I used to think lip smacks and wet mouth noises were just part of streaming until viewers started leaving polite — and then not-so-polite — comments. After a lot of testing, I’ve built a mic chain and a set of habits that stop lip smacks dead in their tracks and keep my voice sounding clean on Twitch, all without buying new gear. Below I lay out the exact software settings, physical tweaks, and workflow I use. You can copy them and tune to taste.
Why lip smacks happen (so you know what you’re fighting)
Before we jump into settings, it helps to understand the problem. Lip smacks are short, high-energy transients with spectral content mainly between ~2 kHz and 8 kHz. They’re loud and brief, so compressors alone don’t always catch them. De-essers can help but are tuned for sibilance around 5–8 kHz and may not get everything. The best results come from a combined approach: control the source, then use gating/expansion, transient-sensitive processing, targeted EQ, and a dynamic de-esser.
Start at the source — physical and behavioral tweaks
Fixing your mic chain is easier when you reduce the amount of smacking to begin with.
The practical mic chain (software only) I run in OBS/Voicemeeter/ReaPlugs
All of the following can be implemented with free tools (ReaPlugs: ReaEQ, ReaGate, ReaComp, ReaXcomp; or use OBS filters and stock plugins). I’ll give concrete settings to start with — tweak by ear.
1) High-pass filter (clean the rumble)
Plugin: ReaEQ or OBS High-Pass
| Type | 1st-order HPF |
| Cutoff | 70–120 Hz (male voices lower; female a little higher) |
| Purpose | Removes low rumble and mic stand noise so the rest of the chain behaves predictably |
2) Noise gate / expansion (kill quiet smacks)
Plugin: ReaGate (expander mode) or OBS Noise Gate
| Mode | Downward expander (if available) or gate |
| Threshold | -40 to -30 dB relative to your speaking level — start conservative |
| Ratio (expander) | 1.5:1 to 4:1 |
| Attack | 3–10 ms (fast to catch transient smacks) |
| Release | 70–200 ms (slower to avoid choppiness) |
Why: Lip smacks are short high-energy bursts that often sit just at or below speech level. An expander reduces their level without making your voice sound gated. If a gate is too heavy-handed, expand instead.
3) Transient-sensitive compressor / multiband compressor
Plugin: ReaComp or ReaXcomp / OBS Compressor
| Mode | Multiband (if available) or a compressor with sidechain EQ |
| Band of interest | 2–8 kHz for smacks |
| Threshold | Low so it reacts to transients (-20 to -10 dB depending on input) |
| Ratio | 3:1 to 6:1 on the upper band |
| Attack | 0.5–5 ms (ultra-fast to catch peaks) |
| Release | 30–120 ms |
Why: A multiband compressor lets you squash only the frequency range where smacks live without dulling the entire voice. If you don’t have multiband, use a fast full-band compressor but follow with de-essing and EQ.
4) De-esser tuned like a smack guard
Plugin: ReaTune De-esser or FabFilter Pro-DS (if you own it)
| Frequency | 2.5–6 kHz depending on where smacks sit for your voice |
| Threshold | Set to catch sibilance and smacks without affecting speech — aim for 2–6 dB reduction on hits |
| Range | Narrow Q (1.5–3) for surgical control |
Why: De-essers are not just for "s" — with the right freq and narrow Q they can clamp down on saliva transients while leaving consonants intact.
5) Surgical subtraction EQ
Plugin: ReaEQ or any parametric EQ
6) Light compression for cohesion
Plugin: ReaComp / OBS Compressor
| Attack | 10–30 ms |
| Release | 60–150 ms |
| Ratio | 2:1 to 4:1 |
| Makeup | Gain to taste |
Why: Smooths level so viewers don’t get shocked by occasional louder words, but keep it moderate — heavy compression can exaggerate smacks.
7) Optional: dynamic EQ or sidechain to gate smacks
If you want to get surgical and you have dynamic EQ or RX-style spectral repair: create a dynamic band at the smack frequency and make it duck very quickly (fast attack, short release) when the energy spikes. This mimics a focused, frequency-aware gate and is very effective when combined with the above steps.
Practical OBS filter chain order I use
Monitoring and tuning tips
Household hacks that help (no purchase needed)
Finally, don’t expect perfection overnight. Tuning thresholds and EQ often takes a few sessions. I tweak my chain after every big stream session and keep a “safe” preset for casual streams and an “aggressive” preset for high-stakes recordings. Follow the order and params above, record samples, and adjust by increments — you’ll shave off most lip smacks without buying anything new.
