Hi Matt, please introduce iZotope and the development team behind RX:
iZotope is a leading audio technology company that develops award-winning audio software and plug-ins for mixing, mastering, music production and live performance. There is a dedicated team that works on our post production product line, including RX.
And how did you get involved with iZotope and RX?
I was doing a combination of sound design (for video games) and dialogue editing (independent film). I was using RX pretty heavily in those scenarios, and as I was living in the Boston area, I got to know the iZotope crew pretty well. My roommate at the time (this was a few years back) played in a band with the Sound Design Manager at iZotope at the time too. It’s a small world. I began working freelance for iZotope, and ultimately came on board full time, where I’ve focussed heavily on their post production product line for the past few years.
How did the idea for RX come about?
RX was originally conceived as an audio restoration solution, at a time when a lot of audio was being transferred from older storage formats. Elegant, great sounding and affordable solutions for removing hum, surface noise, clicks, crackles and other artifacts were few and far between, and this is where the tools in RX 1.0 were primarily, though not exclusively focused.
The visual spectrogram display was always a part of the RX suite, and the ability to see and interact with your audio offered a huge potential beyond this scope, which sets the scene for the story of RX from versions 1 through 4.
RX has a big following in the field recording and post production communities – considering how many audio repair and noise reduction plugins that are out there, why do you think RX stands out?
What makes RX stand out? Taking sound quality for granted, I think it’s the comprehensive suite of tools, and the way it integrates into and speeds up post workflows that makes RX indispensable. It’s rare that we’ll see a user that uses every single feature of the RX suite. Depending on their workflow, different users will use very specific tools without ever using a feature someone else might use daily. Yet it feels like a customized suite of tools no matter which parts they use.
We’ve certainly reached a point where people treat RX as a verb and say “oh, just RX it”
It’s this experience that we work hard to curate effectively, and it’s my hope that this shows. We’ve certainly reached a point where people treat RX as a verb and say “oh, just RX it”.
How long has RX4 been in the works and were there any unusual challenges in developing it?
With the release of RX 3, we reached a large number of new users in the post production community, thus opening many eyes to the possibility of audio repair, which in turn generates many more feature requests. Regardless of the feature request, a common underlying theme seemed to be workflow and efficiency, and as such we wanted to respond quickly in order to better integrate into and speed up people’s daily workflows. Those seconds add up! As such, RX 4 was a year long process. Developing cutting edge DSP always throws up unusual questions or test results. We were certainly able to produce certain unique features solving problems that could only be solved manually before, such as ambience matching.
What are some of the highlights of this new version?
The overall highlight of the new version is really the way RX integrates into your workflow allowing you to spend less time on mundane editing or repair tasks and to spend more time to be creative. Big module highlights of the new version include:
• RX Connect enables you to use the powerful tools in the RX standalone application in conjunction with your audio or video editing software without slowing down productivity.
• Ambience Match, which matches the environment noise of different audio recordings.
• Loudness, which quickly makes any mix compliant with the latest network standards across the world.
• EQ Match, which matches the sonic differences between audio recordings.
In general, what’s your process for deciding what features to include?
It’s a combination of internal and external validation. We ourselves are serious users of our products, and also have a private beta.
RX gives users a very visual approach to sound, and is sometimes referred to as ‘Photoshop for sound’. When RX removes a sound what does it use to replace it?
It depends on which Spectral Repair algorithm is selected. It could be interpolation or resynthesis, or even just intelligent copy/paste.
I’m sure every dialog editor has a story about being asked to remove reverb from dialogue that was recorded in a reverberant space. How did you approach it in your De-Reverb module?
With the Dereverb module, the key was not to simply attenuate tails after a transient, but to try and intelligently identify what was wet and what was dry signal, which is easier said than done since there is often quite some overlap. This initially took over 9 months of testing before we had something we felt was appropriate and useable to continue to develop.
Your Dialogue Denoiser was previously only available in RX Advanced. Now, it’s included with RX Standard. What makes your Dialogue Denoiser different from your regular Denoiser module?
The Dialogue Denoiser is realtime and zero latency for a start, which means it can be run as an online plug-in in any DAW or NLE without sync issues. Additionally, it works in auto mode by default, allowing you to just turn it on and achieve instant results without having to waste time learning, which is key in time-limited post mixing sessions. Additionally, whereas the Spectral Denoiser divides the audio into thousands of frequency bands under the hood, the Dialogue Denoiser uses 64 bands. This lightens the load, but also has a potentially better sounding result on certain pieces of dialogue, as there are less crossover points, though these are pretty transparent.
The Hum Removal plug-in is now “adaptive”. How does that make it different?
Let’s say your hum is 50Hz. Let’s also say you recorded to tape, or are working on an older recording that has wow and flutter. That hum may not always be exactly 50Hz, so a fixed filter won’t do a consistent job of removing it effectively. Adaptive mode tracks a hum profile that changes over time, allowing us to do a better job of removing it accurately. That’s one example.
What’s been the feature users are most excited about in RX4 so far? And is there any feature in RX4 that you’re particularly proud of?
We’ve heard from many users that RX Connect alone was worth the upgrade, before even getting to any of the new DSP or features.
People’s reactions to Ambience Match have been really great too, because it’s kind of an automagical process that saves a lot of time.