Studio Mix · concept guide

De-essing: Taming Sibilance Without Lisping the Vocal

Where sibilance lives, why a static EQ cannot fix it, the difference between a de-esser, a dynamic EQ and a multiband compressor, and a drill that has you set the threshold by ear.

Play it first · De-esser

Free drill, no account

De-esser

Tame the S without lisping the singer.

6 rounds, about two minutes. Headphones help. Nothing is saved until you choose to.

Sibilance is the one vocal problem that gets worse the better the rest of the mix gets. Add air to the vocal, and every S and T comes with it. Compress the vocal, and the quiet consonants come up to meet the loud ones. The result is a vocal that is bright and present and impossible to listen to at volume, and the wrong fix, a static cut in the sibilance range, dulls every vowel to solve a problem that only exists for a few milliseconds at a time.

A de-esser is a compressor that only listens to one band, and setting it well is an ear job: too little and the esses remain, too much and the singer lisps. The drill above has you find the threshold that does the first without the second. The cards cover de-essing as a spectral process, dynamic EQ against static EQ and against multiband compression, resonance suppression and when not to use it, and the two tracking decisions, mic distance and proximity, that decide how much sibilance you have to fix in the first place.

The concepts

8 lessons from the Studio Mix deck

  1. 1

    De-essing as Spectral Processing

    A de-esser is, underneath, a frequency-specific compressor. It targets sibilance, the sharp "s," "sh," and "t" sounds concentrated around 4-9 kHz, using a sidechain tuned to that range, so it applies gain reduction only when those frequencies spike. It comes in two modes: wideband ducks the whole signal when sibilance hits (more natural), split-band reduces only the high band (more transparent, targets sibilance alone).

    In practice: A vocal with harsh "s" sounds: a de-esser sidechained to 4-9 kHz clamps down only on the sibilant moments, split-band mode if you want to touch nothing but the sibilance.

    A de-esser is a compressor sidechained to 4-9 kHz sibilance, wideband ducks all, split-band the highs.

  2. 2

    Dynamic EQ vs Static EQ

    A static EQ band applies its boost or cut constantly, no matter what the signal is doing. A dynamic EQ band only engages once the signal at that frequency crosses a threshold, it is, in effect, a frequency-specific compressor (or expander). That distinction matters when a tonal problem is intermittent. A dynamic cut leaves the natural tone untouched whenever the problem is not occurring.

    In practice: An acoustic guitar boomy only on hard strums: a dynamic cut at 200 Hz engages on the loud strums and stays inactive during fingerpicking, a static cut would thin the quiet passages too.

    Static EQ acts constantly; dynamic EQ engages only above a threshold, a frequency-specific compressor.

  3. 3

    Dynamic EQ vs Multiband Compression

    Dynamic EQ and multiband compression both apply level-dependent gain by frequency, but they are built differently. A multiband compressor splits the signal into fixed bands with crossover filters and compresses each. A dynamic EQ applies a normal parametric bell or shelf at a specific frequency without splitting the signal. That makes dynamic EQ more surgical and more transparent, it avoids the crossover artifacts and the wide band boundaries of multiband.

    In practice: Taming fiddle bow noise at 3 kHz: a multiband band catches everything in its 2-5 kHz crossover range, but a narrow dynamic EQ bell hits just 3 kHz and leaves note fundamentals alone.

    Multiband splits the signal at crossovers; dynamic EQ is a surgical, artifact-free parametric band.

  4. 4

    Soothe2: Resonance Suppression

    Soothe2 is a resonance suppressor, not an EQ. A static EQ notch is fixed at one frequency forever; Soothe2 continuously analyzes the signal and attenuates resonant peaks on a per-frequency basis, in real time. The difference matters when resonances move, it suppresses only frequencies that are actually ringing or building up at that moment, leaving everything else untouched.

    In practice: A fiddle whose harsh resonances shift as the player moves across strings: Soothe2 tracks and tames each ringing peak per note, where static notches fix one frequency and dull the others.

    Soothe2 dynamically suppresses resonant peaks per-frequency in real time, static EQ can't track them.

  5. 5

    When NOT to Use AI/Spectral Processing

    Spectral and AI tools are powerful, and that is the danger. Over-applied, they produce a lifeless, synthetic quality, every acoustic instrument's natural resonances are part of its character, and stripping too many removes its identity. The rule: use spectral tools to fix genuine problems, not to "improve" a sound that was recorded well. One tool per problem, conservative settings, constant level-matched A/B.

    In practice: A bluegrass album with Soothe2, Gullfoss, and smart:EQ on every track sounding "dead": strip it all back, re-add only where a real problem exists, one tool per problem.

    Over-processing with spectral tools strips an instrument's character: fix real problems, don't "improve."

  6. 6

    Vocal Mic Distance: 4" Intimate vs 8" Natural vs 12" Room Blend

    Vocal mic distance sets the character. At 4 inches, the vocal is intimate and breathy with strong proximity effect, every breath and consonant audible, ideal for confessional performances. At 8 inches, the tone is natural and balanced, like a voice heard from nearby, the usual starting point. At 12 inches, room ambiance enters and proximity effect fades for an open, airy sound. A pop filter is essential up close to control plosives.

    In practice: A folk ballad that should feel like "whispering in your ear": move the mic to 3-5 inches for proximity warmth and breath detail, with a pop filter for plosives.

    Vocal distance: 4" intimate/breathy, 8" natural, 12" open and airy with room.

  7. 7

    Proximity Effect

    Proximity effect is a bass boost that appears when a directional mic (cardioid, figure-8) is moved close to a source, the closer it gets, the bigger the low-end lift. Omnidirectional mics do not have it. It is a tool: move a directional mic in to add natural warmth, move it back to thin out a boomy source. The same effect causes problems if you do not account for it.

    In practice: A thin, reedy voice: move the cardioid in to 4-6 inches and proximity effect adds free low-mid warmth, no EQ needed.

    Proximity effect = directional mics gain bass up close, use it or back off it.

  8. 8

    FabFilter Pro-Q Dynamic Bands

    A dynamic EQ band needs four controls beyond frequency and Q. Threshold sets the level at which the band activates. Range sets the maximum dB of cut or boost it will apply. Attack sets how fast it responds, and release how fast it returns to neutral. In FabFilter Pro-Q, every band can switch independently between static and dynamic, so one EQ instance handles both fixed tone-shaping and level-dependent problem-solving.

    In practice: Setting up a dynamic band: pick the frequency and Q, set the threshold just above normal level, set range to ~5 dB of cut, and dial a fast attack with a medium release.

    A dynamic EQ band is threshold + range + attack + release on a parametric band.

These lessons are the teach side of knowledge cards from the Studio Mix deck. In the daily plan each one is followed by a quiz, spaced over weeks, so it stays known rather than read once.

The plan

Reading this once is not the same as hearing it.

A daily session schedules the drill above at your level, quizzes these cards until they stick, and moves you on when the numbers say you are ready. Twenty minutes a day, in the browser, no plugins.

Related concept guides