Track 01 — Voice

Clear voice for your videos: a home recording guide

How to get a voice that sounds close, clean and easy to listen to, using an ordinary room and modest equipment. Most of the improvement comes from decisions you make before you press record.

8 min read Beginners Room to final mix

A woman wearing headphones speaks into a condenser microphone on a desk stand in a softly lit room at home.

Editorial note

Brisk Walnut Studio is an independent editorial site operated by Graphitti Sign & Graphic Studio. This guide is general information written by our editors. It is not sponsored, it contains no affiliate links, and we are not selling microphones, software, courses or services. Equipment is described by type, not by brand. How your recordings sound will depend on your own room, equipment, voice and practice, so treat every figure here as a starting point to test rather than a rule.

The guide

From a quiet room to a finished voice track

Viewers will tolerate a slightly soft picture far longer than a voice that sounds distant, boomy or buried in hiss. The good news is that clean speech is mostly a matter of habits, not budget.

Pick the quietest room, not the prettiest one

Before thinking about microphones, walk through your home and listen. Stand still in each room for a minute with your eyes closed and note what you hear: a refrigerator compressor, a heating vent, traffic, a neighbour’s television, the hum of a computer fan.

The best recording room is usually the one with the fewest noise sources and the most soft surfaces: a bedroom with a bed, curtains and a full wardrobe often beats a bare home office with a lovely view. Small, cluttered rooms tend to sound closer and drier than large, empty ones.

Timing matters as well. Early mornings and late evenings are often quieter, and it is worth switching off anything you control — air conditioning, fans, a ticking clock — for the length of each take, then switching it back on.

Soften the room with what you already own

An echoey voice comes from sound bouncing off hard, flat surfaces — bare walls, desks, windows and floors — and arriving at the microphone a fraction later than your direct voice. Soft, uneven materials absorb and scatter those reflections.

You do not need foam panels to make a noticeable difference. A heavy duvet or moving blanket hung behind the camera, facing you, catches the reflections bouncing straight back at the microphone. A rug on a hard floor, cushions on a sofa and closed curtains all help.

Two terms are often confused. Treatment controls how sound behaves inside the room; isolation stops sound getting in or out. Blankets help with the first and do very little for the second, so they will not silence a road outside.

Room check

Five minutes before you record

Listen for a full minuteNote every hum, tick and rumble before you decide where to sit.
Do a clap testOne sharp clap. A long ring means the room needs more soft surfaces.
Add soft layersBlanket behind the camera, rug underfoot, curtains closed.
Silence what you canFans, vents, notifications and the phone you are not filming on.

FIG. 01 · A quick room check usually reveals the biggest problem before any equipment is involved.

Four microphone types, in plain terms

Microphones for video voice work fall into a handful of broad families. Each is a trade-off between convenience, how much of the room it hears, and how it fits on camera.

Lavalier (lapel) microphones clip to clothing a short distance below the chin. Because they stay the same distance from your mouth as you move, levels remain steady. They can be wired to a recorder or phone, or used with a wireless transmitter. Their weak point is rustle from fabric and cables.

Shotgun microphones are long and directional, most sensitive to what they point at. On a stand or a boom just outside the frame they give a natural, open sound and keep clothing free, but they need to be close and they hear more room than a lavalier does.

USB microphones connect straight to a computer, which makes them a simple choice for narration and voiceover at a desk. They are less suited to moving around on camera.

Dynamic microphones are generally less sensitive than condenser types, so they pick up less of a lively room and less distant noise. The trade-off is that they need to be very close to the mouth and may need more gain from your recorder or interface.

There is no single correct choice. A sensible approach is to start with whatever gets a microphone closest to your mouth in the kind of video you make.

Placement and distance do most of the work

The single most effective change for most beginners is moving the microphone closer. The nearer it is to your mouth, the louder your voice is compared with the room and the background noise, so less of either ends up in the recording.

Lavalier: clip it at the centre of the chest roughly a hand’s width below the chin, pointing up toward the mouth. Make a small loop in the cable and tape it so tugs do not reach the capsule, and keep it away from necklaces, scarves and collars that move.

Shotgun: aim it at the mouth from slightly above or below, as close as the frame allows. A few centimetres of extra distance is audible, so creep in until it nearly enters the shot, then back off a little.

Desk microphone: a hand-span or so from the mouth is a common starting point. Speak slightly across the capsule rather than straight into it, which helps with popping on “p” and “b” sounds; a pop filter or foam windscreen helps too.

A man clips a small black lavalier microphone to the collar of his shirt, framed through the ring of a light. — **FIG. 02** · A lavalier clipped near the collar stays at a fixed distance from the mouth, which keeps levels steady as the speaker moves.

Gain staging and levels, without the jargon

Gain is how much your recorder or interface amplifies the microphone signal. Set it once, before recording, and then leave it alone.

