Microphone technique for voice-over

• Pick a working distance, commonly 10 to 20 cm (4 to 8 inches), and keep it for the whole session.
• Aim your voice slightly past the mic, or angle the mic, so bursts of air from p and b don't hit the capsule head-on.
• Directional mics boost bass as you get closer (the proximity effect). Use it on purpose, not by accident.
• Reduce sibilance with angle, distance and placement before you reach for a de-esser.
• Set gain so your loudest lines stay well under 0 dBFS. A slightly quiet recording is fixable; clipping isn't.
• The room matters as much as the mic: a treated corner or a closet of clothes beats an expensive mic in a bare room.

A microphone captures everything you do: every change in distance, every head turn, every burst of air and every reflection from the room. A consistent performer on a modest USB mic often sounds better than a careless one on an expensive condenser. Editors care about consistency because they may need to match pickups recorded days later.

Most voice-over is recorded on a large-diaphragm condenser or a dynamic mic with a cardioid pickup pattern, which picks up mainly from the front and rejects sound from behind. Short shotgun mics are also common in voice-over and ADR booths. Whatever you use, start by finding out which part is the front.

Many large-diaphragm condensers are side-address, meaning you speak into the side, usually the one with the logo, not into the top. Speaking into the top of one is a common beginner mistake that makes the voice sound thin and distant.

A common starting distance for voice-over on a cardioid condenser is 10 to 20 cm (about 4 to 8 inches) from mouth to mic. A spread hand, thumb to little finger, is a quick way to measure. Closer gives a more intimate, bass-heavy sound with more mouth noise and plosives. Farther gives more room sound, a thinner tone and a lower level.

Small movements matter more the closer you work. In open space, sound level drops by about 6 dB each time you double your distance from the source. Real rooms complicate that, but the principle holds: moving from 10 cm to 20 cm is a big change, while moving from 30 cm to 40 cm barely registers. So the closer your working distance, the more still you need to be.

Make consistency easy. Put the script at eye level on a stand or tablet so your head doesn't drop as you read. Use the pop filter as a reference point and keep your lips a fixed distance from it. For quiet, intimate narration you might work closer; for shouts and big character moments, back off a few centimetres or turn slightly away on the loudest words. Whatever you choose, keep it consistent within a project.

Hold your palm about 10 cm in front of your mouth and say "Peter picked a pack of pickled peppers." You'll feel bursts of air on the p's. Move your palm a few centimetres to the side and say it again: most of the bursts miss. That's the principle behind angling the mic.

Instead of pointing the mic straight at your mouth, place it slightly to one side, or a little above or below mouth level, and angle it toward your mouth. Your voice still reaches the capsule, but the air stream travels past it. Many voice-over performers set the mic around nose height and angle it down toward the mouth, or position it just off to one side.

Off-axis placement also tames a little sibilance and mouth noise, because high frequencies from the mouth are more directional than low ones. Don't overdo it, though. Too far off-axis and a cardioid mic sounds duller and quieter, and you lose the clarity you were after. Small angles do most of the work.

Plosives are consonants made by stopping the air and releasing it: p and b most of all, and t, d, k and g to a lesser degree. When that burst hits the diaphragm, you get a low thump that's distracting and tedious to fix. A pop filter, a nylon or metal mesh screen placed a few centimetres in front of the mic, breaks up the burst before it arrives. Foam windscreens do a similar job and are common on dynamic mics.

Use technique and gear together: the pop filter, a slight angle and a sensible distance. You can also soften the hardest p's slightly without losing the consonant. Monitor on headphones while you record so you hear pops as they happen and can retake the line.

In editing, a high-pass filter, often set somewhere around 70 to 100 Hz for speech, removes rumble and some of the weight of a pop, and individual pops can be repaired with editing tools. But fixing them one at a time is slow. It's far quicker to prevent them.

Sibilance is the harsh, hissy edge on s, sh, z and ch sounds. It gets worse with a bright-sounding mic, speaking straight on-axis, working very close and leaning hard on your s's. Fix it in that order: adjust the angle, try the mic a little higher or lower, back off slightly, and lighten your s's a touch. Only then reach for a de-esser in editing, which turns down just the sibilant range when it spikes. Too much de-essing makes you sound as if you have a lisp.

Mouth noise, the clicks and smacks of a dry mouth, is the other close-mic problem. Drink water steadily in the hours before a session, not just a gulp before you start, and keep room-temperature water nearby. Many voice actors swear by a bite of green apple to cut mouth clicks; it's folk advice, but it's cheap to try. Part your lips slightly before the first word so you don't start with a smack.

Directional microphones, including cardioid, supercardioid and figure-8 models, boost low frequencies as the source gets closer. This is the proximity effect. Omnidirectional mics show little or none of it. Broadcasters use it deliberately for a warm, close, late-night radio sound, and trailer voices often lean in for weight.

Used by accident, it causes problems. A narrator who leans in and out sounds as if someone is turning a bass knob up and down. Working very close for a whole audiobook can make the voice muddy and boomy. If your recordings sound thick, back off a few centimetres, use the mic's low-cut switch if it has one, or apply a gentle high-pass filter.

Set your input gain with the loudest line in the script, not a polite "testing, testing". A widely used rule of thumb is for normal speech to peak somewhere around -12 to -6 dBFS, leaving headroom for louder moments. Never hit 0 dBFS: digital clipping distorts the recording and can't be undone, while a recording that's a little quiet can be raised cleanly later.

Don't adjust the gain during a take. Keep it fixed and manage dynamics with distance and performance, then use compression in editing if needed. Riding the gain mid-read makes the recording inconsistent in ways that are hard to repair.

Check the noise floor too. Record a few seconds of silence, known as room tone, at the start of every session and listen to it on headphones with the volume up. Computer fans, fridges, traffic and air conditioning show up clearly. Delivery specs vary by client. ACX, for example, asks that audiobook files peak no higher than -3 dB, measure between -23 and -18 dB RMS, and have a noise floor no higher than -60 dB RMS.

A microphone can't tell the difference between your voice and your voice bouncing off a bare wall. Hard, parallel surfaces create a hollow, echoey sound that's very hard to remove later. Soft furnishings help: a closet full of clothes, a corner with heavy blankets or duvets behind and beside you, or acoustic panels if you have them.

Working closer to the mic improves the ratio of voice to room, which is one reason home voice-over talent tend to work at 10 to 20 cm. But it trades off against the proximity effect and mouth noise, so treat the room first and use distance second. Run a quick mic test in a few spots around your home and compare.

1. Quiet the room. Turn off fans and notifications, close windows and doors, and move away from the computer if it's noisy.
2. Position the mic. Make sure you're addressing the front of the mic, set it about a hand-span away, angle it slightly off your mouth and put the pop filter in place.
3. Set gain with your loudest line. Perform the loudest moment in the script and adjust the gain until it peaks comfortably below clipping, with normal speech around -12 to -6 dBFS.
4. Record room tone and a test read. Capture five to ten seconds of silence and a short read with the mic test or the voice recorder.
5. Listen on headphones. Check for pops, sibilance, room echo, mouth noise and background hum, then adjust the angle, distance or room.
6. Mark your position. Note your distance and keep it for the whole session so every take matches.