How to de-ess a vocal with EQ and dynamic EQ

Sibilance (S and SH consonants) usually peaks between 5-9 kHz, often around 6-8 kHz on female vocals and 4-6 kHz on male.

Updated 2026-05-19

Short answer

Sibilance (S and SH consonants) usually peaks between 5-9 kHz, often around 6-8 kHz on female vocals and 4-6 kHz on male. A static EQ cut helps for consistent harshness; a dynamic EQ or de-esser only cuts when the sibilance actually triggers. Cut 2-4 dB at the offending frequency with a Q of 4-6.

Find the offending frequency: solo a narrow boost (+12 dB, Q 6) and sweep 5-9 kHz until S sounds sharpest
Female vocals: usually 6-8 kHz; male vocals: 4-6 kHz
Static EQ: cut 2-3 dB with Q 4 — works for constant problems
Dynamic EQ / de-esser: cut 3-6 dB only when level exceeds threshold
Don't over-cut: vocals need air above 8 kHz, killing it makes them sound muffled

Try this in Cue

Find the sibilant frequency by sweeping

Insert an EQ on the vocal. Set a narrow band (Q 6) with +12 dB boost, then sweep slowly across 4-10 kHz while the vocal plays. The harshest, most painful frequency is the sibilant peak — usually somewhere between 5-9 kHz. Remember the exact Hz value, then pull the boost back to a cut at that frequency.

Try a static EQ cut first

Cut 2-3 dB at the sibilant frequency with Q 4. Listen — if the S sounds tame without the vocal sounding dull, you're done. Static EQ is the simplest fix and works for vocals where the sibilance is fairly consistent across the take. If only certain words sound harsh, switch to dynamic.

Use a dynamic EQ or de-esser for intermittent sibilance

If only some S sounds are harsh, use a dynamic EQ band or a de-esser plugin (FabFilter Pro-DS, Waves DeEsser, Logic's stock DeEsser). Set the threshold so the cut only triggers on the loudest sibilants. Cut 3-6 dB. The vocal stays bright except during sibilant moments.

Don't kill the air

Vocals need presence and air above 8 kHz to feel modern and intimate. Cutting too much in the sibilant range (more than 4-5 dB) makes the vocal sound dull and lispy. If you need that much cut, the problem is recording or compression — go fix it at the source instead.

Frequently asked

What frequency is sibilance?

Usually between 5-9 kHz, depending on the voice. Female and higher-pitched voices peak around 6-8 kHz; lower male voices peak around 4-6 kHz. Sweep with a narrow boost to find the exact frequency for the specific vocal — every voice is slightly different.

De-esser vs dynamic EQ?

Functionally similar — both cut a specific frequency only when it triggers a threshold. De-esser plugins are optimized for sibilance (preset frequencies, easier UI). Dynamic EQ is more flexible if you also need to control other frequency-dependent problems (mud that swells on chorus, harshness only at high notes).

Should I de-ess before or after compression?

After compression usually — compression brings up sibilance by raising the average level. De-essing after catches what compression amplified. The exception is heavy compression on already-sibilant takes: a light de-ess before compression keeps the compressor from over-reacting to S sounds.

Why does my vocal sound lispy after de-essing?

Over-cutting. If you cut more than 5 dB or use too wide a Q, you remove the consonant clarity along with the harshness — vocals start sounding like the singer has a lisp. Reduce the cut depth, narrow the Q, or use a dynamic EQ so the cut only triggers when needed.

Apply this in Cue

Open the app with this question pre-loaded. Free to use, no signup.

Try this in Cue