- Pro Video
- Lighting & Studio
- Pro Audio
- TVs & Entertainment
- A/V Presentation
- Shop Categories
- Used Dept
When you’re shooting a video or recording music, it’s generally considered taboo to rely on fixing problems in post, rather than getting them right the first time. After all, if you record something properly, there will be no need to fix it later. This is a good rule to follow; however, there will still be occasions where unwanted sounds get recorded, even if you do everything possible to avoid them. When fixing something in post is a necessity, many professionals turn to iZotope RX 2 and RX 2 Advanced as their secret weapons for making troubled audio sound its best.
iZotope RX 2 is software that can run as a stand-alone application, or as a plug-in within a host (it’s both Windows and Mac compatible). It features five sound-fixing “modules” that can be used on their own or in any combination, and applied to tiny snippets of audio, or to entire lengthy recordings. It also has a powerful graphical “spectrogram” display that transforms the act of analyzing and repairing audio into a visual task. The spectrogram makes operating RX 2 feel comfortably familiar; it’s similar to editing images in Photoshop (there are even Lasso, Brush and Magic Wand tools that are simple and intuitive to use). RX 2 was designed for users of all levels of expertise, from total beginners to working professionals. iZotope provides a very clearly written guide entitled Restoring Audio with iZotope RX, that you can download and read to familiarize yourself with using this powerful restoration software.
I had the chance to test out RX 2, and I was very impressed, to say the least. I had never used the software before, so I dedicated fifteen minutes or so to reading the guide, which was clearly written by Nat Johnson. The guide did a great job of getting me up to speed with the basics of the software. The key elements you need to know about are the Waveform/Spectrogram Opacity Slider in the lower left-hand corner of the graphic user interface (the name of this tool sounds intimidating, but don’t worry, it’s easy to use). This tool allows you to change the view of the loaded audio from a basic waveform to the spectrogram display, which shows you changes in pitch on the vertical axis.
The first audio file that I loaded into RX 2 was a short recording of my own voice. Even though I had recorded this in my home studio, I was surprised to hear audible noise in the silence between my words. The guide suggested that I try the Remove Hum module and the Denoise module to fix the audio. I had spent some time studying RX 2, so I already knew that I needed to drag the cursor to highlight a couple of seconds of the audio that had no vocals, just the background noise. I then clicked the Learn button in the Remove Hum module. Next, I clicked into the main window of the software and highlighted the entire audio clip with the token Command + A shortcut (Control + A in Windows). The shortcut happily worked like a charm, so I clicked the Process button. The Remove Hum module quickly processed the entire audio file. I played it back, and I wasn’t terribly impressed with the result. However, I undid that work, and went through the same steps using the Denoise module. This time the result was striking. RX 2 had completely eliminated the noise, and my vocal recording still sounded full and rich.
It was immediately apparent that RX 2 would have many uses outside of just restoring recordings of old tapes and vinyl records. It had taken a sloppy vocal recording that I had made with a modern computer audio interface and a good microphone, and made it sound ten times better. I had only been using the software for five minutes and my mind was already blown. I was only scratching the surface, as there were several more modules to explore.
The next file that I loaded into RX 2 was a recording of a vinyl record. I had captured the vinyl using a low-priced, modern USB turntable. I had recorded the audio with Audacity, a program that anyone can download from the Internet for free. The recording had all of the crackles and pops that you would expect an old, poorly maintained vinyl record to have. Also, when I made the recording, I couldn’t get a very strong level in the software. The waveform of my recording was rather small and weak, but the music was there.
Again, referring to the guide, it was suggested that I try the Declick & Decrackle module. The window of this module has a Presets drop down menu. I opened the menu and selected the Vinyl Record preset. Then I simply used the Command + A shortcut to highlight the entire recording, and I clicked the Process button. This time it took my 2.4 GHz i5 Dual-Core MacBook Pro a little over two minutes to process the entire file (the stereo recording was around eight minutes in length). It was worth the wait. The Declick & Decrackle module had found and removed 26,630 clicks from the recording. The music was much cleaner sounding on playback. You could still hear a little bit of vinyl crackles, but as a huge vinyl fanatic, I didn’t mind this at all.
