Audio description (AD), also called described video or video description, is an additional narration track that describes the essential visual content of a video, such as actions, scene changes, on-screen text and facial expressions, for people who are blind or have low vision. In video accessibility, it is the counterpart of captions: captions make sound available to people who cannot hear it, while audio description makes images available to people who cannot see them.
The narration is placed into natural pauses in dialogue and important sound, so the viewer hears both the original soundtrack and the description. Accessibility guidelines such as the Web Content Accessibility Guidelines (WCAG) and broadcasting regulations in many countries treat it as a core requirement for accessible video.
How audio description is produced
Audio description is created in a structured production process that resembles dubbing more than captioning. The work starts with a careful viewing of the program to find gaps in the soundtrack and to decide which visual information is necessary to follow the story or message.
- Scripting: a trained describer writes short descriptions timed to fit each gap, usually in the present tense and third person.
- Review: an editor, ideally including a blind or low-vision reviewer, checks accuracy, clarity and timing.
- Voicing: a narrator records the script, or a text-to-speech voice renders it for faster production.
- Mixing: the description is mixed with the program audio, which is lowered slightly during narration so both remain intelligible.
- Delivery: the result is published as a separate audio track, a premixed version or a timed text file.
The central skill is selection under time pressure. Gaps may last only a few seconds, so the describer must prioritize what is essential, such as who is present, what changes and what text appears, and leave out what the audience can infer from the sound.
Types of audio description
Audio description varies in how it fits into the program timeline and in how it is delivered. The main distinction is whether the video plays continuously or pauses to make room for narration.
| Type | How it works | Typical use |
|---|---|---|
| Standard description | Narration fits into existing pauses | Films, television series, broadcast |
| Extended description | Video pauses while narration plays | Training, lectures, dense information videos |
| Integrated description | Visual information written into the main script or voice-over | Corporate and educational video planned for accessibility |
| Text-based description | Timed text read by a screen reader or synthetic voice | Web players supporting description tracks |
Integrated description is often the most efficient approach for new content. When the presenter says what is shown, such as reading the key figure on a chart aloud, no separate track is needed, and the video becomes accessible by design.
How audio description relates to captions and subtitles
Audio description and captions address different sensory gaps, and a fully accessible video usually needs both. Their relationship to the other Terms in video accessibility is summarized below.
| Feature | Audience served | What it conveys | Format |
|---|---|---|---|
| Audio description | Blind and low-vision viewers | Essential visual information | Audio track or timed text |
| Closed captions | Deaf and hard-of-hearing viewers | Dialogue and relevant sounds | Text the viewer can switch on |
| Open captions | Deaf and hard-of-hearing viewers | Dialogue and relevant sounds | Text burned into the image |
| Subtitles | Viewers who do not know the spoken language | Translated dialogue | Text, usually switchable |
| Transcript | Deaf-blind and many other users | Dialogue plus visual descriptions | Separate text document |
A descriptive transcript, which combines dialogue with written descriptions of visuals, is especially important for deaf-blind users who rely on refreshable braille displays and can access neither audio nor video directly.
Requirements and delivery formats
Audio description is required by accessibility standards and media regulations. WCAG 2.1 requires audio description for prerecorded video at level AA (success criterion 1.2.5), and extended audio description appears at level AAA. Many regulators also set quotas for described programming on television.
| Delivery method | Advantage | Limitation |
|---|---|---|
| Separate audio track | Viewer chooses with or without AD | Player must support track selection |
| Premixed second version | Works in any player | Doubles the number of files to manage |
| Timed text description track | Cheap to update and translate | Limited player and screen reader support |
| Broadcast secondary audio service | Familiar to television viewers | Depends on receiver settings |
For web video, the most reliable approach is often a second version with description mixed in, linked clearly from the standard player, because support for selectable description tracks still varies between players and platforms.
Writing guidelines and common mistakes
Good description is objective, concise and well timed. It describes what is visible rather than interpreting it, uses vivid but precise vocabulary and never talks over dialogue or important sound effects.
- Interpreting instead of describing: say that a character clenches her fists, not that she is furious, unless the emotion is unmistakable.
- Covering dialogue: narration that overlaps speech removes information instead of adding it.
- Ignoring on-screen text: names, captions, signs and slide titles must be read or summarized.
- Describing too late: information should come just before or as the visual event happens, not after the plot has moved on.
- Overloading short gaps: when the gaps are too short for essential information, extended description is the correct solution.
The idea in brief
Audio description is a voice that tells you what is happening on screen when you cannot see it. It speaks in the quiet moments between dialogue. It mentions who walks in, what someone does and what text appears.
It is like a friend sitting next to you at the cinema. The friend whispers what is happening, but only when nobody is talking. A blind viewer can then follow the whole story, not just the parts carried by sound.
A practical scenario
A public university publishes a series of recorded lectures with many slides and diagrams. A blind student reports that the lecturer often says phrases such as “as you can see here”, which leaves out the content of the slides. The pauses in the lectures are short, so standard description cannot fit the necessary information.
The accessibility team produces extended audio description: the player pauses the video at each new slide while a narrator reads the title and summarizes the diagram. For new recordings, lecturers are trained to describe their visuals aloud, which makes later description almost unnecessary. The course reaches WCAG level AA for prerecorded video, and the student can follow every lecture without separate help.