top of page
  • Writer's pictureElliot Hardman

iZotope RX - A Crash Course for Voice Actors


Ah, iZotope RX. The strongest tool in every voice actors arsenal, misused or underutilized more than any other plugin on the market. Love it or loathe it, it's here to stay, purely because of how much it can do for a poor recording.

First off, I'll answer a good question - do I need iZotope RX? It's so expensive!

The straight answer is no. If you've got an impeccable setup and you find absolutely no undesirable noises in your recordings, then firstly; good for you. Secondly, it just really won't do that much for you. Most of the issues that RX fixes are also tackled by just having a better recording, though this is easier said than done.

So, what is iZotope RX?

Essentially, iZotope RX is an audio repair toolkit. It contains a suite of different plugins that can help you get your audio polished up. I'm going to be going over the absolute basics of this today - my mindset is that if you find you need to excessively use RX to make your recordings listenable, then you have a problem with your actual recording. Fix that first.

Over the years, I've found a large array of uses for it, from editing my voiceovers to removing clicks from guitar parts for simply just surgically removing something from a sample. It's great. However, I'll be covering usage specifically for voiceover in this article, as that's what we're all here for.

iZotope is applicable in 2 ways - you can either load individual plugins onto your recording track, or you can process a specific audio clip inside of RX itself. We'll be doing both in this post.


First up, the glorious de-plosive plugin. Subtle and yet so effective at eliminating those nasty plosives.

For the uninformed, plosives are hard 'P' sounds in your recordings. These occur when you, as a voice actor, blow a gust of air into your microphone when pronouncing certain words. You'd be amazed how many of these I hear in so many recordings. To me, as a techy audio nerd, they do sound quite unprofessional, especially considering that they're so easy to avoid.

So, how should you be using de-plosive? I use the de-plosive plugin in 2 ways. Despite the fact that I record with good distance from the microphone and a pop filter, plosives do occasionally happen, especially when you're reading something more intimate.

To work as a catch-all de-plosive, I insert a VST instance of RX's de-plosive plugin (the same way you would a compressor or gate, etc). The image below shows what I have my settings on for this. Super, SUPER subtle. This is just here to quash any small plosives that may have found their way into a recording. This does a super nice job of saving me a bit of editing time for some of these slight plosives.

Sometimes, that plugin alone isn't enough! So, I'll have to run through RX itself to clean it up slightly more. Let's give you an example.

Firstly, your process here will be made a lot easier if you have this setting turned all the way to the right - it lets you see the waveform more from a frequency standpoint. You'll find it in the bottom left of your RX window (this may differ from version to version - I'm on RX 7):

So, look at the image below. This is me saying some classic tounge-twister, plosive heavy phrase. Notice the heavy block of orange at the bottom of waveform? It might seem hard to spot at first, but trust me when I say that spotting these comes naturally after only a little time playing around with this.

That little block of orange is our plosive. With this waveform view inside of RX, a brighter/thicker block of colour means that the frequency there is more prominent. So, we need to get rid of it!

There's two things I do when removing a plosive inside of RX. Firstly, another de-plosive. I do this as a more surgical procedure, as to not hurt the rest of the audio. Here's how I do it:

  • Selecting the paintbrush tool (seen below), I select the area that the plosive in (trying to be fairly precise)

  • With the plosive selected, I can specifically target it with a stronger de-plosive plugin. From the toolbar on the right, select the de-plosive option to open a window just like the plugin from before.

  • I then hit the plosive with this plugin using the settings below. This will kill off most of that plosive and soften the attack on that frequency, making it already sound a fair bit nicer.

  • Finally, I almost always hit that same area with a slight gain reduction of about -6dB. DO NOT go too extreme with this, else it will just thin out your voice. You're looking for a very slight reduction in gain to make it seem as if there was never a plosive there. RX has a gain option in the right sidebar - I'd recommend using this.

And you're done! That plosive has been tackled, with it hopefully looking a bit more like this. Easy!

One final important note with using de-plosive - as mentioned before, going too crazy with this WILL make your voiceover sound very thin. You should always be very gentle with whatever you apply to the whole chain, leaving any heavier processing until you can be a little bit more surgical inside the RX app. Don't go crazy blasting every orange blob you find - only if you hear a plosive in the recording!


On the absolute flip side, we have the de-esser. This is there to tackle sibilance.

Again, for the uninformed, sibilance is the harsh 'S' like sounds we get when pronouncing certain words. Pretend that you're making the sound of a snake right now - that's what sibilance sounds like. It's sharp, harsh, and never sounds nice. So, we use a de-esser to tone it down a bit.

Now, I have to confess that I don't use the RX de-esser for my work. Personally, I use Fabfilter Pro-DS, which is fantastic and, in my view, a lot easier to use. However, I'll cover the RX plugin here.

I won't be able to give such specific guidance for de-essing here, as it's often very varied as to your recording space and the timbre of your voice. So, you'll need to use a lot of your own initiative on when it sounds right. I'll explain what sounds right shortly.

RX's de-esser is fairly simple in interface, which is a blessing and a curse. I use Pro-DS for the freedom I get with it, but RX does a great job itself. Let's explain the settings:


This is the level at which sibilants that are above it are attenuated/affected, measured in dB. Generally, you want this to be quite high (closer to 0). The bar on the left is the level of the sibilants, while the bar on the right shows you the gain reduction, based on how low your threshold is.

You should be toying with the threshold enough so that it's catching only particularly harsh sibilance and not a lot else. You shouldn't be dulling your voiceover completely with a de-esser, but instead just simply dampening those sharp hisses.

