Sibilance is the bane of any mixer since many times it’s so hard to correct. While modern de-esser plugins work admirably, they usually only hit one frequency band, while most sibilance occurs at two frequencies or more. Enter the Antares Auto-Tune Vocal De-Esser that not only affects multiple frequency bands, but also uses artificial intelligence to help do it.
The Auto-Tune Vocal De-Esser has a simple easy-to-use interface comprised of a frequency display, a Threshold control and two main frequency controls. The first frequency control hones in on the S-Sh-Z parts of a vocal, and the second on the T-Ch-K part. These both can be linked together with a Link selector in between the controls. An audition button allows you to hear exactly what’s being attenuated.
An AI Helping Hand
The clever part here is the Assist button. Click it and play 5 seconds of the vocal and it will suggest a threshold level that you can either apply or not. Click it a second time and it will set what it thinks is the ideal settings for both the S-Sh-Z and T-Ch-K controls.
The display shows you which frequency bands are being attenuated in blue and red, and a Gain Reduction meter on the right tells you the overall amount of de-essing that’s occurring.
As you’re probably aware from the hallucinations of Large Language Models like ChatGPT, AI doesn’t always give you what you need, only what it thinks you want. That’s apparent when it’s applied to many facets of audio, but vocal de-essing is one place where its application is most welcome. De-essers can often be finicky in setup, working on one part of a vocal while not on another. Anything that will make setup go faster should be something that a mixer will really embrace.
Antares Auto-Tune Vocal De-esser costs $99 with a 14 day free trial period. It’s available for Mac or PC in all plugin formats. You can find out more here, or watch the video below to see it in operation.