With most of the pops and clicks removed, the level of my captured vinyl recording was still too low. Without referring to the guide, I opened the Gain module in RX 2. I selected the entire file with the Command + A shortcut and adjusted the gain slider control. The size of the waveform did not change in real time when I adjusted the gain level (as it does in some audio software). I needed to be able to hear what the audio sounded like, so I would know if it was too loud or too soft, so I clicked the Preview button in the module. Now I could hear the recording, and how it reacted to the gain slider, and see the audio level rise and fall in the software’s meter. I chose a setting that was full and loud (but not too hot), peaking at -6dB on the meter. I clicked the Process button, and seconds later the waveform of the recording was nice and fat.
Boosting the gain was a simple fix that made everything sound better. However, I could see in the waveform that there was one spike that had clipped. I moved the playhead to that spot and listened. It sounded like I had accidentally bumped the record player as it was recording, because there was a heavy, low frequency thump in that spot that obviously didn’t come from the mellow music in the recording (the song was Here, There and Everywhere, by The Beatles).
I needed to get rid of this thump, so I opened the Spectrum Analyzer module in RX 2 to visually identify which frequency was out of control (I had learned how to use a spectrum analyzer by using the Spectrum device in Ableton Live). The thump was really low and I could clearly see it jumping out at 40 Hz and below on the Spectrum Analyzer. I opened up the EQ module in RX 2, and highlighted the unwanted peak. I fiddled around with the EQ bands for a minute or two, only cutting the frequencies below 50 Hz on that short spot in the recording. I clicked the Preview button, and it sounded good, so I processed it. The spike in the waveform went away, and everything sounded good.
With my digitized copy of Revolver sounding sweet, I decided to move on. I opened up a recording of an electric bass guitar I had made years ago. It was a short, six second clip that I had used as a loop. It was a cool bass line, but it was poorly recorded. The signal was clipped, and the sound was covered with unpleasant sounding digital distortion. The clipping peaks were obvious to see in the waveform (the waves were too tall and squared off at the edges). Since there was distortion all over this recording, I opened the Declip module and highlighted the whole thing. There is a Presets menu in the module, and I chose the Digital Clipping preset and clicked Preview. It was amazing. All of that nasty, embarrassing digital distortion was gone, and my bass guitar still sounded full and rich.
I kept the audio playing in Preview in the Declip module. I wanted to hear what the other presets did to my bass line. I was happy to discover that RX 2 let me preview all of the other presets in real time. I just clicked around as the bass line played and listened to how the different parameters altered the sound of the recording. It was fast and easy to find the settings that made the audio sound its best (which turned out to be the Digital Clipping preset). I clicked Process, and my once squared off and badly clipped audio file looked and sounded like it had been properly recorded.
It’s difficult to express the profundity of the innocuous and easy-to-use Declip module. When you work in the pro audio department at B&H, there are a few questions that you get asked over and over again, and many of them have to do with fixing problems in audio after its been recorded. The stock answer to “The sound on my video is too loud and distorted, can I use EQ to fix it?” used to be: “Sorry, it’s incredibly difficult and often impossible to remove digital distortion after it’s been recorded. The only thing you can do is to try to correct it with some expensive restoration plug-ins.” RX 2 changes this completely. While there are still going to be distorted audio files that the Declip module can’t save (hey, it’s not a miracle worker), it does rewrite the audio rulebook about what is and isn’t possible.
RX 2 is also incredibly useful for fixing audio issues that arise in video production (and as we know, there’s no shortage of those). The Declick & Decrackle module that so effectively cleaned up my digitized Beatles album is also very good at removing the rustling and contact noises that are created when lavalier mics are hidden under clothing. Ambient noisemakers like ventilation ducts, HVAC systems, idling vehicles and generators have long been the bane of location sound recording for film and video. RX 2 enables you to eliminate these sonic pests in post production, too. And when your on-screen talent switches from whispering to screaming in a millisecond, the Declip module with be indispensable.
