Background noise is the fastest way to make a good mic sound cheap. A humming PC fan, a buzzing air conditioner, clacking keys, the neighbor’s dog, all of it rides right into your voice track and drags the whole thing down. The good news is that most of it is fixable in about twenty minutes without buying anything, and the rest comes down to a few software toggles and smarter placement. This walkthrough covers both the free fixes and the small upgrades that genuinely help, so your voice lands clean on Discord calls, streams, and recordings. If your mic itself is the weak link, our guides to the best USB microphones for podcasting and the best streaming microphones are worth a look after you’ve squeezed everything you can out of the setup you own.
What you’ll need
Almost nothing, honestly. You need your existing mic or headset, a quiet-ish room, and about twenty minutes. Software-wise, Windows Sound settings and whatever app you talk in (Discord, OBS, Zoom) already have the noise controls built in, so there’s no separate download for the basics. If you want a hardware assist, a headset with a dedicated noise-canceling mic does a lot of the heavy lifting before software ever gets involved. The Picun G2 Wireless Gaming Headset runs about $59.99 and holds a 4.5 rating, and its ENC (environmental noise cancellation) mic is built to strip steady background sound out of your voice at the source. It connects over 2.4GHz, Bluetooth 5.4, or wired, with a rated 5ms latency and 100-hour battery, so it doubles as a daily driver and not just a call mic. A boom arm or a foam windscreen helps too, but you can start with what’s on your desk right now.
Pros
- Three connection modes on a single headset: 2.4GHz dongle, Bluetooth 5.4, and 3.5mm wired fallback.
- Vendor-rated 5ms 2.4GHz latency keps audio synced in competitive shooter scenarios.
- Long rated runtime, up to 100 hours in Bluetooth with RGB disabled for travel use.
- RetractableENC microphone tucks away, useful when switching between gaming and general listening.
Cons
- Limited owner feedback at time of writing, real-world reliability and QC signal is unverified.
- 7.1 surround is virtual DSP, not true multi-driver, so positional accuracy varies by title andear.
- RGB lighting materially cuts battery life, cited100-hour figure requires lighting turned off.
The Picun G2 is a budget-tier wireless gaming headset built around 40mm drivers, virtual 7.1 surround, and tri-mode connectivity across2.4GHz, Bluetooth 5.4, and 3.5m wired. Target buyer is a multi-platform player rotating between PC, PS5, PS4, and Switch who wants low-latency wireless without moving up to established brand pricing.
The defining feature is the tri-mode radio stack. The 2.4GHz USB-A or Type-C dongle is rated at 5ms latency, which is competitive for the tier and adequate for most FPS titles, while Bluetooth 5.4 handles phones and tablets. Headsets in this class typically deliver serviceable rather than reference-grade imaging, and virtual 7.1 leans on DSP rather than discrete drivers.
Trade-offs are typical at this tier. Battery figures are optimistic: 100 hours assumes Bluetooth with RGB off, and 2.4GHz drops the number to 42 to 45 hours. Driver material, impedance, frequency response, clamp force, and weight are not specified, so tonal balance and long-session comfort cannot be verified from source data alone.
Buy this if you want one headset that jumps between a PS5, a Switch dock, and a work laptop over Bluetooth, and you accept spec-sheet-only marketing. Skip this if you rely on measured frequency response, discrete suround, or detachable bom mics with broadcast-grade pickup for streaming.
Drivers and tuning: Picun specifies 40mm dynamic drivers paired with virtual 7.1 surround processing. Impedance, sensitivity, and frequency response are not specified, so bass extension and treble air cannot be quantified from source. Virtual 7.1 is DSP-based positional widening, useful for FPS cues but not equivalent to true multi-driver surround.
Connectivity and latency: Three modes are covered: 2.4GHz via USB-A or USB-C dongle, Bluetooth 5.4, and 3.5mm analog. The 2.4GHz link is rated at 5ms latency. Bluetooth codec support (SBC, AAC, aptX) is not specified, which maters for wireless music quality on phones and tablets.
Battery and power: Rated up to 100 hours on Bluetooth with RGB off, and 42 to 45 hours on 2.4GHz wireless. Play-while-charging is supported, so a USB cable can extend sessions past the rated ceiling. Charge time and battery capacity in mAh are not specified.
