Video, audio and captions: what WCAG 2.1 AA requires

Video has become a normal part of business websites: a welcome message on the home page, a product demonstration, a recorded webinar, a looping background clip behind a headline. Each of these can create barriers for people who are deaf or hard of hearing, people who are blind or have low vision, and people who are distracted or disoriented by movement and sound.

WCAG 2.1 has a group of criteria for time-based media, plus two that deal with content that starts on its own. This article explains each of them and what they mean in practice.

The media criteria

Note the difference between 1.2.3 and 1.2.5. At Level A, a full text alternative can satisfy 1.2.3. At Level AA, which is the level Manitoba's standard refers to, 1.2.5 requires audio description for prerecorded video. A transcript is still valuable, but on its own it does not meet 1.2.5.

Captions

Captions are a synchronized text version of everything in the soundtrack that matters: the words spoken, who is speaking when it is not obvious, and meaningful sounds such as [phone rings] or [music stops]. They help people who are deaf or hard of hearing, and anyone watching in a noisy place or with the sound off.

Captions should be closed captions, meaning a text track the viewer can turn on or off, rather than words burned permanently into the picture where possible. Closed captions can be resized and restyled by the viewer and can be read by other software.

Why automatic captions need editing

Most video platforms can generate captions automatically. These are a useful first draft, but they should not be published unchecked. Automatic captions commonly:

• mishear names of people, businesses, streets and products;
• misspell technical, medical or trade terms;
• omit punctuation, which makes long passages hard to follow;
• leave out speaker changes and non-speech sounds;
• get words wrong when there is background music, an accent or more than one person speaking.

A single wrong word can change the meaning of a sentence, for example a price, a date or a dosage. Review the captions against the audio, correct errors, and check that the timing matches the speech.

Audio description

Audio description is narration that describes important visual information that is not conveyed by the soundtrack: on-screen text, actions, scene changes, or a chart being shown. It helps people who are blind or have low vision.

Many business videos need little or no extra description if they are planned well. If the presenter says what is being shown ("You can see the new patio, which seats 40 people") and reads out any on-screen text, the visual content is already in the audio. Planning this into the script is usually easier than adding description afterwards.

Transcripts

A transcript is a text version of the media, placed on the page or linked next to it. A good transcript includes the dialogue, speaker names, and descriptions of important sounds and visuals. Transcripts satisfy 1.2.1 for audio-only content such as podcasts, can satisfy 1.2.3 for video, and are useful to people who are deafblind and use a refreshable braille display. They also allow anyone to skim or search the content.

Content that starts on its own

1.4.2 Audio Control (Level A)

If audio plays automatically for more than three seconds, there must be a way to pause or stop it, or to control its volume separately from the computer's overall volume. Sound that starts on page load can make it very hard for a screen reader user to hear their screen reader. The simplest solution is not to autoplay sound at all.

2.2.2 Pause, Stop, Hide (Level A)

Moving, blinking or scrolling content that starts automatically, lasts more than five seconds and is shown alongside other content must have a way to pause, stop or hide it, unless the movement is essential. This covers:

• auto-advancing image carousels and testimonial sliders;
• looping background videos behind headings;
• scrolling news tickers and animated banners;
• auto-updating content such as live feeds.

A visible pause button that can be reached by keyboard is the usual fix. Movement can be a serious distraction for people with attention-related or cognitive disabilities, and moving text can be difficult to read for people with low vision.

Practical steps

1. List every video and audio file on your site, including embeds from other platforms.
2. Remove any that are outdated rather than fixing them.
3. Add or correct captions on the rest, and add a transcript next to each.
4. Check whether important visual information is spoken; if not, add audio description or a described version.
5. Turn off autoplaying sound, and add pause controls to anything that moves.
6. Check that the video player itself can be operated by keyboard. Our keyboard test explains how.

If your videos are hosted on an external platform, our article on third-party tools covers how to assess the player. Our checklist includes media checks you can do yourself.

Talk to us

We review media as part of every audit and can prepare corrected captions, transcripts and pause controls without changing the look of your pages. See our services and method, or contact us to discuss your site.

Last reviewed October 2026.