How to De-Ess a Vocal with iZotope Alloy 2
What you want to do is use a processor that’s going to turn down those parts of the vocal performance that sound harsh. And the rest of the performance to pass through unattenuated. The old school way of doing a de-esser goes back to the days of the analog console, where what the mixing engineer would do is use the equalizer setting to intentionally dial in an EQ curve that made the vocal sound extra harsh, and what you would do is sidechain that equalizer into the onboard compressor. Whenever the vocal got too harsh, the compressor would just turn the volume down, or the amplitude down, during that time period of the signal.
Now that we’re in the digital era of recording we’ve got these great plugins that can accomplish the same exact things, and in many ways even do it better and improve upon the old school way of doing it.
I’m gonna be showing you iZotope Alloy 2 because it’s got great controls and it’s actually one of my favorite ways of doing de-essing. So, let me bring up the plugin right now you and show you how I like to use it. To start out with, I’ve got my vocal over here. It’s a female vocal. We’re gonna be focusing on different parts of the song and show you how I like to really control those utterances that can be harsh.
I’ve switched over to the de-esser module of the plugin. I’ve turned everything else off. First thing I like to do is watch the visualization up here. That’s one of my favorite parts of the plugin. You can see the specific frequency range during those utterances that can be harsh. What you’re focusing on then is the high frequencies up here. So I’ll find the part of the song where the singer is making the sound “Sss.”
[music + female vocals]
You’ll see there’s a lot of high frequency content up here around 10k to 12k. So I’ll just dial this in. You can even solo it and then solo inside of it and just focus on the area here.
[solo female vocals]
This way I’m dialing in very narrow settings for the filter so I can really focus on the frequency content where that sound is occurring. Next is just a matter of dialing in the threshold here.
I’ve switched to broadband mode because I don’t just want the compressor to turn down these frequencies, I actually wanted them to turn down all of them. I just want it to respond to this frequency range. Next let me play you a different part of the song, and show you what it’s like for the “shh” sound.
What you’ll see is the difference between “Sss” and “shh” is there’s actually frequency content at a different location for the “shhh” sound than the “Sss” sound. Now the old school way of doing it with an equalizer side-chained to a compressor is that you basically have to cast a wide net. And you have to basically capture all the frequencies that can sound harsh and feed them into the compressor. What I like to do with plugins and having a lot of these fine detailed controls, is just bring up another version of the plugin, and set it up specifically just for the “Shh” sounds. So I’m having these plugins specialize and focus on these different kinds of utterances and do it that way rather than just having one plugin do all the heavy lifting. So it’s like this.
[female lead vocals]
This is between 6k and 8k. Just a matter of setting the threshold then.
I’m even gonna speed up the attack and the release because I don’t really like it to take a long time to respond. I’ll bring them up side by side and show you what these different utterances, what they look like through the processor. Start with the “Sss”.
Next up, “Shh”.
So that’s the way that I like to use these plugins to de-ess a vocal. It gives me a lot of very detailed control over dialing one exactly how much compression and de-easing I want to take for these different utterances. That’s really all there is to it. I’ll play you back some parts of the song and allow you to hear it out.
[music + vocal de-essing]