Subtitles
Subtitles are burned into the picture, because most feeds play without sound and most players ignore a separate subtitle file. What you see on the stage is what is rendered.
The styles
| Style | What it looks like |
|---|---|
| Small white text | Light white lettering with a thin outline. Stays out of the way. |
| Bold white text | Medium bold white lettering with a black outline, up to two lines. |
| Big white capitals | Large uppercase lettering with a heavy outline. Loud on a phone screen. |
| White on a black bar | Bold white lettering on a shaded bar, up to three lines of speech. |
| Gold word highlight | The word being spoken picks up colour as the line plays. |
| None | No subtitles at all. |
The style is chosen per clip, so one recording can produce a loud vertical clip and a quiet wide one.
Where they sit
Top, middle or bottom of the frame. Bottom suits most things; the middle is worth trying on vertical clips, where the lower third often ends up under a caption or a profile name.
Subtitles are sized for the frame they are rendered into, so the same style stays readable whether the clip goes out vertical or wide. You do not have to adjust anything after changing the shape.
Getting the words right
Names, places and technical terms are what a transcript gets wrong. Fix them in the transcript panel - find the line, correct the word - and the next export carries the correction.
Correcting the transcript is free, and so is exporting again afterwards.
Captions while you work
Separately from all of the above, the line currently being spoken can be shown over the video while you work. That one is a working aid, not part of the clip: drag it somewhere it does not cover the speaker's face, change its size, or switch it off in Settings. It is never rendered.