Microphone Selection and Placement, Done Properly

Microphone selection and placement are the two decisions that shape a voice recording more than anything that happens afterward. The microphone you choose decides how the capsule responds to the voice, the room and the noise around it. Where you put it decides the balance between those three things. Once the balance is recorded, it is baked in. Equalization, noise reduction and compression can polish a good recording, but they cannot fully pull a voice back out of a room it was recorded into.

This guide is for anyone who records speech for a living or for their business: podcast hosts, marketing teams recording video, in-house trainers, voice talent working from home, and producers setting up interviews in offices and hotel rooms. It answers the questions people actually ask when a recording sounds thin, boomy, echoing or noisy, and it explains the reasoning behind each answer so you can adapt it to your own room and equipment.

The answers lean on a few pieces of physics that do not change from one brand of microphone to another: the inverse square law, the proximity effect, the shape of polar patterns and the way hard rooms reflect sound. Get those right and a modest microphone will outperform an expensive one placed badly.

Why does microphone selection and placement matter more than editing?

Because the microphone records the ratio of voice to everything else, and editing can only change that ratio a little before the voice itself starts to suffer. A microphone does not know which sounds you want. It captures the direct sound from the speaker's mouth, the reflections of that same voice bouncing off walls, desks and windows, and every other sound in the space: air conditioning, computer fans, traffic, the next room. Placement and pattern choice decide how loud each of those is relative to the voice.

Consider what happens when you try to fix the problem later. Noise reduction tools work by identifying what is not voice and turning it down. When the noise is steady and well below the voice, that works cleanly. When the noise is loud relative to the voice, the tool has to cut deeper, and it starts removing parts of the voice that overlap with the noise in frequency and time. The result is the familiar watery, lisping, underwater sound of over-processed dialogue. Reverb reduction has the same limit: room reflections arrive a few milliseconds after the direct sound and share its exact tone, so separating them is far harder than separating a hum.

The practical consequence is simple. Every decibel of improvement you get at the microphone is a decibel you do not need to claw back with processing, and processing always costs something. Ten minutes spent moving a microphone closer and turning it away from a noisy window routinely saves hours of repair work and produces a better result than the repair ever could.

There is also a consistency argument. Editors can match two recordings made with the same microphone at the same distance in the same room. They struggle to match a close, dry recording with a distant, echoing one, because the difference is not just tone but space. For series content such as podcasts, courses and recurring video, the setup is the brand's sound, and the setup is mostly placement.

Which type of microphone should you choose: dynamic, condenser, lavalier or shotgun?

Choose a dynamic microphone for untreated rooms and close speech, a condenser for quiet, treated spaces where you want detail, a lavalier when the speaker must be hands-free or on camera, and a shotgun when the microphone has to stay out of the shot. The right answer depends far more on the room and the job than on the price of the microphone.

Dynamic microphones

Dynamic microphones use a diaphragm attached to a coil moving in a magnetic field. They are generally less sensitive than condensers, which sounds like a disadvantage but is often the opposite for speech. Lower sensitivity means that a voice a few centimeters away is strongly favored over sounds a meter or two away, so the room and background noise are recorded more quietly relative to the voice. Dynamics need no power, tolerate high sound levels and are physically robust. Their main demand is gain: many broadcast-style dynamics need a clean, high-gain preamplifier or an inline booster, because their output is low.

Condenser microphones

Condenser microphones use a charged diaphragm next to a fixed backplate. They need power to operate, usually phantom power supplied by the interface or mixer, commonly 48 volts. They are more sensitive and generally capture more high-frequency detail and air than dynamics. In a quiet, acoustically treated booth that detail is flattering. In an untreated room with hard walls, that same sensitivity records the reflections and the refrigerator hum with great fidelity. Condensers in untreated, echoing rooms are one of the most common reasons a recording sounds amateur despite expensive equipment.

Lavalier microphones

Lavaliers are small condensers clipped to clothing, usually on the sternum or near the collar. They keep a consistent distance from the mouth as the speaker moves, which is valuable for presenters and interviews on camera. The trade-offs are clothing rustle, a slightly chesty tone because the capsule sits below and off-axis to the mouth, and dependence on either a cable or a wireless transmitter. If you use wireless systems, the planning in our guide to choosing wireless microphone frequencies will save you dropouts and interference on the day.

