How to Remove Harsh Sibilance and Vocal Piercing Frequencies
Sibilance is the burst of high-frequency energy in S, T, Z and SH sounds. It is a normal part of speech that becomes a problem through recording rather than through speaking: a bright condenser microphone, a position too close to the mouth, and a little compression can turn ordinary consonants into something that stabs at the listener on headphones. Once it is in a recording it does not go away on its own, and it is one of the most fatiguing faults in spoken audio. The energy sits mostly between about 5.5 and 8 kHz for adult voices, and this filter is centred at 6.8 kHz with a Q of 3, which spreads its effect across roughly that range. The intensity control sets how deep the cut goes, up to 12 dB at the maximum. Now the part that other tools tend to leave out. A studio de-esser is dynamic: it listens for sibilance and reduces the band only while sibilance is present, leaving the rest of the performance at full brightness. This one is a fixed filter. It cuts that band all the time, including through vowels, cymbals and every other sound that lives up there. The practical consequence is that a heavy setting removes the harshness and also takes some life out of the whole recording, making it sound duller and slightly further away. That makes the technique straightforward to use well: apply the smallest amount that solves the problem, and judge it on the whole passage rather than on the S itself. Two or three dB frequently does the job. Going to the maximum because the S is still slightly audible usually trades one flaw for a worse one. If the sibilance is severe enough that a gentle cut cannot fix it, the real answer is at the microphone, not in the file: move it off-axis, or a few centimetres further away.
How to Use the Audio369 Online Audio De-Esser (Step-by-Step)
This is a tool for voices. On full music mixes the same cut removes cymbals and air along with the sibilance.
Judge the setting on the worst passage you have, since that is what determines how much cutting the recording actually needs.
The control runs from 0.1 to 1.0, which corresponds to a cut of roughly 1 to 12 dB. Stop as soon as the harshness stops bothering you.
Because the cut is always active, listen for the recording sounding dull or distant. If it does, you have gone too far and should back off.
Technical Architecture & Audio Engine Specifications
A fixed cut against a dynamic de-esser
| Behaviour | This tool | A studio dynamic de-esser |
|---|---|---|
| When it acts | ✓ Always, throughout the file | Only while sibilance is detected |
| Effect on vowels | ✓ Reduced along with everything else | Left alone |
| Effect on cymbals | ✓ Reduced | Untouched unless they trigger it |
| Setting up | ✓ One control | Threshold, range, frequency, timing |
| Right amount | ✓ The least that works | Can be pushed harder without dulling |
Who Uses Audio369 Audio De-Esser? (Practical Creative Workflows)
Podcast episodes recorded too close
Speaking a few centimetres from a condenser microphone exaggerates sibilance badly, and a couple of dB here makes an episode far easier to listen to.
Voiceovers destined for headphones
Harsh S sounds that pass unnoticed on speakers are unpleasant on in-ear headphones, which is how most narration is actually heard.
Interviews with a bright microphone
Some microphones have a presence lift right in the sibilance range. A modest cut compensates without touching anything else in the chain.
Recordings that were compressed hard
Compression raises quiet detail, including sibilance, so heavily compressed speech often needs a little of this afterwards.
Cleaning up before publishing
A last pass over a finished read catches the harshness that was inaudible while you were concentrating on the words.
Frequently Asked Questions (FAQ)
Why does my whole recording sound duller now? +
Because the cut is always active rather than only during the S sounds, so vowels and air are reduced too. Lower the intensity: the smallest setting that removes the harshness is the correct one.
Is this a real de-esser? +
It is a targeted cut in the sibilance band, not a dynamic processor. It genuinely reduces sibilance, and it does so by reducing that band all the time, which is a meaningful difference from a studio de-esser and the reason to use it lightly.
Can I use it on a full song? +
You can, but the same band holds cymbals and the top end of most instruments, so a mix will lose air along with the sibilance. It is far better suited to isolated voice.
What if the sibilance is still there at full intensity? +
Then the problem is beyond what a fixed filter can solve, and pushing further will only dull the recording. Re-record with the microphone slightly off-axis or further back, which fixes sibilance at its source.
Should I de-ess before or after compressing? +
After. Compression raises the level of quiet detail including sibilance, so de-essing first leaves the compressor free to bring it back up again.
Why does vocal sibilance concentrate around 5 kHz to 8 kHz? +
When pronouncing 's', 'sh', and 't' sounds, air rushes past the teeth and palate, creating high-frequency acoustic turbulence that microphone capsules often exaggerate into harsh, piercing peaks.
How does dynamic attenuation differ from static EQ cutting? +
Static EQ cuts high frequencies permanently, making vocals sound lisping and dark. A de-esser acts dynamically, pulling back volume only during the brief milliseconds when harsh sibilant spikes occur.
Can this de-esser smooth out harsh podcast voice tracks recorded on condenser mics? +
Yes. Inexpensive USB condenser microphones often have aggressive treble boosts that make speech fatiguing. Adjusting the threshold slightly softens sharp spikes for broadcast-ready listening comfort.