Ask the speaker to read the loudest line they will deliver, including any laugh. Adjust the gain so normal speech peaks somewhere around −12 dBFS and the loudest moments stay below about −6.

If the meter hits 0 dBFS, the signal clips: the tops of the waveform are cut off, and that distortion cannot be cleanly repaired later. Recording a little quiet is the safer mistake.

Turn off automatic gain control where you can. It raises the background between sentences and makes the room seem to breathe in and out.

Some recorders offer a safety track at a lower level, or a 32-bit float mode that is far more tolerant of overloads. Both are useful, but neither replaces a sensible level check.

Record at 48 kHz, the usual sample rate for video, so audio and picture agree when they meet in the edit.

Listen on headphones while you record

A level meter tells you how loud something is, not what it sounds like. Closed-back headphones plugged into the recorder or camera let you hear what is actually being captured, including problems a meter will never show.

Listen for the things that are cheap to fix now and expensive later: a cable brushing against a shirt, a distant siren, a buzz from a nearby charger, or the slight hollowness that means the microphone has drifted too far away.

Closed-back designs leak less sound back into the microphone than open ones, which matters when you are recording yourself.

Protect your hearing. Keep headphone volume at a comfortable, moderate level and take breaks during long sessions. Turn the volume down before plugging in, because some devices start loud. If you notice ringing in your ears or discomfort, stop and rest.

Monitoring

What to listen for

Rustle and scrapeClothing or cable rubbing against the microphone.
Hum and buzzChargers, lights or a cable running beside a power lead.
Pops and breathBursts of air on “p” and “b” sounds.
Distance driftA voice that suddenly sounds thinner or roomier.

FIG. 03 · Four problems that are easy to hear on headphones and hard to see on a meter.

Room tone, and getting sound back in sync

Room tone is a recording of the room with nobody speaking: the same space, the same microphone, the same gain. Thirty seconds to a minute per location is a common habit.

In the edit, room tone fills the gaps where you cut out a cough or a stumble. Without it, the background drops to digital silence and the cut becomes obvious.

Silence in a room is never truly silent. Record what it sounds like, and your edits will stop drawing attention to themselves.

If your voice is recorded on a separate device, keep the camera’s own microphone running as a guide track. Then give one clear clap in view of the camera at the start of every take.

In the editor, line up the sharp spike of the clap in both waveforms, or let the software’s automatic waveform sync do it. On long takes, check the end as well as the start; some devices drift very slightly over time.

Name files and say the take out loud before you begin, so you can find the right recording later.

A gentle cleanup pass

Cleanup tools are best used lightly, in a sensible order, and always compared with the original. Heavy processing can make a voice sound smaller, harsher or strangely underwater.

Noise reduction. Teach the tool what the background sounds like using a stretch of room tone, then apply the smallest amount that helps. Steady hum and hiss respond well; sudden noises do not.
EQ (equalisation). A high-pass filter removes low rumble below the voice. Small, careful cuts can reduce boominess or harshness; big boosts usually add problems.
Compression. Narrows the gap between quiet and loud words so the voice sits evenly. Start with a gentle setting and listen for pumping or flattened energy.
De-essing. Optional. It softens sharp “s” sounds if they stand out.

Toggle each effect off and on. If you cannot hear what it is doing, you may not need it.

Voice chain · suggested order

Fix, shape, then even out

1. CleanLight noise reduction using a room-tone sample.
2. ShapeHigh-pass filter, then small corrective EQ cuts.
3. Even outGentle compression for a steady level.
4. FinishSet overall loudness and a peak limit last.

FIG. 04 · A common, cautious order of operations. Every voice and room is different, so trust your ears over any preset.

Loudness for online delivery, and common slips

Many video and streaming platforms adjust playback volume so that different uploads sound roughly equally loud. Loudness is measured in LUFS, and a frequently quoted target for spoken online video is around −14 LUFS integrated, with true peaks kept below about −1 dBTP. Platforms revise their practices, so check current guidance for wherever you publish.

More important than the exact number is consistency: the voice should stay at a similar level from the first sentence to the last, and music should sit clearly underneath it. Before publishing, listen on phone speakers, cheap earbuds and good headphones.

Recording from across the room and hoping software will fix the echo later.
Leaving automatic gain switched on, so the background swells between sentences.
Skipping headphones and discovering rustle or hum only in the edit.
Forgetting room tone, which makes every cut audible.
Over-processing with heavy noise reduction until the voice sounds artificial.

None of these require new equipment to fix. Small changes in the room, the distance and the routine usually make the biggest difference.

Session order

Five steps for every recording session

A short routine to follow each time you record. Each step heads off a problem that is harder to deal with once it is baked into the file.

