SDH assumes you cannot hear the mix. When two mouths share a frame, or when the voice is off screen, the words alone are not enough. The ID is the prefix: a name, a role, a short tag in the same cue as the line. Closed captions can carry it as text; the file format does not have a special field for it in SubRip. Multi-speaker dialogue is the generate that casts distinct voices. Speaker ID is the caption of that cast, and it is still required when the picture already shows the face — because a mute autoplay viewer may be reading, not watching lips.
Auto captions transcribe speech. They will not reliably invent IDs, and they will collapse two talkers into one cue when the overlap is tight. That is a pass after transcription: split the cue, label the off-screen line, drop the ID once the speaker is visually obvious for a stretch. Karaoke and word-highlight still need the ID on the first token of a new speaker, or the accent walks across a line whose owner is unknown.
Do not confuse this with a lower-third overlay you authored. A lower-third is a graphic. A speaker ID is part of the caption track, timed to the line, and it belongs in SDH even when you also have a name plate.
In practice
- Add an ID when the speaker is off screen, new, or overlapping — not on every line of a single-host read.
- Split overlapping speech into two cues rather than stacking two names on one block.
- Keep the ID short. A long role title will blow caption line length on a vertical frame.
The mistake to avoid
Shipping a speech-only transcript of a two-hander and calling it captioned. Without IDs a mute viewer has dialogue with no cast.
Where you will run into it
- Add Captions to a Video — Speech in, styled subtitles out — no manual timing.
- Add Multi-Speaker Dialogue to a Video — A whole cast, one API call.
- AI Caption Generator — Drop a video, get timed captions. Free, on your device, no upload.
Related terms
SDH
SDH is a caption track that includes speech and non-speech audio — [door slams], [music], speaker IDs — so the file is usable without sound.
Closed captions
Closed captions are a separate subtitle track the player draws, which the viewer can switch off and which each platform styles itself.
Multi-speaker dialogue
Multi-speaker dialogue is one speech generation with a distinct voice per speaker alias, so a conversation is a single take instead of several stitched reads.
Hook rate
Hook rate meaning: the share of people shown a video who are still watching a few seconds in. It grades the opening, not the body.
Retention curve
Retention curve meaning: how many starters are still watching at each second. Watch time is the area under that curve.
The all-in-one AI studio for creators. 60+ models for video, image, voice, music and lipsync, in your browser or on your phone.
Free on the web and on iPhone, one account.