Shotgun microphones

Shotgun microphones are highly directional condensers designed to be pointed at a source from a boom pole or camera mount. They reject sound from the sides better than most patterns, which is why they are standard on film sets. They are not magic long-range microphones: they still obey the inverse square law, and indoors their rear and side pickup still captures room reflections. A shotgun half a meter above the speaker sounds good. A shotgun on a camera three meters away mostly sounds like the room.

USB and headset microphones

USB microphones combine a capsule, preamplifier and converter in one body. Headset microphones place a small capsule a few centimeters from the mouth on a boom. Headsets look less polished on camera but solve the distance problem completely, which makes them excellent for live events, gaming-style streams and remote meetings where people move around.

The distances above are working ranges used by many engineers rather than rules; start there, record a test and adjust by ear.

Which polar pattern should you use, and where should the rear of the microphone point?

For a single speaker in an ordinary room, use a cardioid or a tighter pattern and point the rear, the least sensitive side, at the loudest noise source. A cardioid is most sensitive at the front and least sensitive at the rear, so it rewards careful aiming more than any other single decision.

Polar patterns describe how sensitive a microphone is to sound arriving from different directions. Omnidirectional microphones hear roughly equally in all directions. Cardioid microphones have a heart-shaped pickup with a deep null at the back. Supercardioid and hypercardioid patterns are narrower at the front but have a small rear lobe, so their deepest rejection sits at an angle behind the microphone rather than directly behind it. Figure-8 microphones hear front and back equally and reject the sides. The full detail, including how off-axis tone changes with each pattern, is in our guide to microphone polar patterns.

Aiming is where the pattern pays off. Walk the room before you set up and list the noise sources: an air vent, a window onto the street, a computer tower under the desk, a door to a busy corridor. Then position the speaker so the microphone's null points at the worst one. With a cardioid, that means the back of the microphone faces the noise. With a supercardioid or hypercardioid, rotate slightly so the noise sits in the angled null, typically well off the rear axis, rather than directly behind where the small rear lobe listens.

Two practical points are often missed. First, the pattern applies to reflections too. A cardioid facing a bare wall with the speaker between it and the wall will hear the voice bounce off that wall right into its most sensitive side. It is often better to have the speaker face into the room with a soft surface, such as a bookshelf, curtain or sofa, behind them, so that what the front of the microphone sees is absorbent. Second, patterns are frequency dependent. Most directional microphones become less directional at low frequencies, so a low rumble from a heating system will leak in even if the microphone is aimed well. That is a job for a high-pass filter, which is safe on speech in most cases.

When should you choose omnidirectional? Omnis have no proximity effect and handle breath and movement more gracefully, which is why most lavaliers are omni. In a well-treated room, an omni placed close can sound more natural than a cardioid. In a noisy or reflective room, it will record more of the problem.

How close should the microphone be to the speaker?

Close enough that the voice dominates the room, and consistent enough that the speaker cannot drift in and out of it; for most spoken-word work with a directional microphone, that means somewhere between a few centimeters and about a hand span. This is the hardest question in the guide, because distance changes three things at once: level, room sound and tone.

Level and the inverse square law

The inverse square law says that the level of sound from a point source falls about 6 dB each time the distance from the source doubles. Move from 10 cm to 20 cm and the voice arrives about 6 dB quieter. Move to 40 cm and it is about 12 dB quieter. Move to 80 cm and it is about 18 dB quieter. A speaker who leans back from 10 cm to 20 cm in the middle of a sentence produces a level change large enough that listeners hear it as a volume jump.

Room sound and the direct-to-reverberant ratio

The room reflections, meanwhile, do not fall off in the same way. In an enclosed space, the reverberant sound builds up to a fairly even level throughout the room. So as the microphone moves away from the mouth, the direct voice gets quieter while the room stays roughly the same, and the ratio between them worsens. When you then raise the gain to bring the voice back up, you raise the room and the background noise by the same amount. This is why distant recordings sound hollow and echoing even when the level is fine.

Halving the distance is therefore one of the most powerful tools you have. In the free-field approximation, going from 30 cm to 15 cm gains the voice about 6 dB relative to the diffuse room sound and steady noise. No affordable acoustic treatment gives you that much improvement so cheaply.

