Whether you are narrating a tutorial, a video essay, or a product review, clean voiceover is the difference between content that sounds professional and content that sounds like it was recorded in a bathroom. You do not need a commercial studio. With the right microphone technique, a little room treatment, and sensible levels, you can capture broadcast-adjacent narration at home. This guide covers the room, the gear, the technique, and the settings.
The room matters more than the mic
Beginners obsess over microphones, but the room does more damage to voiceover quality than any mic upgrade can fix. Hard parallel surfaces create reflections that muddy your voice with echo and boxiness. Before spending on gear, deaden your recording space: record in a room with soft furnishings, hang a blanket behind and around you, or build a simple fort of pillows. Even recording inside a closet full of clothes dramatically reduces reflections.
Microphone choice and placement
A dynamic microphone rejects room noise better than a sensitive condenser, which makes it forgiving in untreated spaces. Position the mic 15–20 cm from your mouth, slightly off-axis so plosives do not punch the capsule, and use a pop filter. Being close to the mic raises your voice above the room’s noise floor, which is one of the simplest ways to sound cleaner.
| Setting/parameter | Target | Why |
|---|---|---|
| Mic distance | 15–20 cm | Strong signal, less room |
| Sample rate | 48 kHz | Video standard |
| Bit depth | 24-bit | Headroom for editing |
| Peak level | -12 to -6 dBFS | Loud but no clipping |
| Final loudness | -16 LUFS (web) | Consistent playback |
Getting levels right
Set your gain so normal speaking peaks land between -12 and -6 dBFS, leaving headroom so a sudden loud word does not clip. Clipping is unrecoverable, so it is always safer to record a touch quieter and raise the level later. Record at 48 kHz and 24-bit to match video standards and give yourself editing headroom. Do a test read and listen back on headphones before committing to the full session.
Performance beats processing
The best-sounding narration usually comes from a good performance, not heavy plugins. Warm up your voice, drink room-temperature water, and stand or sit up straight for better breath support. Read slightly slower than feels natural, because narration that feels a touch slow in the booth usually sounds right on playback. Leave a beat of silence between paragraphs so you have clean edit points.
Light processing chain
After recording, a simple chain cleans things up: a high-pass filter around 80 Hz to cut rumble, gentle compression at a 3:1 ratio to even out dynamics, a light de-esser if your S sounds are harsh, and a final loudness normalization to about -16 LUFS for web playback. Resist the urge to stack effects; over-processed voiceover sounds artificial and fatiguing.
FAQ
Do I need an expensive microphone for good voiceover?
No. Room treatment and mic technique matter far more. A modest dynamic microphone in a blanket-deadened room will outperform an expensive condenser in a bare, echoey room every time. Spend on treatment first, then upgrade the mic.
What loudness should I export my voiceover at?
For web and video platforms, normalize to around -16 LUFS integrated. That level is loud enough to compete with other content without clipping and keeps your narration consistent across different viewer devices and volume settings.
Editing your reads efficiently
Clean narration is as much about editing as capture. Record in takes and, when you flub a line, pause, leave a beat of silence, and simply re-read the sentence rather than stopping the recorder; the silence gives you an obvious visual marker to cut on later. Read a little more slowly and deliberately than feels natural, because narration that sounds right on playback almost always felt slightly slow in the booth. Keep a glass of room-temperature water nearby and take small sips to manage mouth clicks, which are tedious to remove one by one. When you assemble the final track, breathe life into it by keeping natural pauses rather than deleting every gap, since over-tightened narration sounds robotic. A relaxed, well-paced performance beats surgical editing every time.
Bottom line
Deaden your room, get close to a dynamic mic with a pop filter, and record at 48 kHz/24-bit with peaks at -12 to -6 dBFS. Focus on a relaxed, slightly slow performance, then apply a light processing chain and normalize to -16 LUFS. That workflow delivers clean, professional narration without a studio.
