To de-ess in Ableton: add Multiband Dynamics, turn off the Low and Mid bands, set the High band crossover to about 5.5 kHz, set the High band's above-threshold ratio to around 4:1, and lower its threshold until S sounds trigger 3–6 dB of gain reduction. Place it after your compressor.
- Sibilance (harsh S, T, CH sounds) lives roughly between 5–9 kHz — lower for deep voices, higher for bright ones.
- Multiband Dynamics with only the high band active is a fully functional de-esser.
- De-ess after compression, because compression makes sibilance louder.
- Aim for 3–6 dB of reduction on the harshest esses; more than that lisps the vocal.
- Static EQ cuts can't fix sibilance without dulling the whole vocal — you need level-dependent processing.
What sibilance is and why EQ can't fix it
Sibilance is the burst of high-frequency energy in S, T, Z, and CH consonants, concentrated between roughly 5 and 9 kHz. It's natural speech — the problem is that microphones, compression, and brightness boosts exaggerate it until every 's' stabs the listener.
A static EQ cut at 7 kHz reduces esses, but it also dulls the vocal for the 95% of the time no ess is happening. What you need is a cut that appears only during sibilant moments — in other words, a compressor that listens to just the sibilant band. That's precisely what a de-esser is, and Live's Multiband Dynamics can be configured as one in under a minute.
The Multiband Dynamics de-esser recipe
- Drop Multiband Dynamics on the vocal, after your compressor.
- Turn off the Low and Mid bands (click their activator buttons) so only the High band processes audio.
- Set the Mid/High crossover frequency to about 5.5 kHz. This is the split point: everything above it can be compressed, everything below passes untouched.
- In the High band's Above section, set the ratio to around 4:1 by dragging the ratio value.
- Play the harshest line of the vocal and lower the High band's Above threshold until esses trigger 3–6 dB of gain reduction.
- Set Attack fast (≤1 ms is fine here — there's no transient to protect in this band) and Release around 20–60 ms so the cut vanishes the instant the ess ends.
Dialing it in for different voices
- Deep male vocals: sibilance often sits lower — try a 4.5–5.5 kHz crossover.
- Bright female or airy pop vocals: energy centers higher — try 6–8 kHz so you don't dull the body of the voice.
- Whispery or breathy styles: use a gentler ratio (2:1) and less reduction; breath is part of the sound.
- Rap vocals with heavy saturation: consider a second, lighter de-essing stage after the saturator, since distortion regenerates high harmonics.
The audible test is simple: lyrics should sound identical except the esses no longer jump out. If the singer suddenly has a lisp, you're cutting too deep — raise the threshold or lower the ratio.
Where the de-esser goes in the chain
After the main compressor, before saturation and brightness boosts. Compression raises sibilance, so de-essing earlier means the compressor partially undoes your work; saturation and air-shelf boosts both amplify whatever sibilance survives, so they should receive an already-tamed signal. The full sequence is covered in our vocal chain order guide.
PhantomRack's generated racks include a Multiband Dynamics de-essing stage automatically whenever the analysis hears strong sibilance in your dry vocal or the reference — with the crossover placed from where that sibilance actually sits, rather than a fixed default.
Frequently asked questions
Does Ableton Live have a built-in de-esser?
Not as a named device, but Multiband Dynamics configured with only its high band active (crossover ~5.5 kHz, 4:1 ratio, 3–6 dB reduction) works exactly like a dedicated de-esser and ships with Live Standard and Suite.
What frequency should I de-ess at?
Start the crossover at 5.5 kHz and adjust by voice: 4.5–5.5 kHz for deeper voices, 6–8 kHz for brighter ones. Solo the band and sweep until you isolate the 'sss' energy with minimal vowel tone.
How much de-essing is too much?
If the vocal starts to sound like the singer has a lisp, or words like 'so' lose their consonant entirely, back off. 3–6 dB of reduction on the harshest esses is the normal range.
Should the de-esser go before or after the compressor?
After. Compression raises sibilance along with everything else, so de-essing post-compression catches esses at their final level. The exception: sibilance so severe it distorts the compressor — then tame it first.
Hear it onyour voice.
PhantomRack turns everything in this guide into a finished rack: upload a reference, download a stock-Ableton .adg tuned to it. First rack is free — no card required.