Remove Harsh S Sounds (Sibilance)
Harsh, piercing "S" and "T" sounds — sibilance — come from close-miking and bright EQ. This de-esser detects the sibilant band and tames only the sharp peaks, smoothing the voice.
How it works
We isolate the 5–9 kHz sibilance band, track its level over time, and apply gain reduction only when a sharp "ess" spikes. Because it is dynamic, the rest of the voice keeps its brightness.
What it's good for
- Close-miked voiceovers
- Bright vocal recordings
- Podcast hosts with sharp Ss
- AI/TTS voice cleanup
Details
- Engine
- DSP
- Formats
- MP3, WAV, M4A, FLAC, OGG, AAC
- Price
- Free to try
Frequently asked questions
Cutting treble dulls the whole voice even during non-sibilant words. A de-esser only acts on the brief sibilant peaks, so clarity stays.
At the default amount it tames harshness without lisping. If you push it harder on very sibilant audio, ease the amount back down.
Yes — de-essing is standard on sung vocals as well as speech; the same 5–9 kHz targeting applies.
Yes. Sibilance lives in the 5 to 9 kHz band regardless of language, so it tames sharp S and T sounds in speech of any tongue.
On a full music mix it can dip high-frequency percussion when they spike, so it works best applied to an isolated vocal rather than the whole mix.
Yes. There is an amount control; start at the default and increase only if the sibilance is still harsh, to avoid a lisping sound.
No. It is a dynamic DSP filter that processes quickly on ordinary hardware with no GPU required.