Microphone: Retractable omnidirectional bom with ENC noise reduction. Sample rate, bit depth, and pickup pattern beyond omni are not specified.ENC filtering targets background noise on the transmit side, adequate for party chat but below dedicated USB mic quality.
Step 1: Fix mic placement before you touch any software
This is the single biggest lever and it costs nothing. Every mic has a pickup pattern, and the closer your mouth sits to it, the louder your voice is relative to the room. That ratio is the whole game. Get the mic within four to six inches of your mouth and slightly off-axis (angled at the corner of your lips, not straight on) so plosives don’t thump the capsule. A headset boom mic should sit about two finger-widths from the corner of your mouth, not floating up by your cheekbone where it hears more room than voice.
Placement also decides how much steady noise you’ll fight later. Point a cardioid mic away from your PC tower and your AC vent, since the back of a cardioid rejects sound. Keep it off the desk surface if you can, because typing and mouse thuds travel straight through the wood into the stand. A cheap boom arm or even a folded towel under a desktop stand cuts that structural rumble noticeably. Nail this step and every software fix afterward has less work to do.
Step 2: Set the right input device and calm the gain
Open Windows Sound settings, go to Input, and confirm the app is actually using the mic you think it is. People spend an hour fighting noise on the wrong device all the time. Once you’ve picked the correct input, click into its properties and watch the level meter while you talk normally. You want your speaking voice to peak in the upper-middle of the meter, not pinned to the top and not so low you’re barely moving it.
Here’s why gain matters for noise specifically. If the input level is set too low and you’re compensating by cranking software gain or leaning in and shouting, you amplify the room’s noise floor right along with your voice. Set a healthy hardware level first, then leave software gain modest. On a USB mic that means the gain knob or the Windows level slider around 70 to 85 percent, adjusted so normal speech is clear without clipping. Lower the level a touch if you hear hiss rise every time you go quiet. That hiss is your noise floor, and a sane gain structure keeps it buried.
Step 3: Turn on software noise suppression
Modern noise suppression is genuinely good, and it’s already sitting in your apps for free. In Discord, open Voice & Video settings and switch Noise Suppression to Krisp; it’s remarkably effective at killing fan hum, keyboard clatter, and steady hiss while leaving speech intact. In OBS, add a Noise Suppression filter to your mic source and choose the RNNoise method, which is the stronger of the two options for most rooms. Zoom, Teams, and Google Meet all have their own background-noise settings buried in audio preferences, usually with low, medium, and high levels.
Windows itself also offers a shortcut. In Sound settings under your mic’s properties, look for audio enhancements, though results vary by driver and it can dull your voice. Don’t stack three suppressors on top of each other. Pick one strong layer, Krisp in Discord or RNNoise in OBS, and stop there. Over-processing makes you sound like you’re talking through a tin can, with words cutting in and out as the algorithm chews on quiet syllables. One good filter beats three mediocre ones.
Step 4: Add a noise gate and light EQ
A noise gate is the tool that mutes your mic when you’re not talking, so silence stays silent instead of leaking room tone. In OBS, add a Noise Gate filter to the mic and set the close threshold just below the quietest sound you make and the open threshold a few decibels above it. A typical starting point is a close threshold around -40dB and an open threshold near -32dB, then tweak by ear. If the gate chops the start of your words, lower the open threshold and lengthen the attack.
A gentle high-pass filter helps too. Rolling off everything below roughly 80 to 100Hz removes low rumble, desk vibration, and AC hum without thinning your voice, since almost nothing useful in speech lives down there. Add a small EQ dip around 200 to 400Hz if you sound boomy or boxy from a close mic. Keep the moves subtle. The goal is a clean voice that still sounds like you, not a heavily scrubbed robot. Once the gate and high-pass are dialed in, that constant background bed usually disappears.
Step 5: Kill the sources software can’t
Some noise is easier to remove at the source than to filter out later. Your PC is often the culprit. Dust-clogged fans spin up loud, so a can of compressed air and a fan-curve tweak in your BIOS or a tool like FanControl can drop the ambient roar by several decibels. If your mic sits near the tower, move it or the tower. A mechanical keyboard with loud switches will always fight you on a live mic, so push it slightly out of the pickup zone or add a desk mat to deaden the impact.
