Skip to content
Faisal Mahmud TusherDigital marketing, SEO & analytics
with Faisal Tusher

Glossary term

Closed Captions

On-screen text that carries a video's speech and meaningful sound, needed for accessibility and useful to viewers who watch with the sound off.
StrategyconceptUpdated

Definition

Captions are synchronised text of what is said in a video, with the speaker identified where that is not obvious and important sounds described. They differ from a transcript, which is the full text of the audio or video read on its own, without the media playing.

The site's guides describe two ways of delivering them: burned into the video, so they appear wherever it plays, or uploaded as a caption file where the platform accepts one, which some viewers can restyle or switch off. The word closed usually refers to the second kind, captions a viewer can turn on and off; captions burned into the picture are always on screen. For accessibility the form matters less than the content: what matters is that the captions are there and right.

The Web Content Accessibility Guidelines, listed in this page's sources, are the reference. The guide to accessible brand identity lists captions for video among the requirements that fall directly on brand and content work, beside text alternatives for images and readable contrast. Which level of conformance a particular organisation must meet differs by jurisdiction and sector, and is a question for someone qualified to advise.

Captions and subtitles

The two words are often used loosely, and the site's localisation guide plans them together. Subtitles carry the spoken words into another language for viewers who can hear but do not understand the original. Captions carry speech and meaningful sound for viewers who cannot hear it or are watching without sound. A video localised for another market needs captions in that market's language too, so the guide to adapting creative for different markets plans captions in each language from the start, keeps speech and music on separate tracks so the master can be adapted, and has the market's reviewer check captions like any other copy, so that they fit the timing of the pictures.

In practice: accessibility, silent viewing, AI drafts and localisation

  • Accessibility. A brand's templates can carry a caption style and caption fields that cannot be skipped, so every video made from them is captioned without anyone having to remember.
  • Silent viewing. The guide to branding in performance ads notes that captions are recommended for silent viewing, which is also why the brand cue in the first frames should be visual rather than carried only by sound.
  • AI and automatic drafts. The guide to AI for alt text, captions and transcripts treats generated captions as drafts: they mishear product names, technical terms, proper names and numbers, miss speaker changes and leave sounds undescribed. Someone who knows the content reads them against the audio before the video is posted.
  • Social batches. The social media content hub puts image descriptions and captions into the production batch, written by whoever made the asset while they still know what it shows.
Illustrative example — not a real result

A plumbing firm corrects its automatic captions

Baurrott Plumbing, an invented plumbing firm, posts short how-to videos on fixing common leaks. Its first 12 videos went out with the platform's automatic captions. A customer who is hard of hearing writes to say that "compression olive" appeared as "compression olive oil" and "isolating valve" as "I slating valve" in 9 of them, and that the sound of the tap running, the cue to stop turning, was never captioned. The firm edits all 12 caption files, adds a line for the sounds that matter, and makes a person's read-through with the sound on a step before each new video is posted. Viewers watching on mute can now follow the steps in the order shown.

WCAG, the Web Content Accessibility Guidelines, is the standard the site's guides use when they ask for captions on video. Alt Text does for images what captions do for sound: it carries the information to someone who cannot see it, and both are drafted quickly by AI and checked by a person. And Transcreation, adapting copy so it works in another market rather than translating it word for word, applies to the supers and scripts around a video, while captions in each language follow the words the video actually speaks.

Common confusions

  • Captions are only for viewers who are deaf. They also serve everyone watching without sound.
  • Automatic captions are enough. They are a starting point; names, terms and numbers need a person's check.
  • Captions are the words only. Speaker labels and meaningful sounds are part of them.
  • A transcript replaces captions. A transcript is read apart from the video; captions are synchronised with it.
  • A post's caption is a closed caption. The text written under a social post is copy, not a caption of the video's sound.

More terms to look up

Where this term is used

Sources

  1. Web Content Accessibility Guidelines (WCAG) — W3C

Related