How Far Should Your Microphone Be From Your Mouth?

Read this site's audio hub the way a buyer would — mic guide, boom arm guide, headset guide — and you will find advice about capsules, patterns, sample rates, weight capacities and cable channels. What you will not find is a straight answer to the question that governs all of it: how far should the microphone actually be from your mouth?

It is not that we avoided the question. It is that we answered it four different ways without noticing. Our Blue Yeti write-up positions it 6–8 inches away. Our closing FAQ said a boom arm lets you position the mic 4–6 inches away. Our HyperX SoloCast entry says the capsule is fine "at normal speaking distances" and leaves it there. And our Best Overall pick, the Shure MV7+, carried no distance at all — in a guide whose entire argument for it is that it rejects background noise. Writing this piece is what surfaced that, and the microphone guide has since been corrected: the MV7+ entry and the two FAQs involved now carry the figures below.

Shure publishes the number. The MV7's own documentation gives two positions: a Near setting for "1–6 inches away from your mouth" and a Far setting for "6–18 inches away." That is the manufacturer of our top pick specifying a working distance that starts three times closer than the closest figure anywhere in our guide, for the one microphone we gave no figure to at all.

The boom arm guide has the same shape of gap, and it is the version this site has learned to look for: a hub that names a spec in its own criteria and then never quotes it. Our boom-arm criteria list "reach and range — enough to bring the mic to your mouth without crowding the frame." The comparison table publishes maximum mic weight, cable routing and, for one arm, 29 inches of reach. It never publishes a working distance, because working distance is a property of where you put the arm, and no manufacturer can print it on a box.

This piece is the missing number and the arithmetic behind it. The short version is that distance is not one factor among several — it is the dominant one, it moves in large steps, it is free, and it is the only variable in desk audio that you control completely and that nobody can sell you.

Sound from a point source falls off with the inverse square of distance, which in decibels is a clean and slightly startling rule: every doubling of distance costs 6 dB, and every halving gains 6 dB. Your voice obeys it. The noise in your room mostly does not — a fan, an air conditioner, traffic through a window and a partner on a call in the next room arrive as roughly diffuse sound that is about the same level everywhere on your desk.

That asymmetry is the whole game. Moving the microphone closer does not make your room quieter. It makes your voice louder relative to a noise floor that stays put, and it does so in 6 dB steps, which are enormous. Six decibels is a doubling of sound pressure and a quadrupling of power.

The arithmetic needs one anchor, and it is a published one. DPA gives the sound pressure level of speech at normal vocal effort as 58 dB at one metre, with raised voice at 64 and loud speech at 70. ANSI S3.5 uses approximately 60 dBA for the same condition; 58 is the more conservative number, so that is the one used below. Extrapolate it inward and you get the level arriving at a capsule at any desk distance.

For the room, the reference figures are equally well established. Archtoolbox's room sound level table puts a private office or workroom at 40–45 dBA and an open office at 45–55 dBA. A home office with an air conditioner running, a desktop fan, or a household happening around it sits comfortably in the upper band. The table below uses 40 dBA for a genuinely quiet room and 50 dBA for a normal one.

The threshold column matters as much as the levels. DPA's guidance is that intelligibility holds up when speech maintains around 15 dB(A) of signal-to-noise ratio against a background above 40 dB(A). That gives a pass mark to hold the geometry against — and in a 50 dBA room, a microphone sitting on its bundled desk stand behind your keyboard lands right on it.

Two honest caveats, both of which make the far end of the table worse rather than better. Your mouth stops behaving like a point source in the last inch or two, so the very closest rows deliver slightly less than the formula promises. And past a room's critical distance, reflected sound stops falling off at all — so at 24 or 36 inches in an untreated room, the real figures are lower than these, not higher. Nothing in this table is optimistic about sitting far away.

Every microphone guide on the internet, ours included, leads on the polar pattern. Our own criteria list opens with "voice clarity — cardioid pickup pattern that rejects keyboard and HVAC noise," and the category page says cardioid mics reject keyboard clatter behind them. That is true. The question nobody asks is how much it is worth, and the answer is published and specific.

The figure is random energy efficiency: a microphone's response to diffuse sound relative to its on-axis response. For a cardioid it is 0.333, or −4.8 dB. That is the number. A cardioid microphone rejects diffuse room noise — the HVAC drone, the traffic, the reverberant wash of the room — by 4.8 dB compared to an omnidirectional mic, and no more. A hypercardioid gets you to 6 dB. That is the entire ceiling of the technology.

