Remove Mouth Noises in Voice Over or Podcast Audio Pt.3

In the previous post, we learned how to use the mouth de-click module in iZotope RX to remove mouth sounds from an entire podcast, sermon or voice-over for video.

We talked about doing this by setting up and using a gentle pass with mouth de-click in Part 2. This setting allowed us to transparently remove problem sounds without noticeably affecting the voice recording.

But before we get into any specific mouth de-click settings for remaining bad spots, we need to grasp some important concepts to harness RX's editing power.

How do we organize our processing steps in RX to save time and achieve the clearest sounding voice recordings?

What's Order Of Operations Got To Do With It?

First, let's touch on the order of operations in RX. This just refers to which modules in RX should be used before others when you're correcting your voice recordings.

It's important because it can affect the overall quality of your finished voice recording if things are done out of order.

And running the mouth de-click module is one of the earlier processes that should be used when you're doing the correcting on your voice recordings in RX.

This is according to iZotope RX's official order of operations here.

And here is iZotope's handy order of operations chart that may be helpful for you.

The only issues they suggest addressing before running mouth de-click are clipped audio or stereo issues interestingly enough:

1. Clipped Audio ========>

2. Stereo Issues =========>

3. Mouth Noise =========>

4. Next Module =========>

5. Next Module etc ======>

However, using my ears as the guide, I've unknowingly strayed from their official order a bit and have still come up with great sounding voice recordings.

For instance, I've dealt with background noise before dealing with random clicks and pops and haven't noticed any degradation in sound.

I've found that denoising a dialogue recording has sometimes helped me hear clicks and pops that were hidden by the background noise.

Possibly, iZotope may be referring to passes over an entire recording with some of the thinking in their order of operations rather than to small isolated problem areas, but I don't know for sure.

With that being said, I would advise following their recommended order as much as possible. But just know that their order of steps are not set in stone.

You can change the order of processing when you need to. As always, let your ears be the guide! 

Workflow And Individual Bad Spots

Sometime after running a gentle pass with the mouth de-click module, you'll need to zero in on any remaining troublesome mouth noises.

We saw what they look like in RX back on this post.

You could zero in directly after running a mouth de-click pass over the whole recording and treat individual spots all the way through.

Or, you could wait and try to consolidate these kinds of correcting tasks to one go of systematically working through all the trouble spots in your voice recording.

This is what I do with my dialogue recordings. I do think it saves some time to consolidate every possible correcting task down into one detailed process of working through the recording.

For example, after the gentle overall pass with the mouth de-click module, I'll usually deal with the problems listed below all together.

But this is after clearing out issues of background noise in my voice recording which we'll get into in another post.

The types of problems I usually lump together for my one go through the recording are: 

• clicks and pops (this includes individual spots of bad mouth noise - mouth clicks and smacks, but there are other types of clicks like digital clicks)
• plosives
• bumps and thumps
• crackles
• any other weird, problem sounds (like sounds from your mouse or keyboard, rustling sounds, passing cars etc.)

If you didn't do it this way, I fear you might end up working all the way through your dialogue recording more than once.

And I would love to spare you from that if possible (it's mind numbing and time consuming)!

Nonetheless, anytime you pore over your recording in this way, it will be time consuming, but there is a benefit.

Transparent Editing In RX

Working all the way through your voice recording, systematically dealing with bad spots, can sound fantastic for your audio! 

This is because the process can be very transparent (unnoticed) when treating individual bad spots as opposed to full passes through the recording with RX's modules.

This method preserves the overall sound of your voice recording when you do it carefully.

When I say "careful", I mean, listening to your edits to make sure they blend in well with the surrounding audio. Also, it means not over processing with the wonderful array of modules RX has to offer.

The goal is for no one but you to be able to hear that any editing was done when listening through your clear sounding recordings.

But don't worry if there are small, noticeable spots after editing... I've found that most people aren't bothered or don't notice them nearly as much as we, the audio editors, do!

Sometimes we are just so close to the process that we need another set of ears to give us some perspective on the overall sound. You know, the classic example of "can't see the forest for the trees".

Summary

Remember, the order of processing in RX is important because it can affect the overall quality of your finished voice recording.

Next, consolidate problem areas that remain after full processing passes into one go of working through your voice recording. Things like mouth noises, plosives etc.

Finally, get into the habit of carefully listening to your edits to make sure they blend in well with the surrounding audio and avoid over processing.

Try out these concepts in your voice recording today and let me know how it's working for you!

Cheers, Bryan
Your Dialogue Editing Mentor