The De-esser

Watching only the sibilant band, so a sharp s softens without flattening the whole word.

Related

EqualizerCompressorPopular singing and the microphone

The question

Sibilance sits between 5 and 10 kHz and lasts a few tens of milliseconds. Full-band compression flattens vowels and consonants together. A de-esser watches only that band.

How it works

A de-esser detects sibilant energy with a band-pass or high-pass filter, then attenuates an adjustable band once the threshold is crossed. Wideband mode turns down the whole signal, while narrowband mode turns down only the detected band, preserving more of the word. A split design compresses only the high path. The detection frequency follows the singer, often 4 to 6 kHz for men and 6 to 9 kHz for women.

Settings and what you hear

Sweep the frequency to the sharpest sibilant, then set the threshold so normal singing does not trigger it and only an s does. Start with 3 to 6 dB of range, a fast attack and a release of 20 to 80 ms. Too much reduction turns s into th and gives a lisping sound.

What to listen for

On a soprano with piano, bypass the de-esser and remember where the s bites, then engage it and deepen gradually. The s should soften while the word stays clear. On a choir, take care not to dull the whole top end.

Common misconceptions

That a de-esser is only for vocals; that a lower threshold is always safer; that it can remove every harsh sound.

Sources

• Udo Zölzer, DAFX — Digital Audio Effects, equalizers and dynamics sections — background checked, not quoted
• Huber and Runstein, Modern Recording Techniques, vocal processing chapter — typical ranges checked, not quoted

Used for fact-checking only; all prose is original.