Set that against the previous section and the comparison is almost unfair. The complete benefit of the directional pattern the category sells on is 4.8 dB. Moving your microphone from eight inches to four inches is 6 dB. One is a design feature you pay for; the other is you leaning in, or spending twenty minutes repositioning an arm.

It gets worse for the pattern when you check it against the two noise sources our criteria actually name. HVAC is diffuse, so it gets the 4.8 dB. A keyboard is not behind the mic at all — on a normal desk it is below and in front of it, somewhere near the 90-degree axis, where a cardioid is down 6 dB, not nulled. The null is at 180 degrees, pointing at the wall behind you. So against the specific noises we cite, the pattern is worth roughly 5 to 6 dB, and the keyboard is the one it handles worst because of where keyboards physically are.

None of which is an argument against buying a cardioid microphone. It is an argument about ordering. Pattern is a fixed, one-time 4.8 dB that you buy once and cannot improve. Distance is 6 dB per halving, repeatable, free, and available right now on the microphone you already own. The category leads with the small fixed term and omits the large variable one — and that is the whole finding.

Our Shure MV7+ entry made a claim that appears in almost every microphone guide ever written, and this site should not have repeated it without qualification. It read: "Dynamic capsules are inherently less sensitive than condensers — they reject keyboard clicks, air conditioning drone, and street noise that condenser mics pick up readily." That sentence has since been rewritten, but it is worth spelling out why, because the same claim is still on almost every competing page.

The advice that follows from it is correct. The mechanism is not. Sensitivity is a transduction efficiency figure — how much voltage a capsule produces per unit of sound pressure. It is not a directivity figure. A cardioid dynamic and a cardioid condenser sitting at the same distance from your mouth in the same room receive the same ratio of voice to room noise, because that ratio is set by the polar pattern and the geometry, and both of those are identical. Lower sensitivity means you turn the preamp up, which raises the voice and the room noise by exactly the same amount.

The same error was in our closing FAQ, phrased as a mechanism: "Closer positioning means lower gain, which reduces room noise pickup." Gain does not reduce room noise pickup. Gain multiplies everything downstream of the capsule equally. What closer positioning does is raise your voice relative to the room before the gain stage, and the fact that you then need less gain is a consequence of the improvement, not the cause of it. Both sentences described a real effect and credited the wrong term, and both have been corrected in the guide.

So why do dynamics genuinely sound cleaner in bad rooms? Because of how they get used. A dynamic capsule's low output pushes you to work it close — Shure's own Near mode starts at one inch — and because dynamics are typically built with high maximum SPL and heavy low-frequency handling, they tolerate being that close. Condensers are sensitive enough to sound fine at 8 or 12 inches, so people place them at 8 or 12 inches. The transducer type does not reject the room. It changes the distance you end up working at, and the distance rejects the room.

The practical consequence is worth stating plainly, because it cuts against the upsell in our own guide: a condenser worked at four inches will beat a dynamic worked at twelve, every time, by about 9 dB. If you own a Yeti or a SoloCast and you have been considering a $249 dynamic to fix a noisy room, move the mic you own first. The upgrade may turn out to be unnecessary, and if it is not, you will at least be comparing the two at the distance that makes the dynamic worth buying.

Closer is better for signal-to-noise without limit, but it is not free, and a guide that only tells you to get closer is giving you half the picture. Three things push back.

Proximity effect. Directional microphones work on a pressure gradient, and at close range that produces a low-frequency lift — commonly 6 to 12 dB of bass boost within about six inches, centred around 200 Hz. Close-miked without correction, this is the boomy, muddy voice that makes people sound like they are talking into a barrel. It is also completely fixable, and the fix is a high-pass or low-cut filter, which is why broadcast dynamics have a bass rolloff switch and why the MV7+ has a high-pass filter in its MOTIV software. Our guide lists that filter among a bundle of DSP features; it is more specific than that. It is the thing that lets you work the microphone at two inches.

Plosives. The blast of air from a P or a B is a physical event, not an acoustic one, and it scales badly as you close in. Inside about four inches you want either a pop filter, a foam windscreen, or — the free option — the mic offset so you talk across it rather than into it, typically positioned slightly above your mouth and angled down toward it. Speaking past the capsule at 30 to 45 degrees costs almost nothing in level and removes most of the problem.

