Captions and Transcript
Captions are text timed with speech and useful sounds. A transcript is a readable text version of the whole audio or video.
Another example
A product tutorial labels each speaker and the sound “notification chime.” Its transcript also has headings and the commands shown in the demo. An editor fixes any model number that automatic captions hear incorrectly.
Main parts
- 01Timed cues
- 02Speaker and sound labels
- 03Structured transcript
Use it when
Give prerecorded speech accurate, timed captions. Add a clear transcript when people need to scan, search, or use deafblind access.
Do not use it when
Do not publish unchecked automatic captions when names, numbers, safety steps, or technical terms matter.
Name used in code
video track kind=captions with WebVTT + adjacent structured transcript Before you ship
Before release, check the page structure, keyboard access, visible focus, contrast, and zoom. Also check loading, empty, error, success, cancel, and reduced-motion states.
Outcome: Use Captions and Transcript to solve this user need without losing the current context. User context: A product tutorial labels each speaker and the sound “notification chime.” Its transcript also has headings and the commands shown in the demo. An editor fixes any model number that automatic captions hear incorrectly. Recommended pattern: Captions and Transcript. Why it fits: Give prerecorded speech accurate, timed captions. Add a clear transcript when people need to scan, search, or use deafblind access. Do not use it when: Do not publish unchecked automatic captions when names, numbers, safety steps, or technical terms matter. Product requirements: Define the default, loading, empty, error, success, and recovery states. Keep user input. Give clear feedback after every action. Accessibility and resilience: Use semantic HTML, full keyboard access, and visible focus. Make the layout work when zoomed and support reduced motion. Follow the project’s current rules. Do not add a new UI library unless the project already uses it. Acceptance criteria: The main task works with keyboard only and at 320px width. Focus and status changes are announced. Check loading, failure, retry, cancel, and reduced-motion paths. Evidence and limits (evidence boundary): No single official standard controls the whole term. Its meaning may change across teams, platforms, or frameworks. Unknowns to confirm: Product wording, data source, analytics events, supported devices and languages, and platform limits.
Check this request
Review the current use of Captions and Transcript. Definition: Captions are text timed with speech and useful sounds. A transcript is a readable text version of the whole audio or video. Release checks: Before release, check the page structure, keyboard access, visible focus, contrast, and zoom. Also check loading, empty, error, success, cancel, and reduced-motion states. Before changing code, report the evidence you found, gaps, severity, and the smallest safe fix.