The Experts team wondered what we thought people would want to find behind the windows of a dream studio advent calendar. Some of the things we own, some we wish we did. Owned or not, we think any of the things we name in the next 18 days would be a gift to a professional studio owner.

Day 7 iZotope RX Advanced

Mike Thornton talks about iZotope’s audio restoration software RX Advanced. I have been using iZotope RX since RX2. When I started using it, I went searching for tutorials on how to use it, and in the main, all I found at the time were videos that showed the before and after but gave very little clue as how to get there. As a result, 10 years ago, I was asked to produce a series of videos for iZotope to show how you could fix 10 regular issues with RX in less than 2 minutes, and that is how we changed RX from a noun into a verb.

Since I started using RX all those years ago, I have lost count of how many times I have been able to save unusable audio that my clients and friends have brought me over the years using RX from iZotope and the skill in using it that I have developed.

Some highlights over the years have been using Ambience Match and EQ Match to enable my client to cross-cut between two interviews of the same contributor recorded on different equipment on different days. As this documentary was a story without a presenter, it was essential that we could tell the story with actuality and personal testimony from the main character. Without iZotope RX, we couldn't have edited between the two interviews in the way that would best tell the story.

On another occasion, I was asked to clean up a choir recording that had been recorded in a basement with air conditioning. I routinely clean up comedy panel programmes recorded in less-than-ideal locations, as well as self-shot documentaries where the location sound is less than ideal.

Another example was a studio discussion programme that had distortion problems on all the recordings, and it wasn’t just simple clipping but with two different treatments, I was able to retrieve the audio, saving them a complete reshoot,

These stories are just the tip of the iceberg of examples where I have been able to bring audio back from the dead, including saving a complete TV studio programme from being reshoot.

With the release of RX 6 Advanced working mainly in audio post-production, for me, the new machine learning powered tools like De-rustle, De-wind, and Dialogue Isolate are incredible. For example, back in 2017, I had to clean up a video of a baptism on a beach that was originally recorded on a VHS camcorder and then transferred onto DVD with the low frequencies filtered off, but there was still some higher frequency wind noise left. I thought that De-wind wouldn’t be able to fix it, but it did! The dialogue was also embedded in a lot of background noise, and instead of de-noising it, I just ran it through Dialogue Isolate; job done!

Moving onto the music-based tools, it's great to see iZotope adding some more music-focused tools to RX6. The De-bleed module has to be the stand-out module here. I have had to deal with click bleed in the past, especially with larger orchestral recordings, and it was hard work. De-bleed handled it straight away.

RX 7 Advanced came in 2019, and the Dialog De-reverb was one of the standouts for me. I remember when I demoed RX4 in front of around 1,000 folk at a Super Meet event in Amsterdam, and the demo of the then new De-reverb module had 1,000 jaws on the floor. Back to RX 7, and with the help of more machine learning algorithms, iZotope has produced a dedicated dialogue version of their de-reverb module, and from my first test, my jaw was back on the floor. I especially like that it has the same ambience preservation control that we have seen on modules like De-rustle.

The other standout from the RX 7 release was the Music Rebalance module. I first tried it on a track, and I was very impressed with the way I was able to reduce the percussive elements so they didn’t obscure the speech, as well as increase the bass elements which wouldn’t fight with the speech and so produce a version of the mix that was much more suitable as an underscore for the speech and as a consequence be able to retain a good level of music without compromising intelligibility.

When the pandemic hit in 2020, my work outside of the blog has been very different. I have had to learn about video streaming and as well as handling so much self-shoot video content to enable our church to have an online presence, producing two different broadcasts, one for Facebook and one for Zoom.

When it came to the audio side, in 2020, RX 8 Advanced was very busy here in my studio, in many respects, even more than in previous years.

Regular work that year was importing the audio from a video, shot on an iPad using the internal mic or Airpods. Firstly, removing the noise using the spectral denoise module and then using the Loudness Control module in RX 8 Advanced to get the loudness to -16LUFS.

Then there were the times when content had been recorded in all kinds of environments, especially outside using a smartphone, with no windshield, and so De-wind got a serious workout on that one.

There was the repurposed content, which, more often than not, had been mixed with the music bed too high. The improved version of Music Rebalance was invaluable in ensuring that the voiceover or vocals could be clearly understood over the music bed. Then back to Loudness Control to get the loudness to -16LUFS.

Even so-called ‘professionally shot’ videos had issues where every time the presenter moved, we got mic rustle because the mic was being brushed by the presenter’s hair, so De-rustle got a serious workout with me needing to process each hair rustle one by one.

Then, I was approached by a producer responsible for putting together the UK Children In Need telethon charity single because one of the contributors had recorded their vocal part on a smartphone outside in a field. Not only was there a lot of wind noise, but there was a plane flying overhead as well as a lot of other background noise. Using the various modules in RX 8 Advanced, including the new Spectral Recovery module to put the top end back in, as well as developing some new techniques, I was able to resurrect the track, and they were able to use it in the single.

None of this work was on the horizon back in early 2020, but thanks to the very clever people at iZotope, yet again, I have been able to save the lives of some very different clients this year.

RX 9 Advanced brought an improved version of Dialog Isolate with a brand new machine learning neural network, designed to separate dialog more effectively than the previous version used in RX6 to RX8. The RX 9 version is even better than before and can clean up a mobile phone text alert so much more easily than before. But for me, the Dynamic De-hum module was the stand-out in the RX 9 release, making it so much easier to take care of complex buzzes and hum, as well as handling PA feedback much more easily with better results.

RX 10 Advanced brought even more improvements to the RX 9 Dynamic De-hum module, and is much better. Another area iZotope improved was an upgraded Spectral Recovery module, which I have found to be very helpful when handling low bandwidth VoIP recordings.

In the early days, iZotope had very little competition, but nowadays, that is no longer the case. Even so, as I have grown up with RX over the last decade or so, it is still my go-to tool for audio restoration work.