Level and consistency. At two inches your voice is arriving at about 84 dB SPL, with peaks 15 to 20 dB above that, so the microphone needs headroom. Every mic in our guide has it. The harder problem is that at very close range, small head movements are large proportional changes in distance — leaning back four inches from a two-inch position costs you 9 dB, while the same movement from twelve inches costs 2.5. Close-miking buys signal-to-noise and spends positional consistency, which is precisely what compressors and auto-level modes exist to manage, and why Shure ties its Near and Far settings to automatic gain.

Weighing all three, the working ranges below are the practical answer. The important column is the last one: what the distance costs you in signal-to-noise against the best case. Note that our own published 6–8 inch figure for the Yeti is reasonable for a condenser, our 4–6 inch FAQ figure is right for a boom-arm setup, and the gap between them is real but small. The number that is badly wrong is the one nobody wrote down — the mic left on its bundled tripod behind the keyboard, which is where most microphones in the world actually are.

Our boom arm guide sells arms on weight capacity, damping quality, cable routing and desk footprint. All of those are real, and the guide is not wrong about any of them. But it frames the arm as a tidiness and convenience product — a way to get the mic off the desk and hide the cable — when the arithmetic says it is doing something much larger.

Take the two positions most people actually choose between. A microphone on its bundled tripod, pushed back behind the keyboard so it does not block the monitor, sits around 18 inches from your mouth. The same microphone on an arm, brought in to four inches, gains 13 dB. That is a factor of 4.5 in sound pressure and 20 in power, and it is the single largest improvement available anywhere in this category.

Now compare that to the upgrade path the category actually sells. The difference between our $50 HyperX SoloCast and our $249 Shure MV7+, in terms of the ratio of your voice to your room, is small — both are cardioids, so both get the same 4.8 dB from the pattern, and self-noise differences only matter when the source is quieter than the microphone's own noise floor, which a voice at four inches emphatically is not. A $25 boom arm moves the number further than a $200 microphone upgrade does. We have a guide that sells the arm and a guide that sells the upgrade, and until now neither said so.

Which makes reach the wrong headline spec, and this is where the boom-arm guide's comparison table quietly misleads. Reach is horizontal, and the constraint is not horizontal. Our stands explainer establishes that a correctly seated user's eyes sit 20 to 22 inches above the desk surface whatever their height; the mouth is several inches below that, so the microphone's target position is somewhere in the low-to-mid teens above the desktop and 4 to 6 inches out from your face. The arm has to reach up and in, past the near edge of a monitor, without entering your sight line or your webcam frame. An arm with 29 inches of reach that cannot get its last joint above 12 inches is useless for the actual job.

That reframes two picks in our own guide rather than contradicting them. The low-profile arms — the Elgato Wave Mic Arm LP in particular — are not merely a stylistic alternative to a tall parallelogram boom; they solve the vertical-clearance problem from below rather than above, coming up under the monitor instead of over it. And the capacity spec matters for a reason the guide only half states: an arm that cannot hold position at full extension will droop away from your mouth over an afternoon, and drooping away is the failure mode that costs decibels. Damping is not a luxury. It is what keeps the 13 dB you bought.

Our wireless headset guide already makes the right call, and it makes it on the right grounds — it says plainly that a boom beats hidden beamforming in loud rooms. What it does not do is quantify it, and once you do, the conclusion gets uncomfortable for the rest of the hub.

A headset boom puts the capsule one to two inches from your mouth and, crucially, keeps it there regardless of how you move, because it is attached to your head. Against a microphone on its bundled desk stand at 18 inches, that is a 19 dB advantage. Against a well-positioned desk mic on a boom arm at four inches, it is still about 6 dB. There is no microphone, no capsule technology and no amount of software processing that recovers 19 dB.

Run it against the intelligibility threshold and the practical stakes become clear. In a quiet 40 dBA home office, everything works — even the bundled tripod clears the bar with room to spare, which is exactly why so many people never discover they have a problem. In a 50 dBA room — an air conditioner, a street-facing window, a household in motion — the desk mic on its stand lands at about 15 dB before the polar pattern helps, right on DPA's threshold, and every honest description of that is "understandable, and tiring to listen to for an hour." The headset is at 34 dB.

So the recommendation this hub should have been making, and now does: if your calls happen in a room you do not control, buy the headset, not the microphone. Our Logitech Zone Vibe 100 pick is around $99 and our Jabra Evolve2 65 is $180–290; both put a boom where a $249 desk microphone physically cannot. The desk mic is the right purchase when your room is quiet and you want to sound good rather than merely clear — for recording, for streaming, for a podcast, for the times you want to be on camera without hardware on your head. Those are real reasons and we stand behind the guide. They are just not the same reason as "I need to be understood in a noisy room," which is the problem most people think they are solving.