Tone and the proximity effect

Directional microphones add bass as the source moves closer. This proximity effect is why radio voices sound warm and full at close range, and why the same speaker sounds boomy and muddy when they lean right onto the grille. The effect is progressive: every change in distance changes the low end, so a speaker who moves around produces a voice whose tone is constantly shifting, not just its level. Omnidirectional microphones do not show this effect, which is one reason they are chosen for lavaliers and for speakers who move a lot.

A worked example: two distances in the same room

Here is an illustrative scenario, not a client project, to show how the numbers interact. A marketing manager records a weekly update in a small office with plasterboard walls, a hard desk and a laptop fan. On a trial take with a cardioid condenser at 40 cm, the voice peaks around -18 dBFS with the gain set to a comfortable position, and the room tone between phrases sits around -55 dBFS. The recording sounds distant and slightly echoing.

Now move the microphone to 10 cm, two doublings closer. The voice rises about 12 dB, so to keep peaks in the same place the gain comes down by about 12 dB. The fan noise and the diffuse room reflections, which did not change with the move, drop by the same 12 dB, so the room tone now sits around -67 dBFS. The voice-to-noise ratio has improved by roughly 12 dB without buying anything. Switch the condenser for a dynamic microphone at 5 cm with a pop filter, and the improvement grows further, because the dynamic's lower sensitivity to distant sound favors the voice even more.

The trade-off is proximity effect. At 5 to 10 cm the voice will have more low end; a gentle high-pass filter and possibly a small low-shelf cut in editing tame it. That is an easy, transparent correction, unlike trying to remove 12 dB of room.

How to make the distance stick

People drift, especially when they are relaxed or reading from a screen. The fixes are physical. Use a boom arm or a stand that places the microphone at mouth height so the speaker does not lean down to it. Put the script or screen directly behind the microphone so reading keeps the speaker on-axis. Measure the distance with your hand the first time, a spread hand is a handy gauge, and note it in the session sheet. For people who move or gesture, switch to a lavalier or headset rather than asking them to sit still.

How do you stop plosives, sibilance and breath noise at the source?

Stop them by keeping the microphone out of the direct airflow: use a pop filter, angle the microphone slightly off-axis, or place it just above or beside the mouth rather than directly in front of it. Plosives are bursts of air from p and b sounds, and once a burst hits the diaphragm it records as a low-frequency thump that is awkward to remove cleanly.

Plosives

Say the word "puppy" with your hand a few centimeters in front of your mouth and you will feel the jet of air. A microphone diaphragm feels it too, and responds with a large low-frequency excursion. Close-range directional microphones are especially vulnerable, because proximity effect exaggerates the low end of that burst.

A pop filter, either a mesh screen a few centimeters in front of the microphone or a foam windscreen, breaks up the jet before it reaches the capsule. Angling works just as well and costs nothing: turn the microphone about 15 to 30 degrees off the line between mouth and capsule, or place it slightly above the nose line pointing down toward the mouth, so the airflow passes by rather than into it. Many engineers combine both approaches for close vocal work.

Sibilance

Sibilance is harshness on s, sh and t sounds. It is partly a property of the voice and partly a result of the microphone's high-frequency response and aim. Moving slightly off-axis often softens it, because many microphones are a little less bright off to the side. A condenser with a pronounced presence peak aimed straight at a sibilant speaker is a common recipe for trouble. If it persists, a de-esser in editing handles it well, but the less you have to de-ess, the more natural the voice sounds.

Breath and mouth noise

Breaths are part of natural speech and should usually be reduced, not deleted. Keeping the capsule out of the airflow reduces the harsh, gusty character of breaths. Mouth clicks come from a dry mouth more than the microphone; room-temperature water and avoiding dairy and sugary drinks before a session help more than any placement trick.

What should you do if you have to record in an untreated room?

Get closer, choose a less sensitive microphone, and add soft surfaces around the speaker, especially behind and to the sides of them. An untreated room is not a reason to give up on quality, but it is a strong reason to choose a dynamic over a condenser and to shorten the distance.

Hard, parallel surfaces produce flutter echo and a short, bright reverb that sounds like a bathroom or a kitchen. Small rooms add strong low-frequency resonances that make voices boomy. The goal is not a perfectly dead room but a reduced reflection level at the microphone. Our guide to room acoustics for recording covers treatment in depth; here are the quick, placement-driven fixes.

