Definition: Audio description is a narrated account of the important visual content of a video: action, people, on-screen text, changes of scene. It is fitted into the pauses in dialogue and makes the video understandable for blind and partially sighted people. WCAG requires it for prerecorded video in 1.2.5 at Level AA [1].
German: Audiodeskription · Also: described video, AD · Last reviewed: 27 September 2026
How it works
A narrator describes what happens on screen that the soundtrack does not already convey. The description is written as its own script and timed so it does not talk over dialogue. In German, a film with audio description is often called a Hörfilm.
On websites it is rarely about feature films. Product videos, explainers and training videos are typical, and those are exactly where the key facts often never reach the soundtrack:
- A price or discount appears only as an on-screen caption.
- The presenter says "click here" and points somewhere.
- A phone number or address appears at the end as text only.
What WCAG requires
| Criterion | Level | Requires |
|---|---|---|
| 1.2.3 Audio Description or Media Alternative | A | audio description or a full text version that includes the visuals |
| 1.2.5 Audio Description (Prerecorded) | AA | audio description |
| 1.2.7 Extended Audio Description | AAA | the video pauses when the gaps are too short [4] |
So at Level A, a descriptive transcript is enough. At Level AA, the usual target in law and standards, a prerecorded video with important visual content needs audio description [1][2].
When it isn't needed
If everything important is already in the soundtrack, no extra description is needed – an interview where two people talk and nothing else of note is shown, for example. The W3C therefore recommends writing the description into the script from the start: whoever speaks says what they are showing, such as "I'm opening the Settings menu at the top right" [3]. It costs almost nothing on the day of the shoot and saves a second audio track.
Putting it in place
Where that is not enough, there are two usual routes: a second version of the video with the description mixed in, or a player that offers an extra described audio track. Either way, the described version must be easy to find right next to the video. A descriptive transcript is a good addition, because deafblind people read it on a braille display.
Common mistakes
- Confusing captions with audio description. Captions serve people who cannot hear; audio description serves people who cannot see.
- Key facts shown only as on-screen text and never said.
- Description talking over dialogue, so neither can be followed.
- The described version is hidden, on another page with no link at the video.
How Reviseberg handles it
No automated check can tell whether a video shows visual information that is missing from its soundtrack, and Reviseberg has no agent today that watches video. Criteria 1.2.3 and 1.2.5 are therefore manual in Reviseberg's catalogue. They stay "untested" until a person records a result with its method, for example "listened to every video with eyes closed". Guided manual tests set out the procedure step by step. A recorded result is valid for twelve months, after which the criterion is "untested" again.
Related terms
Captions · Transcript · Screen reader · Braille display · Manual accessibility testing