Applies at Level AAA only when pauses in the foreground audio are insufficient to fit the audio description needed to convey the video
1.2.7 Extended Audio Description (Prerecorded)
In Plain Language
1.2.7 Extended Audio Description (Prerecorded) is a Level AAA criterion that applies to prerecorded synchronized media when the pauses in the foreground audio are insufficient to allow an audio description to convey the sense of the video[1]. The mechanism: pause the video at scripted timestamps, play the description audio, then resume playback once the description finishes.
1.2.7 is the escalation path from 1.2.5 Audio Description (Prerecorded) (Level AA), which requires that a description track fit inside existing dialogue gaps. When the dialogue is continuous -- a lecturer talking over a flowchart, a surgeon narrating a procedure -- there is no gap to insert the narration into, and 1.2.5 techniques cannot carry the visual content. 1.2.7 solves this by freezing the video frame so the description can run longer than the natural pause would allow. It is distinct from 1.2.3 Audio Description or Media Alternative (Prerecorded) (Level A), which accepts a full text transcript as an alternative -- 1.2.7 requires a synchronized audio track, not a document[2].
Why It Matters
- Training, lecture, and documentary video typically carries continuous narration with no scriptable pauses. Standard audio description written to 1.2.5 has nowhere to place the visual narration, so the description track drops the chart, the diagram, or the on-screen demonstration entirely.
- Blind and low-vision users depending on the description track receive an incomplete account of the content when the only available pauses are shorter than the narration required to describe a complex visual. The missing content is exactly the content that motivated adding description in the first place.
- Dense technical visuals -- wiring diagrams, surgical field views, code walkthroughs, scientific apparatus -- cannot be compressed into a two-second gap. Extended audio description is the only WCAG mechanism that gives the describer enough wall-clock time to narrate the frame without cutting dialogue.
- 1.2.7 is Level AAA and not required for AA conformance, but it is the correct fix for any video where the description track under 1.2.5 is demonstrably incomplete. Authoring tools and players that support it make it cheaper to ship than re-cutting the source video[1].
Examples
<video controls>
<source src='training-extended-ad.mp4'
type='video/mp4'>
<track kind='descriptions'
src='extended-ad.vtt'>
</video>
✔ Video pauses at key moments so the narrator can describe complex visuals in full
<!-- Extended audio description version pauses
the video when natural pauses are too short
to describe all visual content -->
<video controls>
<source src="training-extended-ad.mp4"
type="video/mp4">
<!-- Extended description track -->
<track kind="descriptions"
src="extended-ad.vtt"
srclang="en"
label="Extended audio description" default>
</video>
<!-- The player pauses the video at timestamps
where extended descriptions are needed,
plays the description audio, then resumes
the video automatically. This ensures all
visual information is conveyed even when
the soundtrack has no natural pauses. --><video controls>...</video>
<!-- Audio description only during pauses -->
✘ Standard audio description omits key visuals because the narrator talks continuously -- extended description is required
<!-- FAILS 1.2.7: important visual information
is omitted because there are no pauses in
the dialogue to insert descriptions -->
<video controls>
<source src="lecture.mp4" type="video/mp4">
<track kind="descriptions"
src="partial-ad.vtt"
srclang="en"
label="Audio description">
</video>
<!-- The audio description only covers visuals
during brief pauses. A chart shown at 3:42
and a diagram at 5:15 are never described
because the narrator is speaking over them.
Extended audio description would pause the
video to describe these visuals fully. --><div class='video-player'>
<video controls>...</video>
</div>
✔ Users can switch between standard and extended audio description versions
<div class="video-player">
<div class="ad-controls" role="radiogroup"
aria-label="Audio description version">
<button role="radio" aria-checked="false"
data-src="lecture.mp4">
Standard
</button>
<button role="radio" aria-checked="true"
data-src="lecture-extended-ad.mp4">
Extended audio description
</button>
</div>
<video controls>
<source src="lecture-extended-ad.mp4"
type="video/mp4">
<track kind="captions" src="captions.vtt"
srclang="en" label="English" default>
</video>
</div>
<!-- Let users choose between the standard
version (descriptions fit in pauses) and the
extended version (video pauses for longer
descriptions). Default to extended for
maximum accessibility. --><video controls>...</video>
<a href='transcript.html'>Read transcript</a>
✘ A separate transcript does not satisfy 1.2.7 -- extended audio description must be synchronized with the video
<!-- FAILS 1.2.7: a transcript is not a
substitute for extended audio description -->
<video controls>
<source src="demo.mp4" type="video/mp4">
</video>
<a href="transcript.html">Read full transcript</a>
<!-- A text transcript may satisfy 1.2.3
(Level A) as a media alternative, but it
does not meet 1.2.7. Extended audio
description must be delivered as an audio
track synchronized with the video, with
the video pausing when needed to allow
complete descriptions of visual content. -->How to Fix It
- Audit the existing 1.2.5 description track for gaps. Walk the video with the description track on and flag every visual element the narrator cannot fit into a natural pause -- charts, diagrams, on-screen text, physical demonstrations, UI state changes. These are the timestamps 1.2.7 has to cover.
- Script the extended descriptions against those timestamps. Write each description to the length it actually needs, not to the length of the available pause. A scripted 12-second description of a flowchart is fine -- the player will hold the frame.
- Choose a delivery mechanism. Two paths satisfy 1.2.7. Path one: author an alternative version of the source file with freeze-frames baked in at each extended-description point, so any standard
<video>element plays it correctly. This is the approach W3C documents as technique G8 (extended audio descriptions via SMIL) and is the most portable option[1]. Path two: provide a player that pauses<video>playback programmatically at cue points -- for example, a WebVTT descriptions track whose cues triggervideo.pause()in acuechangehandler, resumed by a timer once the description audio finishes. Path two keeps the source file unmodified but requires a player that actually implements the pause behavior; the native HTML5<track kind="descriptions">element does not pause video on its own. - Offer both versions and let the user pick. Extended audio description lengthens the runtime for every viewer of that stream, which the W3C Understanding document calls out as disruptive for users who do not need the extra description[1]. Expose a control to switch between the 1.2.5 cut and the 1.2.7 cut so each user gets the version they need.
- Verify synchronization. Each description must start after the visual it describes is on screen and finish before playback resumes. Off-by-one cue timing on the extended track is the most common failure mode once the mechanism is in place.
- Check the fallback chain. 1.2.7 is Level AAA, but confirm the video still meets 1.2.3 (Level A) and 1.2.5 (Level AA) on its own -- an extended description version does not exempt the base video from the lower-level criteria[2].
References
- [1] W3C (2023). Understanding Success Criterion 1.2.7: Extended Audio Description (Prerecorded). W3C, Accessed 2026-04-07. https://www.w3.org/WAI/WCAG22/Understanding/extended-audio-description-prerecorded.html ↩ ↩ ↩ ↩
- [2] W3C (2023). Understanding Success Criterion 1.2.3: Audio Description or Media Alternative (Prerecorded). W3C, Accessed 2026-04-07. https://www.w3.org/WAI/WCAG22/Understanding/audio-description-or-media-alternative-prerecorded.html ↩ ↩