start playback, and click the Learn button in the top left of the plugin. Let it play for at least 10 seconds and when you stop playback it will automatically setup the plugin to what it deems to be the best settings. 

Going one step further, if you are looking for an inexpensive way to get better quality recordings, consider making some acoustic panels! For around $100 you can get the materials needed to put together an acoustic panel. If you make a few of these and hang them strategically in your recording area the room noise will improve. I learned how to make these from Amond Jackson, you can contact him for access to the training if you are interested. 

Next we are going to apply a gate to the dialogue track and this is really going to clean things up. A gate stops a signal from coming through depending on where the threshold is set. If the volume of the signal is above the threshold then it allows the signal through, but if the signal is below the threshold it stops it. It is very important to take special care dialling in the settings so that the gating sounds natural.

Neutron 3 is a very powerful tool which I use extensively for EQing. However, it also has a great Gate tool in it. It is easy to operate: you drag the threshold bar to where you want it set and choose whether you want the gate to open or close when triggered. Then you can set to dialling in the details such as the ratio, attack, hold, and release.

The ratio is how much the signal is being reduced when it falls below the threshold. The higher the ratio the higher the reduction. Attack is how quickly the reduction starts. Hold is how long the gate will stay open after the signal has dropped below the threshold. I think of this like pre-delay on reverb. Lastly you have release which is how long it takes for the gate to move from open to closed.

I've noticed the two I have to be very careful with is threshold and release. If the threshold is too high then the ends of the speakers words will start to get clipped. If it is too low then the room noise becomes very apparent again. There is always a sweet spot somewhere in between those two places. As for the release, it is similar in that you are looking for a sweet spot. Too fast and the transition will be unnatural and stick out, too slow and the room noise creeps back in. But somewhere in the middle is just right.

Now that we have cleaned up some of the room noise we can really hear those mouth noises. We'll tackle that in the next post!