• Choose the room's quietest, softest corner of activity. Rooms with carpet, curtains, a bed or a sofa and filled bookshelves are better starting points than empty meeting rooms with glass walls.
• Keep the speaker away from walls. Sitting directly against a wall creates a strong early reflection. Move the setup a meter or so into the room if possible.
• Put absorption behind the speaker and in front of the microphone's rear. What the front of a cardioid sees, beyond the speaker, is what it hears most. A duvet or moving blanket hung behind the speaker is a remarkably effective temporary fix.
• Avoid the center of small rooms and the corners. Both are where low-frequency build-up tends to be worst.
• Turn off what you can. Air conditioning, fridges, fans and notification sounds are easier to switch off than to process out.

A portable reflection filter, the curved shield that mounts behind a microphone, helps somewhat, but it blocks sound arriving at the back of the microphone, which a cardioid already rejects. Absorption behind the speaker usually does more. Meeting rooms, which combine hard surfaces with several speakers and a distant ceiling microphone, deserve their own planning; installed conferencing systems are usually the better answer there.

How should you set gain so the loud moments never clip?

Set gain against the loudest thing the speaker will do, not their average level, and leave clear headroom: with 24-bit recording, many engineers aim for normal speech peaks somewhere around -18 to -12 dBFS and let laughs and emphasis peak well below 0 dBFS. Clipping happens when the signal exceeds the maximum level the converter can represent, and the flattened waveform it produces is recorded permanently.

Setting gain so high that loud moments clip is one of the most common and costly mistakes in spoken-word recording. Speech is dynamic. A presenter who reads at a steady level will still laugh, raise their voice to make a point or cough. If the gain was set on the quiet reading level with little headroom, those moments distort. Repair tools can reconstruct lightly clipped peaks, as explained in our guide to repairing clipped audio, but heavy clipping cannot be fully undone.

Recording at 24-bit gives you a large dynamic range, so leaving headroom costs almost nothing in noise terms. There is no benefit in recording hot. The noise that matters is the room and preamplifier noise, which are set by placement and equipment, not by how close your peaks get to 0 dBFS.

If your recorder or interface offers 32-bit float recording, the converter clipping risk at the recording stage is largely removed, but the analog preamplifier and the microphone itself can still overload, so sensible gain staging still matters.

Is a USB microphone good enough, or do you need an XLR microphone and interface?

A good USB microphone placed well is good enough for many podcasts, webinars and course recordings; an XLR microphone and interface is worth it when you need several microphones, better preamplifiers, upgrade flexibility or reliable long cable runs. The placement rules in this guide matter more than which connector the microphone uses.

One thing is not a trade-off at all. Recording from across the room on a built-in laptop microphone is almost never acceptable for published content. The laptop microphone sits beside a fan and a keyboard, is typically omnidirectional, and is usually 50 cm or more from the speaker's mouth. Even an inexpensive headset improves on it dramatically, because distance dominates everything else.

For remote contributors, the same principle holds. When you direct guests or voice talent in their own spaces, getting them to set up a decent microphone close and aimed correctly, with headphones on, matters more than any other instruction. Our guide to directing remote voice sessions covers how to coach that setup over a call.

How do you record several speakers in one room without bleed and phasing?

Give each speaker their own directional microphone placed close, aim each microphone's null toward the other speakers, and keep the distance between microphones clearly larger than the distance from each microphone to its own speaker. When several microphones hear the same voice at different times, their signals combine with small delays, which causes comb filtering, a hollow, phasey tone.

A widely used rule of thumb is the 3:1 guideline: place other microphones at least three times as far from a source as that source's own microphone. At 10 cm from host A, the microphone in front of host B should be at least 30 cm from host A, and further is better. The idea follows from the inverse square law: the further voice arrives much quieter in the wrong microphone, so when the signals are mixed, the delayed copy is too low to cause obvious comb filtering. It is a guideline, not a guarantee, and reflective rooms reduce its effectiveness.

Practical layouts for two to four speakers:

• Two hosts across a table. Seat them facing each other with cardioid microphones back to back in the middle; each microphone's rear then points at the other speaker. Dynamics work especially well here.
• Two hosts side by side. Angle the microphones outward slightly so each null faces the neighboring speaker, and increase the spacing between the chairs.
• Three or four around a table. Use one dynamic per person at a consistent close distance, place a soft cloth on the table to reduce reflections, and keep the microphones out of each other's front lobes.
• Panels and audiences. Once there is an audience and loudspeakers, the problem becomes sound reinforcement; see our guide to conference and venue sound reinforcement.

