If you clip a lavalier mic to your shirt and it still sounds hollow, muffled, or noisy, the problem is almost never the mic itself — it is placement, gain staging, and processing order. This guide covers the settings that matter most for chest-mounted lavaliers, whether you are recording voiceover for gameplay videos, streaming, or capturing field audio.

Prioritize these in order: placement first, gain second, noise reduction third, and everything else last. A lavalier worn on the chest sits far from the mouth and points wherever the fabric points, so it picks up more room reflection and clothing noise than a boom or headset mic. No software setting can fully recover from bad placement, so get the physical side right before touching a single slider.

Setting Recommended value Why
Placement distance 6–8 inches (15–20 cm) below the chin, sternum height Close enough for voice dominance, far enough to avoid plosives and chest rumble
Clip orientation Capsule facing up, cable loop secured through the clip Upward-facing capsules reduce fabric rustle transmitted through the cable
Input gain Peaks at −12 dBFS to −6 dBFS, average around −18 dBFS Leaves headroom for laughter or loud exclamations without clipping
High-pass filter 80–100 Hz, 12 dB/octave Removes low-frequency rumble from movement and handling without thinning the voice
Noise gate Threshold set 6–10 dB above the noise floor, attack 5 ms, release 100–150 ms Mutes room hiss between sentences without chopping word tails
Noise suppression Moderate/single-pass; avoid stacking multiple reducers Stacked suppressors cause metallic, underwater-sounding artifacts
Compression Ratio 3:1, threshold −18 dBFS, attack 10 ms, release 100 ms, 3–5 dB gain reduction Smooths volume differences when you lean back or turn your head
Sample rate / bit depth 48 kHz / 24-bit Standard for video; 24-bit gives enough dynamic range to record low and raise volume later

Setting-by-Setting

Placement: 6–8 inches below the chin

Set the mic at sternum height, centered or slightly to the side of your chest. The exact trade-off: closer to the collar (about 4 inches) captures more direct voice and less room, but every head turn swings the volume noticeably, and breathing and plosives get louder. Lower on the chest sounds more consistent but thinner, because you lose high-frequency detail over distance. If you move your head a lot while talking — common in gaming commentary — favor the lower, more consistent position and compensate with compression.

In OBS Studio, confirm placement with the audio mixer: right-click your mic source, select Filters, and watch the level meter while speaking at normal volume. If peaks jump more than 10 dB between head positions, move the mic lower or add more compression.

Cable loop and clip orientation

Most lavalier kits (Rode smartLav+, BOYA BY-M1, Sennheiser ME 2-II style) include a clip with a small cable loop or strain-relief slot. Thread the cable through the loop and secure it to your shirt below the capsule, then run the remaining cable down inside or behind your clothing. This is the single cheapest fix for rustle: cable-borne friction noise travels up the wire and into the capsule unless the loop decouples it.

How to check: record 30 seconds while rubbing the cable near the mic, then play it back with headphones. If you hear loud scratching, the loop is missing or loose. To undo or adjust, simply unclip, re-thread the loop, and re-clip — no software involved.

Input gain: peaks at −12 to −6 dBFS

Set gain so your loudest normal speech peaks between −12 dBFS and −6 dBFS on the meter. The trade-off: higher gain raises your voice above the preamp’s noise floor but pushes laughter and excited reactions into clipping, which cannot be repaired. Recording too low wastes bit depth and hiss becomes obvious when you raise the volume in post.

Windows: right-click the speaker icon, select Sounds → Recording tab, double-click your interface, open the Levels tab, and start at 60–70% for typical USB interfaces. On dedicated recorders like the Zoom H1n or a Rode Wireless GO II, start at −12 dB input level and adjust from there.

