Batch processing audio can save you a huge amount of time and effort. Here's Matt Vanacoro to show you how it's done in iZotope's powerful RX6 Audio Editor.
Ess reduction is aimed at anyone using the Leveler on dialogue or vocals, and utilizes a smart algorithm, inspired by the DBX 902 De-esser, to detect when ess is present in a signal, and then attenuate it accordingly. This avoids adding any boost to esses, which may otherwise be seen as quiet sounds requiring a boost.
RX6 Audio Editor from Izotope has been a saving grace for audio and video professionals everywhere. Nothing cleans out the hiss from a poorly grounded preamp or the rustle from a poorly placed lav mic like RX6. There are times, however, where running RX6 can be very time consuming, and as most of us may know, opening it up 3rd party audio plugins in Final Cut Pro is, well… less than fun. I often have a huge folder of video takes after a shoot that all need the same touch of RX6 love before I use them, and here’s how to process them all at once.
Step 1 - Get it Right the First Time
It’s not just one of my favorite Billy Joel songs, it’s the first step in the batch process! The first thing you’re going to have to do is open one of the files in RX6 and figure out exactly which modules you need to use, and nail down the settings for those modules. If you’ve got more than one module to use (like voice denoise followed by gain to raise the volume up) it may be easier to put them in a module chain to test out the order.
Now, you can adjust the settings in the batch process window later on, but for efficiency, I find it easier to simply save a preset for each module right now. This way I know the settings will be exactly the way I want them, and I don’t even have to open up the specific module’s window.
Step 2 - Line Up the Chain
Go to ‘Window’ and choose ‘Batch Processing’. This will open up the batch processing window for RX6. From here, you’ll want to first ‘add’ the files you plan on processing. RX6 will take audio files, but it will also take video files as well. It will extract the audio from the video file and give you a ‘bounced’ audio file as an output that matches your video file’s length exactly.
After you add your files, line up your ‘chain’ of processing steps. If you’re only using a single one, that’s cool too. It’s on this screen that you can use the drop down menu to select the preset you created for each module when you did your ‘test run’ in step one. If you didn’t save a preset, you can open the settings and tweak them now with the ‘view’ button. Just keep in mind you will need to ‘record’ the custom settings (the button right next to view) before the ‘process’ button in the bottom right will be activated for you.
Step 3 - Output Options
Now that you know what RX6 is going to do to your audio, it’s time to choose where that audio should go once it’s been processed. You can change the file format, choose a different folder, alter the file name.. it’s all here. If you’re batch processing video, it helps to simply keep the same file name, this way if you sort by name every video file will have a corresponding ‘cleaned up’ audio file right next to it. Score!
Move Along
I can’t say enough how much faster processing audio is this way as opposed to trying to tinker with 3rd party plugins in your video editing app. You can even load up your favorite third party plugins to give it that ‘final polish’ before the video ever even hits your editor!
Learn more about iZotope RX: https://ask.audio/academy?nleloc=topic/izotope
Using Leveler For Lead Vocals Izotope Rx 5
Related Videos
Overview
The Leveler module automatically rides the gain in your file to even out the variations of the signal level. The algorithm consists of a compressor with a makeup gain to achieve a smooth signal that’s aiming towards (though may not exactly hit) a desired Target RMS level. The compressor has the ability to prevent pumping on speech pauses or breathing sounds, using the Optimization mode, for either Dialogue or Music, in addition to the Ess and Breath parameters.
The level detector stage includes the K-weighting filter that helps equalize the audible loudness, not just RMS level. However, the Leveler module is designed for the smoothing of overall audio signals, rather than taking an entire signal and using a fixed gain to ensure it hits a loudness compliant LKFS level, which is the goal of the Loudness module.
This all combines to create a transparent, non-destructive Clip Gain curve, without the color or artifacts of a traditional compressor. Unlike the Loudness module, which applies a constant gain based on Loudness compliant analysis to the whole file, Leveler applies a time-variable gain. For convenience of RX users, the time-variable gain is applied as a Clip Gain envelope, which can be viewed and edited by the user.
