RX has many algorithms and processes for cleaning up audio, but it also offers two distinct modes of operatio—as a plug-in and as a standalone app. Certain modules like Spectral Repair are standalone app only, while tools like Voice De-noise work in your DAW.
So when do you use RX as a plug-in or in standalone? Ultimately, it’s your call. But it pays to be informed. Let these recommendations be your primer, no matter your DAW!
A word on standalone integration
Many DAWs are similar in this regard—they require you to use RX in a corollary fashion.
Others, like Pro Tools, utilize RX Connect. RX Connect allows you to send audio to the Standalone application as needed.
Whatever you do, I’d advise the following: always back up any file you plan to process in the standalone application. This way, you can always go back to the original, should something go haywire.
If you’re trying to eliminate a relatively static, unobtrusive room ambience, try inserting RX’s Voice De-noise module as a plug-in within your DAW. Say we’re trying to remove some slight machine whirr from a nearby hard drive, or a fridge buzzing away from some other room. In this case, you can simply load Voice De-noise as a plug-in, have the algorithm learn the noise during a moment of silence, and tweak from there.
Should the noise change throughout the recording—if the hard drive suddenly whirrs louder, for instance, or if the fridge compressor kicks on mid take—it’s better to engage the standalone app for de-noising. Depending on your DAW, you can use an external editor, RX Connect, or, in the case of Pro Tools, simply apply the AudioSuite to the region that needs fixing.
I would advise using De-hum similarly to Voice De-noise. It works well as a plug-in for tracks that have a consistent hum. If you’re working with a track that only has hum some of the time, use in standalone or AudioSuite (in the case of Pro Tools) to better adjust the problematic region.
If you have a vocal with the occasional pop, running De-plosive as a plug-in process should work most of the time. I would insert the plug-in before any vocal denoising, EQ, or compression.
However, not all pops in a track are the same. Sometimes, one pop “hits” higher up in the frequency spectrum than another. Other times, the dynamics of the pop don’t trigger the settings you’ve utilized for other instance.
Here it pays to be selective: try to AudioSuite the plosive, or take it over to the standalone app if you’re using Logic or a similar DAW. In RX Audio Editor, tweak the settings to treat the specific plosive, or run the process subtly a couple of times (this can help in certain situations). Also, on the occasions where De-plosive might add an artifact, you can easily take it out with De-click.
A dialogue track for a film, podcast, or radio broadcast holds different considerations for mixing. An overly polished sound with crispy, expensive high end is less important than character, mise-en-scene, and intelligibility. Furthermore, the vocal often sits in a sound-bed of noise: sound design, room tone, or music can distract the listener from imperfections, allowing you to work quicker without fear of obtaining “the perfect result”—simply put, no one will notice.
In these instances, run Mouth De-click as a plug-in first and see where that gets you. Chances are, this will be all you need. Again, use this plug-in before any vocal de-noising, compression, or EQ.
Music differs from broadcast in that you often strive for the impeccably polished song. Because of this, I rarely use Mouth De-click in a music track. Instead, I pull the track into the standalone application and use this process selectively on the passages that need them. You’ll begin to recognize them by sight, they look like this:
Again, Pro Tools lets you tinker with settings in real-time with an AudioSuite plug-in. That might be the best way to go if you’re trying to save time here.
Breaths are annoying not because they’re hard to edit, but because they’re very time-consuming to edit. Breath Control is meant to automate the process to some degree, which is why I’d recommend using it as a plug-in (before any de-clicking, de-noising, EQ, or compression). The glory behind something like Breath Control lies precisely in using it as a plug-in insert.
Here’s a tip for you: if Breath Control is only catching a majority of the breaths, you can easily automate parameters for the stray offender. I would advise trying this exactly once before you move on to manual editing, because at this point, it will take the same amount of time.
I get the best use of De-clip in both the standalone app and the DAW plug-in. Yes, with the right settings, it can leave everything I don’t need de-clipped untouched—but I’d rather not process all the audio for a couple of selections worth of clipping. That offends my Hippocratic sensibilities.
