Fix it
How to fix harsh s sounds on your mic
Every s and sh comes through the mic like a little dart. That is sibilance, and it has three fixes in a strict order: move the mic, add a de-esser, and only then reach for EQ.
What that sharpness actually is
Say the word "stress" out loud and listen to its edges. The s at the front and the ss at the end are sibilance: short bursts of high-frequency hiss your mouth makes on s, sh, ch, and z sounds. The tone of your voice lives mostly below 4 kHz, but those bursts concentrate their energy roughly between 2 and 12 kHz, with most voices hissing loudest somewhere near 6 kHz, right where human hearing is most sensitive. Microphones make it worse: close miking, the presence boost built into most vocal mics, and bright condenser capsules all push that band up, so a voice that sounds fine in the room arrives with every s sharpened.
Here is the part that saves you an hour of wrong fixes: sibilance is not a volume problem. Look at one word up close and the s bursts stand just as tall as the vowel between them, so pulling the fader down makes the whole voice quieter while the s stays exactly as harsh relative to everything else.
Because the problem is concentrated in one narrow band for a few milliseconds at a time, the fixes go in a strict order: first the physical ones that cost nothing and remove the hiss before it is ever recorded, then the tool built precisely for this job, and only then a careful EQ move.
Move the mic before you touch software
Sibilance is directional. The hiss of an s shoots straight forward out of your mouth in a narrow beam, while the body of your voice radiates much more evenly. That physics gives you the cheapest fix there is:
- Angle the mic off-axis. Instead of pointing the capsule straight at your lips, aim it at your mouth from 15 to 30 degrees off to the side, or point it at your chin from slightly above. The beam of hiss shoots past the capsule instead of into it, the body of your voice barely changes, and the darts land noticeably softer. This is to sibilance what a pop filter is to plosives.
- Back off a little. A fist or two of extra distance softens the direct blast of every consonant and takes the edge off a bright capsule. If backing off makes you too quiet, raise the gain, not the closeness.
- Undo any "clarity" boost. If you added a high-frequency EQ boost to sound crisper, you sharpened every s along with it. Switch it off while you diagnose; you may find it was most of the problem.
A pop filter, for the record, will not help. It exists to break up the blast of air from p and b sounds; the hiss of an s sails straight through the fabric. If the s is still sharp after the mic is angled and at a sane distance, what is left is a job for processing.
Add a de-esser and lower one slider
A de-esser is a compressor with selective hearing: it watches only the sibilant band and turns the signal down for the few milliseconds an s actually lasts, then gets out of the way. That is what makes it better than any static fix. It acts exactly where the problem is, exactly when the problem happens, and does nothing at all the rest of the time.
In patchd the de-esser is a free node on every channel: open the effects rack, add it from the + Add Node menu, and place it after the compressor. That order matters, because a compressor raises your average level and its makeup gain lifts sibilance right along with everything else, so a voice that barely needed de-essing before the compressor clearly needs it after.
The node ships at defaults tuned for exactly this complaint: Threshold -28 dB, Range 8 dB, Frequency 6500 Hz, Bandwidth 1.5x. For most voices the entire setup is one move: speak a few s-heavy sentences and lower the threshold until the harshness softens. The Range control caps the reduction at 8 dB, so the node cannot over-clamp into a lisp unless you push it there. If your voice stops sounding like you, back the threshold off; a de-esser you can hear working is set too hard. The full walkthrough, from the defaults down to the split-band mode and the timing math, lives at how to tame harsh s sounds with a de-esser.
When EQ helps, and when it backfires
There is one case where EQ is the right tool: the whole top end is hot, not just the s sounds. If cymbals, hiss, mouth noise, and consonants all feel brittle and fatiguing, the mic is simply bright, and a gentle shelf fixes the balance. A fresh EQ in patchd already ships with a high shelf at 10 kHz; pulling it down a dB or two darkens the entire top end evenly, s sounds included.
What EQ cannot do is fix sibilance on its own, and this is where most people go wrong. An EQ cut is static: carve out 6 kHz deep enough to tame the s and that presence is gone all the time, including the 95 percent of every sentence when no s is happening, which reads as a dull, distant voice. The de-esser exists precisely because the problem is dynamic. So split the work: a small shelf if everything is bright, a de-esser for the bursts, and never a crater at 6 kHz.
The short version
- Angle the mic off-axis and back off a little; the hiss beams straight ahead, so let it miss the capsule.
- Kill any high-end boost you added for clarity while you diagnose.
- Add the free de-esser after your compressor and lower the threshold until the s softens. The defaults handle the rest.
- Reach for a gentle 10 kHz shelf only if the whole top end is bright, never as a substitute for the de-esser.
patchd gives you the whole toolkit on every channel: a free de-esser with honest defaults, a real EQ for the shelf, and a rack that runs live while you talk, so you hear the s soften as you move the slider. It is in development now. Join the waitlist and we will tell you the moment it is ready to install.