One caveat in the headset's disfavour, since this section is otherwise a commercial for it: proximity to the mouth also means proximity to breath, and boom position matters more on a headset than people assume. Most manufacturers specify the boom should sit at the corner of the mouth rather than directly in front of it, for exactly the plosive reasons in the previous section.

Our microphone guide is titled for Video Calls, and its comparison data includes sample rate and frequency response for every pick: 24-bit/192 kHz on the Audio-Technica ATR2100x, 16-bit/48 kHz on the Yeti and the SoloCast, frequency responses of 20 Hz–20 kHz on three of the five. Those figures are accurate. They are also, for the stated use case, almost entirely decorative.

Video call platforms do not carry them. Zoom, Teams, Meet and the WebRTC applications behind most of the rest run speech codecs — Opus, SILK, and Microsoft's Satin — which operate in bandwidth modes: narrowband at an 8 kHz sampling rate, wideband at 16 kHz, super-wideband at 24 kHz and fullband at 48 kHz. A wideband call carries audio content up to about 8 kHz. Zoom's own description of its high-fidelity music mode is that it raises the codec to 48 kHz and disables the default post-processing — which tells you what the default is not.

Hold the specs against that and two things fall out. The ATR2100x's 192 kHz converter is running at roughly twelve times the sample rate the call will transmit. And more pointedly: the ATR2100x's frequency response of 50 Hz–15 kHz is the narrowest in our table, the one thing that makes our budget pick look technically outclassed — and 15 kHz is still nearly double the bandwidth a wideband call carries. The spec that penalises it in the comparison table cannot affect the use case the guide is about.

The clearest evidence that these numbers are decorative is that a manufacturer deleted one and improved the product. Elgato's Wave:3 MK.2 dropped the 96 kHz sample rate of the MK.1 entirely, keeping only 48 kHz. In exchange, the MK.2 raised dynamic range from 95 dB to 110 dB, raised maximum SPL from 120 dB to 130 dB, and added an on-microphone DSP chip whose five effects include a low-cut filter. Read that list against this article: they gave up the number that cannot reach the far end and spent the budget on headroom, dynamic range and low-frequency correction — which are exactly the three specifications that close-miking stresses. The revision is an endorsement of the thesis by the only party with a commercial reason to resist it.

That correction applies to us directly, and it is the second run in a row that reading every pick's manufacturer page has caught a description of a superseded product. Our guide lists the Wave:3 at "96 kHz / 24-bit," which describes the MK.1; the currently-sold MK.2 is 48 kHz. That spec has been corrected in the guide, along with the MK.2's revised dynamic range and max SPL. The lesson generalises past this hub: when a manufacturer removes a headline spec from a revision and the reviews get better, the spec was never doing any work — and any comparison table still ranking products by it is ranking them by nothing.

None of this requires new gear. It requires moving the gear you have, and the order below is deliberately arranged so that everything free comes before anything that costs money.

1. Measure what you have now. Sit as you actually sit on a call and measure from your mouth to the microphone capsule. Most people are surprised — the bundled tripod behind the keyboard is usually 16 to 20 inches, which the table above puts at the edge of intelligibility in a normal room.

2. Halve it. That is 6 dB, immediately, for free. If the mic is on a stand, bring the stand forward and raise it. If that puts it in your camera frame, that is the actual constraint you are solving for, and it is a geometry problem rather than a hardware one.

3. Talk across it, not into it. Position the capsule slightly above your mouth and angled down, so your breath passes over it rather than into it. This costs almost nothing in level and removes most plosive problems, which is what lets you go closer still.

4. Turn on the low-cut filter. If you are now inside six inches, proximity effect is adding 6 to 12 dB of bass. Every serious mic has a high-pass filter in hardware or software — the MV7+ has one in MOTIV, the Wave:3 MK.2 has one on the microphone itself — and conferencing apps generally do not apply one for you. This is the step that makes close-miking sound professional instead of boomy.

5. Measure again, and find out whether you still have a problem. Re-measure from the seated position you actually use. If you are now inside six inches, you have collected 9 to 13 dB without spending anything — more than the gap between the cheapest and the most expensive microphone in our guide — and the honest outcome may be that you are finished. What the second measurement is really for is naming the constraint if you are not: either the mic physically cannot reach your face from where it has to live, or it can and your room is still too loud behind it. Those are two different purchases, and the bottom line below says which is which.

The one-sentence version: you are not shopping for a microphone, you are shopping for a distance — and the distance is free.