Vocal Sibilance Finder

Knowledge Base

How This Tool Finds Your Sibilance Frequency

Rather than just showing the loudest frequency in your recording (which is usually a vowel, not an "s"), this tool identifies the specific time frames where high-frequency energy dominates the signal: the acoustic signature of unvoiced fricatives like "s" "sh" and "t." It then averages the spectrum across only those frames, revealing the frequency this particular voice's sibilance actually peaks at.

Why This Beats a Generic "5-8 kHz" Guess

Every voice sibilates at a slightly different frequency depending on mouth shape, dentition, and mic placement. Some peak closer to 4 kHz, others past 9 kHz. Feeding a de-esser the generic range instead of the measured peak either misses the problem or cuts frequencies that don't need it.

Using the Result

Set your de-esser's target frequency to the detected peak, and use the suggested bandwidth as a starting point for its Q or frequency-range control, see the De-Essing Explained guide for the rest of the setup.

FAQ

Q: What if the tool doesn't find enough sibilant frames?
A: This usually means the take doesn't have many drawn-out "s"/"sh" sounds, or the recording has significant background noise masking them. Try a longer or more speech-dense sample.

Q: Does mic type affect the detected frequency?
A: Yes, brighter condenser mics tend to emphasise sibilance higher in the spectrum than darker dynamic mics on the same voice, which is exactly why measuring your specific recording matters more than a generic number.

Finding the frequency and treating it

Sibilance is the burst of high frequency energy produced by s, sh, z and ch sounds. Every voice concentrates it in a narrow band, and that band differs from person to person. Most adult voices land somewhere between 5,000 and 9,000 hertz, with higher voices generally sitting further up. Knowing the specific figure for the specific voice is what makes treatment precise rather than approximate.

Once you have the frequency, a de-esser set to that band reduces the harshness without dulling the rest of the recording. The alternative most people reach for first, a broad high frequency cut, removes the sibilance and the clarity together and leaves the voice sounding muffled. A narrow reduction at the right frequency is nearly inaudible on everything except the sounds you wanted to tame.

Sibilance is often a microphone placement problem before it is a processing problem. Speaking directly into the capsule aims the airflow from an s straight at it. Angling the microphone slightly off to one side, so the speaker talks past it rather than into it, reduces the effect at the source and costs nothing. A pop filter helps with plosives and does relatively little for sibilance, since the two happen at different frequencies.

Treat lightly. A voice with no sibilance at all sounds lisped and unnatural, and listeners notice the absence as readily as the excess. The goal is removing the sharpness that makes someone reach for the volume control, not eliminating the sound.

Place the de-esser before compression rather than after. A compressor reacting to an untreated sibilant pulls the whole voice down for that instant, producing an audible dip. Treating sibilance first means the compressor sees a more even signal and works on the performance instead of the consonants.

Vocabulary

Sibilance. The concentrated burst of high frequency energy in s, sh, z and ch sounds. Natural in speech, but exaggerated by close microphones and by compression, at which point it becomes tiring to listen to.

De-esser. A processor that reduces level only in a chosen frequency band, and only when energy there exceeds a threshold. Set to a voice's sibilance frequency, it acts on the harsh sounds and leaves everything else untouched.

Hertz. The unit of frequency, meaning cycles per second. Higher numbers are higher pitched. Human hearing runs to roughly 20,000 Hz, and sibilance sits high in that range at 5,000 to 9,000 Hz.

Plosive. The thump produced when the burst of air from a p or b hits the microphone. A different problem from sibilance, occurring at low frequencies rather than high, and solved by a pop filter rather than a de-esser.