The tools in RX 2 are also handy for creatively modifying the sound in a video project. The Spectral Repair module enables you to alter some kinds of intermittent sounds in recordings to your liking. When you’re editing a project, and specific sounds start to bother you, RX 2 may be able to remove them (or let you move them around). For example, if you shot a scene of a person getting into a car, and in post you decide that you don’t like the chiming sound that the car made to alert the driver that the door was open, the chime sound can be isolated and removed, without negatively impacting the other sounds in the mix.
The Spectral Repair module can also patch holes in gaps in audio. Sometimes you end up with a chunk of silence in a recording (which is usually caused by some kind of digital glitch). RX 2 is so powerful that it can not only remove deeply imbedded sounds, it can generate sounds to make audio dropouts disappear, too. I gave this a try with a recording of a song with vocals and piano. The musician intentionally paused between two verses, and I used Spectral Repair to fill in the pause. I highlighted the silent part of the recording, and I also highlighted a little bit of sound on either side of it. I clicked the Replace tab in the module. It took my laptop a few minutes to process the job, and the result was somewhat strange sounding (it kind of sounded like the song becomes haunted for a moment). But it did successfully replace silence with audio that sounded somewhat like the rest of the track. This is obviously a feature that requires a little more practice and experimentation to master, but it was interesting nonetheless.
Being able to pick out certain sounds and alter them is useful in music production as well. If you’re recording an electric guitar player, and your mic picks up the humming sound of their amplifier, RX 2 enables you to easily remove the hum, without compromising the overall quality of the guitar recording. On top of that, you can even remove the squeaking sound of the players’ fingers sliding across the strings (if you find it distracting). Keep in mind that you have to be gentile when making these kinds of alterations. Too much of a heavy hand ends up altering your sound for the worse. It’s definitely possible to muffle your recordings with these tools, but if you’re careful you can do some amazing things.
The basic version of this software, RX 2, is available at an incredibly reasonable price. It’s capable of doing everything that was described in this article. If you need ultimate power, RX 2 Advanced is available for quite a bit more. The advanced version adds an additional module called Deconstruct, an Adaptive Mode to the Denoise module, 64-bit sample rate conversion, MBIT+Dithering, Radius Time and Pitch control, export as XML capability, azimuth alignment for tape restoration, and third party plug-in support. That last feature is pretty amazing. You can highlight a tiny portion of a single harmonic element of a sound, and apply a third party plug-in just to that material. That’s some real audio muscle.
RX 2 gave me two major revelations. First and foremost, many of the things that I used to consider impossible to fix in audio are now easily fixable. The other thing I realized was how often I create recordings that need help. Seriously. Just about every file I brought into RX 2 needed some kind of surgery. What gives? I’m supposed to be good at this stuff! If you use software to create music, podcasts, news reporting, or you mess around with video soundtracks in any way, I can confidently say that RX 2 is an audio tool that’s absolutely essential to have. Give it a shot and hear how good you can sound.
|RX 2||RX 2 Advanced|
|Operating System||Windows: XP, x64, Vista, 7
Mac: OS X 10.5 or later (Universal Binary)
|Windows: XP, x64, Vista, 7
Mac: OS X 10.5 or later (Universal Binary)
|Plug-In Formats||Pro Tools 7+ (RTAS/AudioSuite/AAX), VST, MAS, Audio Unit, DirectX||Pro Tools 7+ (RTAS/AudioSuite/AAX), VST, MAS, Audio Unit, DirectX|
|6-Band Parametric EQ||•||•|
|Channel & Phase Operations||•||•|
|Lasso, Brush and Magic Wand Selection Tools||•||•|
|Stand-alone Application and suite of plug-ins||•||•|
|Adaptive Mode in Denoiser||•|
|3rd Party Plug-In Support||•|
|Spectral Repair Multi-resolution Processing||•|
|Advanced Denoiser Options for Precise Control||•|
|iZotope 64-bit SRC (Sample Rate Conversion)||•|
|iZotope Radius Time and Pitch Control||•|
|Export History as XML||•|
|Azimuth Alignment for Tape Restoration||•|
*Spectral Repair is not a real-time plug-in and therefore not compatible with some hosts, including Apple Logic and GarageBand, BIAS Peak, Ableton Live and Sony Acid.