I’d rather deal with the selections that need de-clipping on an individual basis. In Logic Pro, I’ll open up the external editor with a key command (Shift+W by default, after you set iZotope RX to be your external editor). In Pro Tools, I’ll run the plug-in in my DAW, or use RX Connect.
Again we reference the Hippocratic oath: “First do no harm” is a good motto for audio. While De-click can be employed in a gentle way, I choose to err on the side of caution by either using the RX plug-in for specific clicks, or porting over to the standalone app through AudioSuite or RX Connect (in Pro Tools) or external editors (in Logic).
However, if time is an issue, you can get away with de-clicking something as a plug-in, especially if heavy processing will be induced downstream. Say I have a really clicky DI bass that will then get the fuzz treatment. I may slap de-click as the first instance and just go about my processing from there.
Ideally, you wouldn’t use Ambience Match on entire tracks, because you’re matching one out-of-place audio source with a reference. The process, in my experience, excels on specific regions—on instances where you need to match ambiences to sell an edit. This could be a line of film dialogue looped in post-production, or a guitar part suddenly missing believable amp hiss.
Making matters more complicated, the algorithm must learn the source ambience for the process to work—getting it to learn the right bit of material can be problematic within your DAW for a variety of reasons. Maybe the source audio plays on a different track—now you need to apply two instances of Ambience Match. Maybe the ambience you’re sampling only lasts for a split second—some DAWs don’t let you loop audio that flies by so quickly.
For all these reasons, I’d advise going with standalone or AudioSuite use over-processing in real-time with the plug-in. Within Pro Tools, I’d rather AudioSuite the good ambience with the learn button, and then hit the target audio with AudioSuite again. For Logic, I’ll use the standalone application through the external editor.
You can use De-reverb as a plug-in, but if you do so try to use it sparingly, after any de-clicking and before any de-noising. This process is best employed as subtly as possible, lest you risk summoning the dreaded Space Monkeys (whooshy-sounding artifacts). It also has a Learn feature.
For the reasons outlined in the previous tip, I prefer to use De-Reverb in standalone—it’s easier to sample the source audio that way.
The use-case varies depending on what you’re de-crackling. If you’re sampling an old vinyl record, and you want to clean it up a little, De-crackle as a plug-in works well. The same is true for an interview recorded over the phone: if there’s a bit of static crackling, a plug-in is your friend.
However, if you come across pieces of audio that need drastic de-crackling in a few spots, and no processing most of the time, I go to my combo of AudioSuite/external editor depending on the DAW.
Take this phone call recording, for instance, from a podcast I recently did for NPR:
We’ve got noticeable distortion only on a couple of words. However, we can use De-crackle on certain words to make it better, like so:
Here, selective processing benefitted us.
RX’s De-ess is exceedingly powerful and transparent. You can also automate it—and automating de-essers can be necessary at times. So try it as a plug-in first. If you’re going to go the route of offline de-essing, I would advise pulling down the gain of each individual ess rather than processing them; that is frequently the most natural way forward.
The AudioSuite Family
De-rustle, Dialogue Isolate, and Music Rebalance—these modules are offered for use as AudioSuite processes in Pro Tools. If you use Pro Tools, you can choose to forego the RX Connect route and apply these algorithms in real-time in a region-specific way. This is exceedingly useful, however, I must recommend again that you backup your unprocessed audio files!
A plug-in on a track affects the whole track. Using AudioSuite on a region, or the Standalone application via external editors, is far more selective. Plug-ins tend to be faster. You tweak the plug-in and you’re good to go. Conversely, standalone/AudioSuite use slows you down, as you’re taking more time to home in on specific bits of audio.
You ultimately make the call for how to use these plug-ins. I only ask you to balance efficiency with efficacy: how will you get the most out of the process in the least amount of time? Or, look at it through a different lens—does your audio require more of a time investment for the purposes/deadline at hand?
Ask yourself these questions when cleaning up audio, and you’ll make better decisions when reaching for the plug-in, or opening up the application.
And backup your clean file. It bears repeating. It always pays to be safe!