Please Note
If additional processing is applied from another module after running Leveler, the clip gain values assigned by Leveler will be destructively written to the file and the clip gain nodes will return to zero. However, the clip gain settings from Leveler will be saved in the Undo History list).
Controls
NUMERICAL READOUTS: Numerical Readouts provide you with the Total, Maximum and Minimum readouts for RMS.
The total value is the overall RMS of your audio signal, which may inform where you choose to set the Target RMS level parameter.
OPTIMIZE FOR: Optimize For switches between two modes, Dialogue and Music. Each mode utilizes a slightly different handling of the noise floor.
Dialogue tends to be audibly juxtaposed against the noise floor, as it’s typically very transient, whereas music often tends to fade into the noise floor, with chords, notes, and other instrumental decays.
Switching between these two modes will affect the behavior of the Leveler, and prevent pumping.
TARGET LEVEL: Target Level sets the desired average RMS level of the recording.
Note that Leveler uses K-weighted RMS to better level perceived loudness, but that it is not a loudness compliant leveling tool. It uses the Target Level as a guide, but with the goal of smoothing out variations in an audio signal much more transparently than a compressor typically would. As such, it is not unusual to see the resulting output of Leveler not be an exact 1:1 with the defined Target Level.
At high target levels, the leveler may not be able to hit the target without clipping, so the target level will not be reached.
RESPONSIVENESS: Sets the integration time for RMS level detection and is similar to the attack/release setting on a compressor.
Lower settings will result in more aggressive Leveling, useful if a signal has a lot of sudden variations.
Higher settings will result in smoother behavior, leveling words or phrases rather than individual syllables.
If you find the Leveler is responding to any sudden unwanted sounds, such as a cough, and boosting it, increase the slider to a higher value to see if this results in less aggressive jumps.
PRESERVE DYNAMICS: This can be thought of the maximal amount of gain applied by the Leveler. The wider the range of gain adjustments allowed, the further away from the original dynamic range the audio signal will be.
At lower values, the Leveler will preserve fewer of the original dynamics in the audio signal.
At higher values, the Leveler will preserve more of the original dynamics in the audio signal.
ESS REDUCTION: Ess Reduction is aimed at anyone using the Leveler on dialogue or vocals, and utilizes a smart algorithm, inspired by the DBX 902 De-esser, to detect when ess is present in a signal, and then attenuate it accordingly. This avoids adding any boost to esses, which may otherwise be seen as quiet sounds requiring a boost. The slider sets the amount of ess reduction, applied in dB.
BREATH CONTROL: Breath Control will automatically detect breaths in your vocal takes and attenuate them. This saves time when editing dialogue or vocal tracks, and streamlines a task that is typically done manually.
Breath Control automatically analyzes the incoming audio and distinguishes breaths based on their harmonic structure. If any piece of the incoming audio matches a harmonic profile similar to a breath, the Leveler will apply a Clip Gain adjustment.
Different from a ‘Threshold’ based process in which the module is only engaged once the audio has risen to a certain volume, this feature will perform its analysis regardless of level.
This allows for accurate breath recognition with a multitude of quiet or loud dialogue / vocal styles with minimal adjustment of the module’s controls.
The slider represents the desired level, in dB, that you wish all detected breaths to be reduced to. This can result in much more natural sounding breath reduction as the detected breaths in your audio are only reduced when necessary.
Loud and abrasive breaths will be reduced heavily, and quiet, natural sounding breaths will be left at the same volume. The volume level specified by this slider is a guide, but may not result in exact values.
LIMITER: The Leveler has a built-in Limiter in order to avoid introducing any clipping to the audio signal once the Clip Gain envelope has been applied.
This cannot be adjusted, but you’ll see the Clip Gain envelope smooth off an audio signal if you’re pushing peaks close to 0 dB.