Room reflections matter as well. A bare room with hard walls bounces sound around and adds a hollow echo no gate fixes cleanly. Soft stuff absorbs it. Curtains, a rug, even a blanket behind you tame reflections for free. And if you’ve done all of this and still hear hiss, the mic hardware may be the limit. A headset with a dedicated ENC mic, like the Picun G2 earlier, cancels steady noise in hardware before it’s recorded. For a broader look, our roundup of the best gaming headsets breaks down mic quality across price tiers.
Troubleshooting common issues
My voice sounds robotic or cuts out
That’s over-suppression. You’ve likely stacked noise cancellation in multiple apps or set the gate threshold too high. Turn off one layer, drop the gate’s open threshold, and lengthen the attack so it doesn’t clip the front of words. Aim for one strong filter, not a pile of them. Your voice should breathe naturally with a little room tone left in.
There’s a constant hum or buzz
A low 60Hz hum usually means a ground loop or interference, common with cheap USB hubs and unshielded cables. Plug the mic directly into a rear motherboard USB port instead of a hub, and keep the cable away from power bricks and monitor cables. A high-pass filter tames what’s left. If it’s a whine that changes with GPU load, that’s coil or electrical noise bleeding in, and a powered USB port or a different cable path usually fixes it.
Suppression works but my keyboard still comes through
Keyboard clatter shares frequencies with speech, so software struggles to separate it. Physical distance is your friend. Move the mic further from the keyboard, angle it away, and add a desk mat to soften each keystroke. Krisp and RNNoise handle a lot of it, but the last bit is a placement problem, not a filter problem.
The noise gate keeps cutting off my sentences
Your open threshold is too aggressive or your attack is too slow. Lower the open threshold so quieter speech still trips the gate, and shorten the attack so it opens the instant you start talking. A longer release, around 150 to 250ms, keeps the gate from slamming shut between words. Small changes here make a big difference.
Before and after expected outcomes
Done right, the difference is obvious the moment someone hears you. Before, listeners catch a steady bed of fan hum, occasional key clatter, and a hollow room echo that makes you sound distant. After placement, gain, one suppression layer, and a gate with a high-pass filter, the background drops away and your voice sits forward and clear. You won’t get silent-studio perfection from a $60 setup, and that’s fine. The realistic target is a clean, present voice where the room simply isn’t part of the conversation anymore.
Variations and advanced setups
If you’re on an NVIDIA RTX GPU, the free Broadcast app runs an AI noise-removal model that’s stronger than most built-in options and works across every app system-wide. Streamers and podcasters who want the cleanest result often move to a dedicated audio interface with a real XLR mic, which sidesteps USB noise entirely and gives you hardware gain control. Our guide to the best audio interfaces for streaming covers that path. For a fully separate voice chain, a hardware noise gate or a small mixer with a compressor lets you dial in a broadcast sound before the signal ever hits your PC.
Common questions
Does noise suppression lower my voice quality?
A little, if you push it hard. Aggressive suppression trims frequencies to remove noise and can make your voice sound thinner or slightly digital. Used at a moderate level with good mic placement, the tradeoff is minor and worth it. The trick is one solid layer instead of stacking several, which is what actually wrecks the sound.
Is a headset mic or a standalone mic better for background noise?
A headset mic wins on consistency because it stays close to your mouth no matter how you move, keeping your voice loud relative to the room. A dedicated desk mic can sound richer but picks up more space unless you keep it close and gated. For noisy environments, a close boom mic with ENC, like the Picun G2’s, is often the simpler path to clean audio.
Why does my mic pick up game audio or my own speakers?
That’s your speakers bleeding into the mic. Switch to headphones and the problem vanishes instantly, since there’s no loudspeaker for the mic to hear. If you must use speakers, lower their volume and increase the mic-to-speaker distance. A noise gate helps between words, but headphones are the real fix.
Can I reduce background noise after recording instead of live?
Yes. Free tools like Audacity have a Noise Reduction effect where you sample a second of pure room tone, then remove that profile from the whole track. It’s very effective for steady noise like fans or hiss. For live calls you’re stuck with real-time suppression, but for recorded podcasts and voiceovers, cleaning it in post gives you the most control.
Write Your Review
No reviews yet. Be the first to share your experience!