Cutoff Frequency

This is the frequency at which sibilants are detected above. In laymans terms, only harsh sounds about this frequency will be taken in by the de-esser. The default is 2500Hz, but it's up to you to play around with it and decide where this should realistically sit. You can get some particularly nasty high-mid sounds that can be cut with a de-esser, but I would strongly remind grabbing a spectral analyzer to see where most of that noise is coming from, before going too deep into a de-esser.

Everything Else

There's a few other settings here on the de-esser which I would be neglecting if I didn't mention, but my simple advice here is to just not worry about them. These are typically more advanced settings that won't be too important for the majority of us - if you'd like to look more into these, I'd suggest looking into a more in-depth RX tutorial.

Default settings for all of this is just fine. If you feel like playing around with them, though, go wild! Just note that the bottom 2 settings are only accessible when in spectral mode.

Mouth De-click

To me, the most essential plugin in RX's aresnal is the Mouth De-clicker. This plugin will detect mouth noises from your recording and eliminate them. Dead.

What do I mean by mouth noises? So, if you're a bit of a gulper of water or you simply just don't drink enough, you'll have a sticky or wet mouth. This causes your mouth to click a lot when you speak, which will be picked up in droves in your recording and sounds dreadful.

So, before even thinking about de-clicking, learn to control your water intake around a session! You should always be sipping your water instead of gulping, but making sure you're hydrated. Also, avoid sugary foods around a session!

If you've followed the above advice, you're probably still finding that there are at least some clicks in your recording. That's ok and is honestly relatively normal and easily fixed.

On to the plugin - I simply run a VST instance on my voice's track in my workstation which catches all of the clicks in my recording. I use the settings below:

This is a fairly heavy de-clicker, but this is what works for me. For you, it will be different. I don't worry about the other settings, as all I'm trying to do is catch the horrible mouth sounds and nothing else. Nice and simple, I need not do anything by change the sensitivty. I've floated the sensitivity around a few points over the years, but this setting captures all the noise without hurting the quality of my recording.

That being said, you really must be careful with this plugin - a sensitivity that's too high will begin to take random chunks out of your voiceover which it is mistaking as clicks. This plugin should NOT affect the quality of your voice at all. If it is doing so, you're pushing the plugin too hard.


I'm not going to linger on these for long for 3 reasons:

  • I never have to use them

  • If you find you need to use them regularly, your recording area likely just isn't suitable for recording

  • In my experience, a client's audio team will do this themselves if it's really necessary

iZotope RX has 2 different de-noisers. Spectral de-noise and Voice de-noise. I won't be seperating them whilst I talk about them, but generally speaking, you'd likely use the Voice de-noise in this application.

De-noisers work by grabbing a noise profile (often specified by the user) and then working to try and reduce the volume of that noise profile. There's a few different ways you can have the plugin grab a noise profile, but there's one specific way we'll look at.

When you open up the Voice de-noiser, you'll be met with a slightly scary screen like the one blow. But don't worry, it's all fairly straight forward.

So how should you use this? Firstly, your optimization should be set for dialogue. We aren't working on music! Filter type can vary dependant on your situation, so it may be worth flicking between the two settings to see what best fits the application.

The first step is to grab your noise profile. Adaptive mode is selected by default, but I personally think that it's a little less reliable than manually grabbing the profile yourself. That's where the 'Learn' button in the top left comes in.

When you uncheck Adaptive mode, you can use the Learn feature. To use this feature appropriately, you should have a blank space of a few seconds before your recording of nothing but the noise/room tone/whatever you're trying to erase. With this room tone playing, click the Learn button. This will grab the profile of the noise and change the main analyzer window to fit the profile.

From there, you can start working at reducing this noise.

I can't stress this enough - this plugin can really affect the quality of your recording. It's a great tool, but it's only as great as the ability of the person using it. So, again, if you're having to push this plugin a lot to get a clean recording, fix your recording space first!

Once again, you'll need to use your own intuition to work out what sounds good here. The two sliders on the right, Threshold and Reduction, are you main controls for this plugin.

Just like most plugins with a threshold, the Threshold slider will offset the noise threshold curve. If you move it up and down, you'll see the blue line move up and down. This is where you should use the noise profile (the gray line on your graph) to see where it would be appropriate for your threashold curve to sit. Ideally, you don't want it sitting significantly lower than the noise profile.

The Reduction curve is then where you will altar the strength of the noise reduction. This should be low. Ideally, so low that you can't tell it's doing anything (except softening the noise). If you flick this setting to the max, you'll quickly hear what I mean by ruining a recording.

Once you've found a good spot for it, you can render it out into RX and de-noise the rest of your recording. But hopefully, you don't need it :-)


Whew, that was a lot. And there's so much more.

RX is choc-full of plugins that are there to fix any audio problem. De-reverb, de-clip, de-crackle... You name it. However, if you're finding yourself using any of the other plugins to any considerable extent, it may be worth diagnosing where the issues begin. As I've already said a few times - fix your recording space before even thinking about audio cleanup. It's the reason you're having issues.

The plugins I've covered here are really the most you should be using (I'd even count de-noising out of this). So, hopefully, this gives you a bit more understanding on the software and how you can use it to your advantage. Each and every one of these plugins, when pushed too hard, will poorly affect the quality of your recordings - so use them sparingly!

I hope this all helps.

Peace out. Be kind.


538 views0 comments


bottom of page