You film yourself talking to camera, add captions, and something looks off. The text sits on your chin, or it crowds the username at the bottom, or it flashes past before anyone could read it. A talking head video has one subject that viewers keep coming back to, your face, and the captions have to share the frame with it.
The short answer. Put captions in the band between your chin and the platform’s interface, usually just below the middle of a 9:16 frame, and film with enough headroom to leave that band free. Keep the caption block smaller than your face and never over your mouth. Break captions at the pauses you actually make, and clear them when you stop talking.
Why talking head captions are different
In most short videos the eye wanders: a product, someone’s hands, the background. On a talking head video it barely moves. People look at the eyes and the mouth, and they return there between lines of text. Expression carries a lot of the meaning. A raised eyebrow or a half smile tells the viewer a line was a joke before the words do.
That changes the captions’ job. They need to sit close enough to the face that reading them isn’t a long trip down the screen, and far enough away that they never cover it. On a muted feed, a caption over the mouth also hides the one cue that helps a viewer when a word is wrong or missing.
Where to put the captions
Start with the framing
Placement starts before you add a single caption. A vertical talking head shot usually puts the eyes somewhere around the upper third of the frame, with the chin a little above the middle. That leaves the lower half for shoulders and chest, which is exactly where captions fit.
The other half of the problem is the platform. At the time of writing, TikTok, Reels and Shorts all lay the username, the post text and an audio row over the bottom part of the frame, with a column of buttons down the right edge. Captions in either zone get covered.
So the usable space is a band: below your chin, above the interface, and kept in from the right edge. On a 1080×1920 frame that’s roughly the area just under the middle. Frame tight, with your face filling most of the screen, and that band disappears. Step back or zoom out a touch when you film, so there’s some chest showing under your chin.
When the face is too low or too big
If the footage is already shot and your chin sits near the bottom third, you have two options. Put the captions above your head if there’s headroom, or crop in and reframe so your face moves up. Captions above the head work, but the eye travels further from mouth to text, so they suit one short line better than a two-line block.
Avoid laying text across your forehead or eyes. It’s the usual fallback when there’s no room, and it’s the one that hurts the video most.
Check the punch-ins
Plenty of talking head edits zoom in on alternate cuts, scaling the shot up to hide a jump cut. Each punch-in makes your face bigger and pushes your chin down the frame. A caption that cleared your chin in the wide shot can land right on it in the tight one. Scrub through the whole edit and stop on the tightest framing before you export.
In Capster, position on a nine-point grid is on the free tier, and the preview uses the same renderer as the export, so a caption that clears your chin on a punched-in frame in the preview clears it in the file. The guide to add captions to a video on iPhone walks through where that control sits.
How big to make them
With one person on screen, the captions are the second most important thing in the frame, and their size should reflect that. A simple test: the caption block should be clearly smaller than your face. Two lines of text as wide as your shoulders and as tall as your head turn the video into a slide with someone standing behind it.
Weight helps more than size here. A heavy font with an outline reads at a smaller size, which leaves more of the band under your chin free. If you want numbers, the post on caption font size for vertical video gives starting cap heights by style. For talking heads, start at the low end of each range and only go up if a test on your phone at arm’s length says so.
Keep lines short enough that they stop before the button column on the right. A centred line a little narrower than your shoulders usually clears it.
Pacing captions for one speaker
One speaker makes the transcript simpler. There are no speaker changes to mark and no voices talking over each other. What’s left is rhythm.
Break where you breathe
People talking to camera speak in bursts: a phrase, a pause, another phrase. Captions read best when they break in the same places. A caption that stops mid-phrase and finishes on the next card makes the viewer hold half a thought. If the auto captions split “I stopped doing this / every morning” across two cards when you said it in one breath, merge them.
Don’t carry a caption across a cut
Jump cuts are normal in this format, and each one is a small jolt. A caption that carries over the cut adds a second one: new framing, same old text. Where you can, end a caption on the last word before a cut and start a fresh one after it.
Clear the screen when you stop
If you pause for a beat, for a reaction or a held look at the camera, let the caption go. Text left hanging over silence looks like a timing mistake, and it pulls the eye down when you want it on your face.
Match the style to your pace
Fast talkers with lots of cuts suit one word at a time. Each word lands on the beat and the frame stays mostly clear. Slower explainers suit one or two lines with the spoken word highlighted, because the viewer can read slightly ahead and then look back up at you. The animated caption styles page lists which presets do which, including Capster’s One Word preset for the fast end.
A quick check before you post
| Check | What to look for |
|---|---|
| Mouth and eyes clear | No caption over them on any frame, punch-ins included |
| Inside the band | Below your chin, above the platform’s bottom text |
| Clear of the right edge | The longest line stops before the button column |
| Smaller than your face | Text the width of your shoulders competes with you |
| Breaks match your pauses | No half-phrases carried onto the next card |
| Nothing across cuts | Each cut starts a fresh caption |
| Pauses are empty | Captions clear during held silences |
Placement details shift a little from one platform to the next. If most of your talking head clips end up on YouTube, the page on subtitles for YouTube Shorts covers where its player puts the title, channel row and buttons.