Always record each microphone to its own track. That lets the editor gate or duck microphones that are not being spoken into, which reduces bleed further, and it lets them fix a problem on one speaker without affecting the others. When two microphones must capture the same source, for example a close dynamic and an ambient condenser, check them for polarity and time alignment; the guide to getting phase and polarity right walks through the method.

How do you place microphones for video, where the mic cannot be in the shot?

Put the microphone as close as the frame allows, just outside the top edge on a boom, or on the speaker as a lavalier, and never rely on the camera's built-in microphone for dialogue. Video raises a constraint audio-only work does not have: the audience sees the setup.

Boom and shotgun placement

Mount the shotgun or a hypercardioid on a boom pole or stand, above the speaker and slightly in front, pointing down toward the mouth at around chest to chin height. Bring it down until it just appears at the top of the frame, then raise it slightly. Indoors, a hypercardioid often sounds more natural than a long shotgun, because long shotguns can color the reflections arriving off-axis. Pointing down toward the floor also means the microphone's sensitive axis sees carpet rather than a hard ceiling if the floor is soft.

Lavalier placement

Clip the lavalier on the sternum, about a hand span below the chin, centered so the voice stays consistent when the speaker turns their head. Hidden lavaliers need care: secure the capsule and the first few centimeters of cable with tape so clothing cannot rub against it, and avoid stiff synthetic fabrics that rustle. Test by having the speaker move their arms and turn in their chair while you listen.

Short-form and mobile video

For phone-shot content, a small wireless lavalier system or a compact on-phone microphone held close is a large step up from the phone's own microphone. Our guide to audio for short-form video covers setups for creators and social teams.

How do you keep a podcast or course series sounding consistent from session to session?

Document the setup and repeat it exactly: same microphone, same distance, same angle, same room position, same gain. Listeners notice when an episode sounds different even if they cannot say why, and editors spend a surprising amount of time matching sessions that were set up from memory.

The headphones matter as much as the microphone for this check. Closed-back headphones isolate you from the room so you hear what the microphone hears, rather than the room directly. Our guide to headphones and monitoring for editing covers what to use and how to listen critically.

If a session must move to a different room, accept that it will sound different, and minimize the change by keeping the distance identical and adding soft furnishings until the room tone and reflections are close to the original. Record a sample in the new room and compare it side by side with a previous episode before committing to a full session.

When should you bring in professional help with microphones and recording?

Bring in help when recordings consistently sound distant or echoing despite following the basics, when you are setting up a dedicated recording space, or when several speakers need to be recorded in one room. These are the points at which experience saves more money than it costs, because the mistakes are expensive to fix afterward.

A professional brings three things a guide cannot. First, they can hear a room and identify its problems in minutes, whether it is a flutter echo between two walls or a low-frequency mode making one seat boomy. Second, they have a range of microphones to try, which matters because voices are individual: a microphone that flatters one presenter can make another sound thin or harsh. Third, they can design a setup for multi-speaker recording that controls bleed and stays consistent, and document it so your team can reproduce it without them.

It also makes sense to involve help early in the life of a series. Changing the setup after 20 episodes is harder than getting it right before the first. Industry bodies such as the Audio Engineering Society publish educational material on microphone technique if you want to go deeper yourself, and tool makers such as iZotope publish practical placement guidance alongside their repair software, which is a useful reminder that even the people who make the repair tools recommend fixing it at the microphone first.

If you already have recordings that need rescuing, or you want a studio to plan and run the recording itself, our audio editing and production service covers setup advice, recording, cleanup and mastering for podcasts, courses and video.

Where this comes from

• Audio Engineering Society — Educational resources on microphone technique
• iZotope — Microphone placement guide

The figures and practices above come from the sources listed.

Working on something like this?

We take on Audio Editing & Production work for teams who want it done once, properly. Tell us what you are building and we will tell you honestly whether we are the right studio for it. Start a project.

Where to go next

• A Practical Guide to De-essing and Sibilance
• A Practical Guide to Room Tone and Ambience