Captions, subtitles, transcripts and audio description
Accessible video is not optional. If your media includes speech or important audio, people must be able to access that information in text. If your video contains important visual information, it must be described in audio or text.
This guide helps you decide what you need and how to do it well.
Understand the three core components
1. Captions (for video with audio)
Captions are synchronised on-screen text that:
- Show spoken dialogue
- Include important non-speech sounds (laughter, applause, music cues, tone changes)
- Appear at the same time as the audio
Captions are essential for:
- Deaf and hard-of-hearing users
- People in noisy or quiet environments
- People who process text better than audio
- Multilingual audiences
Closed captions can be turned on or off. Open captions are always visible and cannot be turned off.
If your video contains speech that is needed to understand the content, captions are required.
2. Transcripts
A transcript is a text version of the content.
There are two types:
Basic transcript
- Includes spoken words
- Includes important audio information
Descriptive transcript
- Includes spoken words
- Includes important audio information
- Includes essential visual information
Descriptive transcripts support:
- Deaf-blind users
- Screen reader users
- People who prefer reading
- Search engine indexing
For audio-only content (like a podcast), a transcript is required.
For video, best practice is to provide both captions and a transcript.
3. Audio description
If important information is shown visually but not spoken aloud, it must be described.
| Instead of saying: | “As you can see here…” |
| Say: | “The chart shows a 30% increase in sales between January and March.” |
If visuals are not explained in the main audio, an additional audio description track may be needed.
What does my media need?
Use this decision guide:
Audio only (podcast)
- Transcript required
- Captions optional
Video with no meaningful audio
- Descriptive transcript or audio description required
Video with speech
- Captions required
- Transcript strongly recommended
- Audio description if visuals are not explained
Live video
- Live captions required
- Edited captions for recordings
- Transcript recommended
Quality standards for captions
Captions must be:
| Synchronised | Text appears at the same time as the audio. |
| Accurate | Words match what is spoken. Do not rely on raw automatic captions. |
| Complete | Include meaningful non-speech sounds and speaker identification where needed. |
| Accessible | Users must be able to turn them on easily. |
Do not rely on automatic captions alone
Speech recognition tools are helpful starting points, but they are rarely accurate enough on their own.
Common problems:
- Incorrect wording
- Missing words
- Meaning changed
- Technical terms misheard
Automatic captions must be reviewed and corrected before publishing.
Caption file formats
For web video:
The most common format is:
WebVTT (.vtt)
Other formats:
- SRT
- TTML
Automatic captions must be reviewed and corrected before publishing.