What Does a De-esser Do? (And When You Don't Need One)
The short version
A de-esser is a compressor specialized for vocals: it only listens to the sharp S and T sounds, and turns them down when they get too loud. It's a repair tool for one specific problem, not a plugin every vocal needs.
On this page
- What does a de-esser do to vocals?
- Wideband or split-band: which kind is it?
- What frequency should a de-esser be set to?
- Do you need a de-esser at all?
- De-esser vs dynamic EQ: what’s the difference?
- Does your DAW already have a de-esser?
- Can you use a de-esser on acoustic guitar or the master?
- Key takeaways
- Our recommendation
A de-esser turns down the sharp “s”, “sh” and “t” sounds in a voice, and only at the moments they get too loud. Everything else in the performance passes through untouched. It’s a repair tool for one specific problem, not a plugin every vocal needs.
If you’ve ever pushed a vocal forward with EQ or compression and found every “s” suddenly hitting like a small whistle, that’s sibilance, and this is the tool built for it. Here’s what it actually does, where to set it, and when the better fix is somewhere else in your session.
What does a de-esser do to vocals?
A de-esser is a compressor specialized for vocals. It only listens to one part of the frequency range, the band where sibilance lives, and when the level there crosses a threshold you set, it turns the signal down for as long as the “s” lasts.
That’s the whole trick. A normal compressor reacts to how loud the voice is overall. A de-esser reacts to how loud the esses are, so a quiet but harsh “s” can still be caught while a loud, warm vowel goes through untouched.
Sibilance itself is the hiss of air through the teeth in consonants like s, z, sh and ch. Every voice has it. It becomes a problem when it’s louder than the rest of the voice, which brighter microphones, close mic positions and top-end EQ boosts all make more likely.
Wideband or split-band: which kind is it?
De-essers come in two main types, and the difference is what gets turned down.
- Wideband turns the whole signal down when an “s” is detected. It’s simple and sounds natural on light sibilance, but heavy settings can make the whole voice dip on every consonant.
- Split-band turns down only the sibilant frequency band and leaves the body of the voice where it was. It can work harder before you hear it, but pushed too far it makes an “s” sound like “th”.
Many de-essers let you switch between the two.
What frequency should a de-esser be set to?
As a starting point, female voices usually carry sibilance between about 5 and 8 kHz, and male voices between about 3 and 6 kHz. Those are the ranges Apple gives in its own guide to Logic’s de-esser.
Treat them as a place to start listening, not a setting. Sibilance depends on the singer, the microphone and the room, so two voices of the same range can need very different bands.
The quickest way to find it:
- Loop a phrase with obvious sibilance in it.
- Use the de-esser’s listen or monitor mode if it has one, which lets you hear only the band it’s detecting, and sweep the frequency until the “s” is loudest and the vowels have mostly gone.
- If there’s no listen mode, put an EQ with a spectrum analyzer before it and watch where the peak lands on each “s”.
- Set the threshold so it only reacts on the problem consonants, then check the whole song, not just the loop.
How much reduction is too much? The best rule of thumb is the simplest one: if you can hear it working, back off. That goes for most corrective tools, and Auto-Tune is the famous example of what happens when nobody follows it.
Do you need a de-esser at all?
Often not, if the problem can still be fixed where it starts. A de-esser is a repair tool, and the best time to deal with sibilance is before it’s recorded.
Start with the microphone. Different microphones have very different frequency profiles, so some pairings of voice and mic are simply sibilant. Choosing the right mic for the voice matters far more than anything you do to the recording later. With a directional mic, pointing it at the singer’s chest rather than straight at the mouth is a good compromise: the voice stays full, and the sharpest consonants don’t land on the capsule head-on.
Use a pop filter. Its main job is stopping the blasts of air on “p” and “b”, but it also keeps the singer at a steady distance from the mic, which keeps the esses more consistent from one line to the next.
Then check what you did to the top end. Raising the high frequencies is usually the fastest way to make a vocal feel present and alive, and to give it that professional sheen. Those are the same frequencies where sibilance lives, so harsher esses come with that particular fix. If they appeared after an EQ boost, try a gentler boost before adding another plugin to undo it.
A few bad esses? Fix them by hand. If sibilance only spikes now and then, turning those moments down with clip gain is quicker and more transparent than a plugin listening to the whole track.
A de-esser earns its place when sibilance is a pattern across the whole recording, to the point where it keeps distracting you, and it can’t be fixed at the recording stage anymore.
Where the de-esser sits in a vocal chain, before or after compression, is covered in our guide to building a vocal chain.
De-esser vs dynamic EQ: what’s the difference?
A dynamic EQ is an equalizer whose bands only cut or boost when the signal in that band crosses a threshold. Set one narrow band in the sibilance range to cut only when it gets loud, and you have built a split-band de-esser.
The difference is mostly convenience. A dedicated de-esser is set up for one job, with a detection mode and controls named for it. A dynamic EQ gives you more control over the shape of the cut and can fix other problems in the same plugin, at the cost of more setup.
Does your DAW already have a de-esser?
Before buying one, check what’s already in your DAW. Two of the big ones document a de-esser of their own:
| DAW | Built-in option | Where to find it |
|---|---|---|
| Logic Pro | DeEsser 2 | Audio Effect slot > Dynamics > DeEsser 2 |
| FL Studio | Maximus, which has a de-esser mode | Producer Edition and up |
Checked 28 October 2026.
If yours isn’t on the list, its stock compressor can usually do the job: set its sidechain or detection filter to the sibilant band, so it only reacts to the esses.
Can you use a de-esser on acoustic guitar or the master?
We wouldn’t. A de-esser is built around the way a voice behaves, and other sources are better served by a multiband compressor.
A multiband compressor splits the signal into several frequency bands and compresses each one on its own. That lets you tame string squeak on an acoustic guitar, or a harsh band across a full mix, with settings chosen for that source rather than for a singer’s esses.
Key takeaways
- A de-esser is a compressor specialized for vocals: it only reacts to the sibilant band, so it fixes harsh consonants without dulling the whole voice.
- Start around 5 to 8 kHz for female voices and 3 to 6 kHz for male voices, then find the real band by listening. If you can hear it working, back off.
- Fix sibilance where it starts first: the right microphone, aimed at the chest, a pop filter, and a gentler top-end boost. Reach for a de-esser when it’s a pattern across the whole take.
- A few stray esses are faster to fix with clip gain than with a plugin.
- For acoustic guitar or a full mix, use a multiband compressor instead.
Our recommendation
Get it right at the recording first: the right mic for the voice, aimed at the chest, and go easy on the top-end boost. If the esses are still a problem across the whole take, use the de-esser your DAW already has, set it by ear, and stop as soon as the esses stop drawing attention to themselves.