ROOM · before setup

Choose and quiet the room

Listen for a minute, pick the softest and quietest space, and switch off fans, vents and notifications for the session.
ROOM · treatment

Add soft surfaces

A blanket behind the camera, a rug underfoot and curtains closed. Repeat the clap test and listen for a shorter ring.
MIC · placement

Get the microphone close

Lavalier below the chin, shotgun just out of frame, desk microphone about a hand-span away. Secure every cable.
LEVELS · test take

Set gain and listen

Read the loudest line, aim for peaks near −12 dBFS, disable auto gain, and keep headphones on at a moderate volume.
FINISH · edit

Room tone, sync, gentle cleanup

Record thirty seconds of room tone, clap to sync, then clean, shape and even out the voice with a light touch.

Quick reference

Small habits that keep a voice clean

Each one costs almost nothing to adopt and saves time in the edit.

REF 01 · DISTANCE

Closer beats costlier

Halving the distance between mouth and microphone usually does more for clarity than upgrading the microphone. Move it in first, then judge whether anything else needs to change.

REF 02 · HEADROOM

Leave space at the top

Normal speech peaking near −12 dBFS
Loudest moments below about −6 dBFS
Never touching 0 dBFS

REF 03 · PLOSIVES

Tame the pops

Use a pop filter or foam windscreen and speak slightly across the microphone rather than straight into it.

REF 04 · CLOTHING

Quiet fabrics

Soft cotton is kinder to a lavalier than stiff or synthetic fabrics. Tape the cable and avoid jewellery near the capsule.

REF 05 · INTERRUPTIONS

Silence the devices

Mute notifications on every phone, tablet and computer in the room, including the one you are recording with.

REF 06 · FILES

Name it, copy it, keep the original

Give each recording a clear name with the date and take, copy it to a second drive before you start editing, and keep an unprocessed version. If a cleanup step goes too far, you can always return to the untouched file.

REF 07 · ROOM TONE

Thirty quiet seconds

Record the empty room at every location. It is the patch for every gap you cut.

Questions

Common questions about home voice recording

Short, hedged answers. Where the right choice depends on your room, voice or equipment, we say so.

Do I need an expensive microphone to sound clear on video?

Usually not. Distance and the room tend to matter more than the price of the microphone. A modest microphone a short distance from your mouth, in a quiet and soft-furnished room, will often sound cleaner than a costly one placed across the room.

Fix the room and the placement first, then decide whether an upgrade is still needed.

Should I use a lavalier or a shotgun microphone for talking-head videos?

Both can work. A lavalier stays at the same distance from your mouth however you move and hides easily, but it can pick up clothing rustle. A shotgun on a stand or boom sounds more natural and keeps your outfit free, but it has to sit close, just outside the frame, and it hears more of the room.

In a lively, echoey room a lavalier is often the more forgiving choice.

Is a USB microphone good enough for a voiceover?

For narration recorded at a desk, a USB microphone is often perfectly adequate. It plugs straight into a computer and needs no separate interface. Its limits are flexibility rather than quality: it ties you to the computer, and it will not help if you appear on camera moving around.

Keep it close, use a pop filter, and switch off any automatic gain in the recording software.

What recording level should I aim for?

A common, cautious starting point is for normal speech to peak somewhere around −12 dBFS, with the loudest moments staying below about −6 dBFS. That leaves headroom for a laugh or a raised voice.

Recording a little quiet is generally easier to fix than recording so loud that the signal clips, because clipped audio cannot be restored cleanly.

How can I reduce echo without buying acoustic panels?

Use soft things you already own. Record in the most furnished room, hang a heavy blanket or duvet behind the camera, put a rug or cushions on hard floors, and keep open bookshelves or a full wardrobe nearby. Moving the microphone closer also reduces how much of the room it hears.

Soft furnishings mainly reduce echo; they do little to block outside noise.

Should I apply noise reduction to every recording?

Not automatically. Gentle noise reduction can tame a steady hum or hiss, but heavy settings tend to make a voice sound watery or metallic. Listen first, apply the lightest amount that helps, and compare with the effect switched off.

Removing the noise at the source, by turning off a fan or moving rooms, usually sounds better than any plug-in.

What loudness should I aim for when publishing online?

Many online platforms normalise playback loudness, and a frequently quoted target for spoken video is around −14 LUFS integrated with true peaks kept below about −1 dBTP. Platforms change their practices, so check the current guidance for wherever you publish.

Above all, keep the voice consistent from start to finish and check it on phone speakers and headphones.

How do I keep separately recorded audio in sync with the video?

Record the camera’s own microphone as a guide track, and make one sharp clap on camera at the start of every take. In the edit, line up the clap spike in both waveforms, or use the editor’s automatic waveform sync.

On long takes, check the end as well as the start, because some devices drift slightly over time.