Glossary
What is a talking head?
By the KeyMo team · Updated
A talking head is footage of one person speaking straight to the camera, usually framed from the chest up, with little else happening on screen. Interviews, vlogs, lectures and most video podcasts are built from talking-head shots, which is why they cut so well into vertical clips.
Talking head meaning in film, documentary and scripts
In a documentary, a talking head is the interview subject on camera explaining or reacting, often cut against archive or cutaway footage. On a call sheet or in a script, the label marks a stretch where someone simply speaks to the lens. In news, the phrase is sometimes used as a mild put-down for panel shows where people only talk.
What unites the uses is the picture: a face, a voice and not much else. The words carry everything, which is the strength and the weakness of the format.
Why podcasts and interviews count
A video podcast is mostly talking heads, switched between a wide shot and singles. That makes it A-roll in editing terms, and it explains why the format is easy to clip. The speaker is already the subject, the audio is clean enough to follow, and the face gives a viewer something to look at while the sentence lands.
Framing and caption rules that follow
| Shot | Framing | Use it for |
|---|---|---|
| Close-up | Head and top of shoulders | Emotional or emphatic lines |
| Medium | Chest up, a little headroom | Most answers and explanations |
| Two-shot | Both speakers in frame | Quick back-and-forth |
- Keep the eyes about a third of the way down and leave a little headroom.
- Put captions below the chin or in the lower middle, never across the mouth or eyes. US broadcast captioning rules say captions must not block faces, and the habit carries over to short-form video.
- In a wide two-person shot, crop to whoever is speaking instead of one fixed slice.
Common mistakes
- Letting a head sit dead centre for ten silent seconds. Trim pauses before they read as dead air.
- Starting on the host's question when the guest's answer is the draw.
- Using a distant shot on a phone screen, where faces become small and expression is lost.
Where KeyMo touches it
Talking-head footage is what KeyMo is built around. It reframes a landscape recording to 9:16 by tracking the face, trims silences and adds word-by-word captions in one of 12 styles. The output is vertical 1080x1920, never 16:9. See face tracking for how that reframe works.
Sources: Electronic Code of Federal Regulations, 47 CFR 79.1 (closed captioning of video programming) (checked 11 October 2026).