Guide
In a voiceover video the voice persuades while the pictures illustrate
A voiceover video is a video where a recorded voice carries the message while the screen shows something other than the speaker. Footage, stills, screen recordings or a generated presenter can hold the frame. The voice does the persuading and the pictures do the illustrating. This page covers the trade against showing your face, and the audio rules that do not relax.
- Jobs split apart: the voice sells, the frame shows
- 2
- Roughly what doubling mic distance costs
- 6 dB
- Credits a minute for a generated voice-over here
- 70
What you trade against a talking head
A talking head buys trust. A visible person appearing to mean what they say is the strongest signal a short video can send, and a voiceover gives it up.
What you get back is freedom of frame. The screen can show the product, the process or the numbers for the full runtime, which a face cannot.
The formats sort themselves by that trade. Tutorials, screen walkthroughs and process videos lean voiceover. Ads, testimony and anything selling a person lean face.
Voiceover against talking head
Neither wins in general. The content decides which job matters more for this video.
| Voiceover | Talking head | |
|---|---|---|
| Trust signal | The voice alone | A visible person |
| The frame | Free for anything | Mostly the face |
| Performance needed | Reading well | Being watchable |
| Natural home | Tutorials, demos | Ads, testimony |
The audio rules do not relax because the face is gone
A voiceover video is all audio, so audio faults are the whole experience. A muddy read over beautiful footage is a muddy video.
The fixes are recording fixes, not mixing fixes. Get the microphone close, because distance costs roughly six decibels per doubling. Record in a soft room, because reflections cannot be removed later.
Music sits under the voice, never beside it. A bed a listener starts noticing during a sentence is a bed that is too loud, and the fix is always turning it down rather than turning the voice up.
This is the whole editor
Highlight a phrase and a clip lands on those exact words. No timeline, no keyframes, no layers.
Where the voice comes from, and one honest limit
Most voiceovers are a person reading into a phone or a USB microphone, and that is fine. Warmth and pacing beat studio gloss for feed content.
Generated voices are now a real option for informational reads. Cutroom generates a voice-over at 70 credits a minute, which it uses to drive avatar presenters.
The limit worth knowing before you plan around it: a Cutroom project starts from a filmed take, so a video that is only audio and stock footage is the wrong fit here and belongs in a conventional editor.
Questions people ask
- Is a voiceover video cheaper to make than a talking head?
- Usually, because nobody has to look presentable and re-reads are painless. The saving is in the recording. The editing effort is similar, and the b-roll burden is heavier because the frame is always footage.
- Do voiceover videos perform worse in feeds?
- Not as a rule, but they compete on different strengths. Without a face the first seconds must earn attention with the picture and the on-screen text, and a large share of viewers watch muted.
- Should the voiceover be scripted or improvised?
- Scripted, lightly. A written script read too stiffly sounds like a script. Write it, then deliver it as if explaining to one person, and keep the sentences short enough to say in one breath.
Split the jobs cleanly: let the voice carry the argument and make every frame earn its place under it. The audio is the video, so record it close.