High-pass filter: 80–100 Hz

A chest-mounted lav picks up low-frequency rumble from shirt friction, footsteps, and handling. Enable a high-pass (low-cut) filter at 80 Hz minimum; 100 Hz is safer for male voices if placement is stable, while 80 Hz preserves a bit more chest tone. The trade-off: filtering at 120 Hz or higher starts thinning bass presence in the voice, making you sound distant. Most interfaces and USB mixers expose this as a physical switch labeled “Low Cut” or “Bass Roll-off.” In software, use the 3-band EQ filter in OBS set to Low Shelf, 100 Hz, and reduce the low band, or the EQ module in Audition/Reaper’s ReaEQ with a high-pass at 90 Hz and 12 dB/octave slope.

Noise gate: threshold 6–10 dB above the noise floor

Measure your noise floor first: stay silent for 10 seconds and read the average level (OBS shows this in the mixer meter; Audacity shows it under Analyze → Plot Spectrum). Set the gate threshold 6–10 dB above that number. Attack 5 ms so word beginnings survive, release 100–150 ms so sentence tails fade naturally instead of cutting abruptly. The trade-off: a higher threshold silences more room noise but clips soft consonants at sentence ends; a lower threshold preserves speech but lets hiss through between phrases.

OBS path: Filters → Noise Gate. In Reaper, use ReaGate on the track. To undo, remove or bypass the filter in the same menu.

Noise suppression: one pass, moderate strength

Use a single suppression tool — OBS’s Noise Suppression filter (RNNoise is the built-in AI option in current versions), NVIDIA Broadcast, or a plugin like ReaFir in subtract mode — never two chained together. Stacked suppressors multiply artifacts, producing the metallic, robotic “underwater” timbre. The trade-off: stronger suppression removes more fan and room noise but degrades voice naturalness, especially on sibilants and breaths. Start with RNNoise at its default; only escalate if your recording space is genuinely loud.

Compression: 3:1 ratio, −18 dBFS threshold

Chest placement means your level changes as you shift posture. A 3:1 ratio with the threshold at −18 dBFS and 3–5 dB of gain reduction evens this out. Attack 10 ms keeps consonant punch; release 100 ms avoids pumping. The trade-off: more compression sounds consistently loud but raises the noise floor during quiet moments and can sound dense or fatiguing if overdone. After compression, add makeup gain so your average level sits around −16 LUFS for YouTube delivery.

Sample rate and bit depth: 48 kHz / 24-bit

Set this identically everywhere: in Windows (Sounds → Recording → Advanced tab → 24-bit, 48000 Hz), in your interface’s control panel, and in OBS under Settings → Audio → Sample Rate 48 kHz. Mismatches between these locations cause crackling and dropped samples. 24-bit lets you record conservatively low with headroom and normalize later with no audible penalty.

By Hardware Tier

Low tier: 3.5 mm lav into a phone, camera, or on-board mic jack

Built-in jacks have noisy preamps and automatic gain control (AGC) you cannot disable, so work around it. Placement becomes critical — go 6–8 inches below the chin and prioritize the cable-loop fix, because these jacks amplify cable noise. Disable AGC where possible: in your camera menu, look for Audio → Auto Level / AGC and turn it Off, setting manual input level instead. On phones, use an app that allows manual gain (Open Camera on Android, Blackmagic Camera on iOS) instead of the stock recorder. Expect hiss; plan to apply one moderate pass of noise suppression in post rather than trying to record clean.

Mid tier: dedicated lav into a USB audio interface

A lav plugged into an interface with a real gain knob (Focusrite Scarlett Solo, MOTU M2, Audient iD4) or a wireless system’s receiver into that interface is where settings start paying off. Set the interface gain knob so peaks hit −12 dBFS, engage the interface’s low-cut switch if it has one, and disable all Windows-level “audio enhancements”: Sounds → Recording → Properties → Enhancements tab → check Disable all enhancements. Add your processing chain in OBS filters in this order: Noise Suppression → Noise Gate → High-pass/EQ → Compressor → Gain. This order prevents the gate from fighting the compressor’s makeup gain.

