Closed captions present spoken dialogue plus relevant non-speech audio - sound effects, speaker IDs, music cues - as on-screen text that viewers can turn on or off. Unlike translation subtitles, captions (and SDH, subtitles for the deaf and hard of hearing) are designed for accessibility and convey more than just the words.
Technically they ride as a selectable text track (WebVTT, TTML/IMSC) referenced from the manifest and rendered by the player with styling controls. Captions are legally required for much content in many markets, so correct authoring, carriage, and rendering across every device is a compliance obligation - and, given how many viewers use captions by choice, a mainstream usability feature too.
