- Level
- A
- Since
- WCAG 2.0
1.2.3 Audio Description or Media Alternative (Prerecorded)
What a prerecorded video shows and does not say has to reach people who cannot see it, through audio description or a full text version.
/ What it requires
A video can show what its soundtrack never says: the price on a slide, the name under a face, the button the presenter clicks, a change of scene. People who are blind or have low vision miss all of it. This criterion asks for one of two remedies for prerecorded video with sound. Either an audio description, narration that describes the important visual details in the pauses of the dialogue, or a full text alternative that reads like a script, with every line of dialogue and a description of everything on screen, in the order it happens.
If the soundtrack already conveys all the important visual information, nothing more is needed. The alternative has to be in the language of the video or the page, and if the video asks the viewer to do something, the text version has to offer the same action. The exception for media that only repeats text on the page and is labelled as such applies here too. At level A the author chooses between the two remedies. At level AA, 1.2.5 requires the audio description.
/ How pages fail it
- A software demo shows the steps on screen while the presenter says "click here, then here", and there is neither an audio description nor a text version that names what is clicked.
- A promotional video reveals the offer and its price only as text over music.
- The transcript next to a training video holds the dialogue only, while the screen shows procedures the speaker never puts into words.
- A recorded webinar shows charts and figures the speaker introduces with "as you can see" and never reads out.
- The described version of a video is offered only in English on a page written in Polish.
/ How to check it yourself
- Watch each prerecorded video with sound and write down every visual detail that matters and is not spoken: on-screen text, names, actions, scene changes and charts.
- Listen to the same video without looking at it. If everything on the list is said, the video meets the criterion as it stands.
- If not, look for a remedy: an audio description button or a second audio track in the player, a described version linked next to it, or a text alternative.
- Read the text alternative against the video. It should hold the dialogue and the visual description in the order of the video, and offer links for anything the video asks the viewer to do.
- Check that the alternative is in the language of the page.
/ How to fix it
- Write the script so the speaker says what the screen shows: "Open Settings, then select Billing" instead of "click here".
- Publish a text alternative with the dialogue and the visual description in sequence, as HTML, and link it right next to the player.
- Offer an audio-described version, either as a second audio track the viewer can select or as a separate described video linked from the player.
- For a talking-head video whose only picture is the speaker, a short static text description of the speaker and the setting is enough.
- Include every piece of on-screen text in the description: prices, addresses, names in lower thirds and the text of slides.
/ What our scanner checks
There is no automated check for 1.2.3 in our scan. Deciding whether the picture carries something the soundtrack does not, and whether a description or a text version makes up for it, means watching the video once with the sound on and once without looking. That comparison is part of a manual audit.
/ Who fails it, in our own measurements
The weekly index has no figures for this criterion, because none of the checks the site scan runs can test it. A person judges it, which is what a manual audit is for.
One page per body, measured by automated checks on a weekly schedule. Any body that has fixed something since the last sweep will show it at the next one, and every entity page carries a right of reply.
/ Questions about this criterion
Do I need audio description or a transcript for 1.2.3?
Either one meets 1.2.3 at level A, as long as the text version describes the visual content as well as the dialogue. At level AA, however, 1.2.5 requires an audio description for prerecorded video. A site aiming for AA usually provides the audio description, which then satisfies both criteria at once.
Does a talking-head video need audio description?
Usually not. W3C notes that if all the important information in the video is already in the soundtrack, no additional audio description is needed. A person speaking to the camera with no slides, captions or demonstrations conveys everything through the audio. Once on-screen text or a demonstration carries information of its own, it needs describing.
Is a transcript the same as a media alternative?
No. A transcript of the dialogue covers what is heard. The alternative W3C means here also describes what is seen: actions, scene changes, on-screen text and the expressions that matter, in the order they occur, so that it reads like a script. A dialogue-only transcript next to a video that shows things the speaker never says does not meet 1.2.3.
