In this article we discuss 4 ways to deal with Sibilance in your vocal recordings. From spoken word recordings like podcasts to vocal performances learn how to treat it.
Have you ever been listening to the radio or TV, and the person talking on air always seems to hurt your ears when they say a word with the `s` sound? Or perhaps the Radio or TV distorts when these `s` sounds come through the speakers. Well this is what you call sibilance and it is going to be the topic for today’s article and more specifically, how to deal with it in your productions.
This process of `dealing with` sibilance is often referred to as `De-Essing Vocals` and is a powerful & important technique to master. We will run through how to De-Ess vocals using four different methods. First we will look at De-Essing plugins, how to use them and what the result sounds like. Next we will look at the sidechaining method and how to set this up on a vocal. Finally, we will try two manual way to achieve De-Essing, one with automation and another with multiple tracks.
But first, technically speaking, what is sibilance? Well sibilance occurs most often when consonants like S, T and Z are spoken since the air pressure of the voice leaving the mouth is quickly constricted by your teeth and tongue. This causes a `burst` of air pressure and so you get the harsh sibilance frequencies often around the 5kHz to 7kHz range. It is like a hard and sharp hissing sound which accompanies the consonant of a word.
So now that we know what sibilance is, how do we deal with it in our productions?
note: the following methods do not consider the microphone type and setup, they assume every effort was made to reduce sibilance at the source, as possible.
Here is the original recording of a sentence with sibilance, play this recording before comparing to the processed/De-Essed recordings below.
Using a De-Esser plugin
The easiest way to deal with sibilance is with a dedicated De-Esser plugin like the Lisp plugin from Sleepy Time which you can download for free as a Windows VST here. The plugin is super flexible and can be left on auto or set manually to target a specific frequency range. For De-Essing and there is even a listen button that allows you to hear what frequencies are being targeted. So all you need to do with this plugin is to increase the reduction amount and listen in for the sibilance sounds in the recording to get a setting of your liking.
So what happens when we apply the Lisp plugin with the above settings to our audio clip?
When you compare this with the original recording at the top of the article, you can hear a slight reduction in the sibilance sounds and the recording is less offensive to the ears. This method is by far the easiest and least time-consuming however, you may be sacrificing control over the De-Esseing since the plugin takes care of everything for you. OK lets try another method!
This method is much more complex than the previous but it can still be considered an automated one. As we get into it, you will be able to see that really all we are doing it replicating what the Lisp plugin above did.
What you will need is two identical tracks with the recording in both. The top track we will consider the `Master` and the lower the `Sidechain Input`. We place a multi band EQ on the Sidechain Input track and apply a band pass filter to the recording from 4kHz to 8kHz. From here what you will get is just the frequencies between these two points coming through. What you want to do is fine tune these so that all you can hear is the sibilance you want to get rid of. Now on the Master track we will apply a compressor and route the Sidechain Input tracks signal into the compressor, so that it reacts to that signal. Now with the Sidechain Input track muted listen to the Master track and apply compression to it. Since the compressor is only reacting to the Sidechain Input track, compression will occur when the sibilance signal you dialed in earlier comes through.
Check out the results of the Sidechain method and compare to the original at the top of this article.
The following methods are quite easy to set up compared with the above sidechain method but, can be quite tedious to implement since you have to go through and draw in the automation (De-Essing levels) yourself manually. This can end up taking a lot of time to do and could result in some missed areas. But if you put in the time you can end up with a better result since you have ultimate control over what is going on.
The gain method is straightforward where all you need to do is automate the reduction in gain around the sibilance areas. Take care not to apply too much gain reduction as this can result in a choppy sounding recording because, the gain is `bouncing` up and down too much and/or too quickly. Here is an example of gain reduction automation and I have purposefully applied a little too much gain reduction for my liking.
The EQ method is very similar to the above Gain method where you will need to manually draw in the automation lines where you want reduction to occur. There is however one difference here, rather than reducing the gain across all frequencies, you are able to pick the frequencies you want to reduce. With patients this method can prove to be one of the best methods for removing harsh sibilance in a recording.
To do so, all you need to do is set up a track with a multi band EQ, dial in the frequencies you want to remove, then draw in the gain reduction automation lines for these frequencies. This is great because say if you have two people talking in the same recording with different sibilance sounds, you can set up two EQs with their own automation for each person. Keep in mind that this will take longer since you are having to manually configure two EQ plugins with automation. Remember that sibilance often occurs around the same frequencies so the pay off for doing it this way may not be worth it if you are in a rush.
Check out the EQ method below and compare with the original.
Our final method for removing sibilance is to use multiple tracks to split out the sibilance sound from the rest of the track, and apply effects accordingly. To set this up first have your `Master` track setup, then listening to the recording in your DAW, cut out the areas of sibilance and move them to the second track. Once you have done this Apply gain or a combination of gain and EQ reduction to the second track. The result when they play together will be a De-Essed track with softer sibilance sounds when compared to the original.
Check out how this method sounds by comparing it to the original at the top of this article.
The Wrap up
So today we have discussed 4 different method for De-Essing and head examples of how they work compared to the original and now we can compare them to each other. Which one to you, sounds the best? I admit that with these example recordings I have over emphasized the De-Essing to help illustrate the example we were discussing.
At the end of the day, the method that works best for you, is the one that sounds the best to you. I personally prefer the sound of a combination of the EQ and Gain methods combined. However, if you were a podcaster who needs to get a hours worth of discussion posted within a very short turn around, then perhaps a De-Esser plugin or the compression method would be best for you.
Like any other art form, it takes practice and lots of it! The best way to get better at something is to do it, so go ahead and grab yourself some vocals, and try out these methods for yourself!
Let me know if you like this article by clicking the like button below, and let me know if there is anything else you would like me to cover, in the comments section.
Thanks for reading!