High tier: wireless lav system or XLR lav with quality preamp

With a clean signal chain (Sennheiser MKE 2 into a quality preamp, or a Rode Wireless Pro / DJI Mic 2 system), you can relax the processing. Set the transmitter’s input gain so peaks sit at −12 dBFS using the on-device metering, enable the transmitter’s built-in low cut at 80–100 Hz, and record the 32-bit float safety track if your system offers it — float recording makes gain settings nearly forgiving of mistakes. In software, reduce noise suppression to the weakest setting or disable it entirely; a quiet room plus a good preamp needs less correction, and skipping suppression preserves the most natural tone. Keep the compressor gentle: 2:1 ratio, 2–3 dB of gain reduction.

Common Mistakes

Fabric rustle (what causes it: the capsule rubbing against synthetic fabric as you move). How to check: record 20 seconds while sitting still, then 20 seconds shifting in your chair; if the second clip has loud scratching, this is your problem. What to do: mount the mic on a cotton layer, use a foam windscreen or a small piece of moleskin under the clip, and secure the cable loop. How to undo: remove the moleskin or windscreen if it dulls the high end.

Boomy, muddy voice (what causes it: proximity bass buildup plus chest-cavity reflection with no high-pass filter). How to check: play your recording and listen for a thick, “in a barrel” quality on vowels. What to do: enable the 80–100 Hz high-pass and, if still muddy, cut 3–4 dB with a wide Q around 200–300 Hz using an EQ filter. How to undo: bypass the EQ filter to compare before committing.

Metallic, robotic voice (what causes it: two noise suppressors running simultaneously, or suppression set to maximum). How to check: look at your filter chain in OBS — if Noise Suppression appears twice, or NVIDIA Broadcast runs while OBS’s suppressor is active, that is the cause. What to do: keep exactly one suppressor at moderate strength. How to undo: simply remove the extra filter from the chain.

Clipping on laughter (what causes it: gain set for average speech, no headroom for excitement). How to check: watch the meter during a loud laugh — if it hits 0 dBFS or shows red, you are clipping. What to do: reduce gain until laughs peak at −6 dBFS, then use compression to lift the quieter parts. How to undo: raise gain back and accept the clipping if you need raw loudness in an untreated room.

Chopped word endings (what causes it: gate release too fast or threshold too high). How to check: listen for words that trail off into silence mid-syllable, especially quiet sentence endings. What to do: lower the threshold by 3–4 dB and lengthen release to 150 ms. How to undo: return to the previous values if the noise floor becomes audible again between phrases.

FAQ

Should the lavalier point up or down on my chest?

Up. Facing the capsule upward keeps it shielded by the clip and clothing above it, which reduces both plosive blasts and fabric contact. Downward-facing capsules catch every downward sliding fiber and shed clothing dust onto the mesh. If your clip only allows one orientation, favor up and add a windscreen.

What distance should the mic be from my mouth?

For a chest mount, aim for 6–8 inches below the chin — roughly 8–12 inches from your mouth. Every doubling of distance costs about 6 dB of level, which is why gain and compression matter more here than with a headset mic. If your commentary involves lots of head movement, place the mic lower for consistency and let compression handle the level.

Do I need a windscreen indoors?

Often, yes. Even indoors, a chest-mounted lav rubbing against fabric produces wind-like low-frequency bursts that a foam windscreen or fur “deadcat” tames far better than any filter. The trade-off: foam slightly softens the highest sibilance frequencies, which is usually a fair price. If your recording is perfectly still and clean, test with and without — the difference is audible in the first ten seconds of playback.

Why does my lav sound worse than the same model in reviews?

Reviews record in treated rooms with the mic a hand’s width from the mouth, often on a boompole or hidden very close to the collar. Chest placement doubles or triples that distance and adds fabric and room reflections. Match the settings in this guide — proper placement, cable loop, high-pass filter, and a single moderate compression pass — and you will recover most of the gap without buying anything new. By the time these settings are dialed in, most viewers in 2026 cannot tell a well-configured budget lav from a premium one in a gaming video.

Related guides