A voice-over sits on top of everything else in a video: screen recordings, product demos, B-roll, explainers, slides. Because nothing visual competes with it, every flaw in the recording is exposed, from a laptop fan to a hollow room sound to a level that jumps between takes.
Here is a step-by-step guide to getting a voice-over that sounds clean and consistent, including the cleanup after recording.
Step 1: Set up for the voice, not the camera
Voice-overs have one big advantage over on-camera audio: you can put the microphone exactly where it should be, because nobody sees it.
- Microphone a hand's width or two from your mouth, slightly off-axis so breath doesn't hit it directly. Use a pop filter.
- Soft surroundings. Recording facing a wardrobe of clothes, a bookshelf or a duvet hung behind the microphone cuts the room sound dramatically. Room echo is the one thing post-production can't remove.
- Quiet the machine. If you record into a laptop, close heavy apps so the fan stays quiet, and keep it as far from the microphone as the cable allows.
- Headphones on, so you can hear what the microphone hears.
Step 2: Record in a consistent way
Consistency is what makes a voice-over sound professional, more than any single setting.
- Same distance and angle for every take. Mark the spot or keep your elbow on the desk.
- Same time of day and room for a series of videos; your voice and the background change through the day.
- Leave three seconds of silence at the start of each session.
- Read in sections, re-reading a sentence immediately when you stumble rather than starting over. You can cut the mistakes later.
- Watch the level. Leave headroom: loud words well below the top of the meter. Clipping can't be cleaned away.
Step 3: Clean the recording before you edit
Remove the background from the raw recording first. Then every cut you make is on clean audio, and you won't hear the background switch on and off at each edit.
- Open the noise remover for audio or the tool and drop in the recording (WAV, MP3, M4A/AAC, OGG/Opus or FLAC).
- Turn on Voice clarity for crisper speech: it trims rumble and boxiness and adds presence, which suits voice-over well.
- Choose Loudness: YouTube (−14 LUFS) for online video, or Off if you'll set the final level in your video editor.
- Choose WAV as the output if you'll keep editing, so the file isn't compressed twice.
- Preview original against cleaned at the same moment, then download.
The cleaned WAV keeps exact timing, so it lines up with your timeline. Our speech AI removes the sound around the voice, steady or changing, without the chopped word endings of a noise gate (see noise reduction vs noise gate).
Step 4: Edit for flow
- Cut stumbles and long pauses, but leave natural breathing room; a voice-over with every breath removed sounds robotic.
- Use short crossfades at each cut to avoid clicks.
- Keep breaths quieter rather than deleting them all. Turning loud breaths down a little sounds more natural than silence.
- Check the pacing against the picture. Slide sections of narration to land with the visuals.
Step 5: Mix with music
If your video has background music:
- Add it after cleaning the voice, never before; a speech-focused cleaner would reduce music mixed into the voice track.
- Keep music well below the voice while you speak. Many editors can lower the music automatically under narration ("ducking").
- Listen on headphones and on a phone speaker; music that sounds fine on headphones can bury a voice on a small speaker.
Step 6: Final level
If you didn't set loudness in Step 3, set it on the final mix. Online video commonly targets about −14 LUFS; spoken-word podcasts about −16 LUFS. Whichever you choose, use the same target for every video in a series so viewers never reach for the volume between episodes.
If your voice-over is already part of a finished video, you can clean the whole video instead: the video noise remover replaces only the audio track and copies the picture untouched. Note that this also reduces any music in the mix.
Troubleshooting
| Symptom | Likely cause | Fix |
|---|---|---|
| Hollow, distant sound | Room echo, microphone too far | Closer microphone, softer surroundings; not fixable by noise removal |
| Constant whoosh | Laptop or room fan | Clean it; quiet the machine next time |
| Popping on "p" and "b" | Breath hitting the microphone | Pop filter, angle the microphone slightly |
| Harsh crackle on loud words | Clipping | Lower the gain and re-record those lines |
| Level changes between sessions | Different distance or room | Mark your position; set loudness on every file |
| Thin, tinny voice | Built-in laptop microphone | Use an external microphone; Voice clarity helps a little |
A checklist for every session
- Microphone placed and marked, pop filter on.
- Soft surface behind and around you.
- Fans off, apps closed, phone silent.
- Headphones on, test take checked.
- Three seconds of silence recorded.
- Clean first, edit second, music third, loudness last.
For more on making the recording itself better, see how to record clean audio at home. If you are making YouTube videos, clean audio for YouTube has the matching preset.