Video Accessibility: Captions, Audio Description and Other Best Practices

Video accessibility ensures people who are deaf, hard of hearing, blind, low vision or neurodivergent can use video content. This guide covers captions, audio description, transcripts, accessible players and a practical workflow you can use now.

What video accessibility covers

Accessible video includes captions or subtitles, audio description, text transcripts and a player that works with keyboards and screen readers. It also covers clear language, good contrast for on-screen text and sensible playback controls. Each part helps different users, so use more than one method where possible.

Standards and legal context

The technical baseline is WCAG 2.1 Level AA, with WCAG 2.2 adding further guidance. Other relevant references include WAI-ARIA for controls and national laws such as the Equality Act or Section 508, which may apply depending on where you publish. Treat WCAG AA as the default compliance target for public content.

Captions: rules and best practices

Closed captions, subtitles and SDH

Closed captions show speech and relevant non-speech sound. Subtitles usually only show speech and help viewers who do not speak the language. Subtitles for the deaf and hard of hearing, often labelled SDH, include speaker IDs and sound cues.

Accuracy, timing and placement

• Match captions to spoken words. Edit automatic captions before publish.
• Include non-speech audio like [applause], [music crescendo] and speaker names when needed.
• Keep caption lines short. Limit to two lines on-screen at once.
• Align timing to readable chunks. Typical display time is 1.5 to 3 seconds per short caption line.
• Use a clear, legible font in the player. Do not burn captions into the image unless required.

File formats and encoding

Use WebVTT (.vtt) for HTML5 players and SRT (.srt) where VTT is not supported. Save files in UTF-8 without BOM to avoid character errors. Keep one caption file per language and label them clearly.

Audio description: what to include and how to add it

What to describe

Describe visual-only information that is important to understanding the content. Focus on actions, scene changes, facial expressions when they carry meaning and on-screen text that is not otherwise spoken.

How to produce an audio description track

1. Script the description in gaps between dialogue. Use brief, neutral phrasing.
2. Record with the same production quality as the primary audio track.
3. Provide the description as a secondary audio track or as a separate file users can choose.

Placement and mixing

Place descriptions in natural pauses. When there are few gaps, use a shortened description and prioritise essentials. Offer a described version of the whole file when detailed description is required.

Transcripts and sign language

Provide a text transcript that includes spoken content and key non-speech audio. Use an extended transcript if you need to detail visual content beyond what captions give. For sign language, offer a picture-in-picture interpreter or a dedicated signed version. Label signed versions clearly.

Accessible video player requirements

• Keyboard operable controls. Every control must be reachable by Tab and activated by Enter or Space.
• Screen reader support. Controls should have descriptive labels using native elements or ARIA where needed.
• Visible focus indicators and logical focus order.
• Toggle for captions and audio description. Allow users to change volume, speed and playback position.
• Support for the HTML5 <track> element and WebVTT files.

Step-by-step workflow for accessible video

1. Plan: decide which accessibility features you must offer. Identify language and audience needs.
2. Script: write narration and on-screen text so it is clear and concise. That reduces caption errors.
3. Record: capture high-quality audio. Use a good microphone and reduce background noise.
4. Create captions: generate a caption file, then edit for accuracy. Use WebVTT for web delivery.
5. Add audio description: script then record descriptions where visual information is essential.
6. Package: include multiple audio tracks and caption files, or provide alternate described files and transcripts.
7. Test: run keyboard, screen reader and caption checks. Do at least one test with a person who uses accessibility features.
8. Publish: include language metadata, a notice of available accessibility features and a transcript link near the player.

Testing and quality checks

• Keyboard only test: try every control without a mouse.
• Screen reader test: check with NVDA on Windows and VoiceOver on macOS and iOS for basic flow and label clarity.
• Caption review: skim for timing, speaker identification and non-speech cues.
• Audio description review: confirm that descriptions fit naturally into pauses and are not disruptive.
• User testing: include at least one person with relevant disabilities where possible.

Common mistakes and fixes

• Burned-in captions only. Fix by providing a selectable caption file instead.
• Relying solely on automatic captions. Fix by editing or commissioning human captions.
• No caption language labels. Fix by adding language tags and clearly named files.
• Missing caption placement for speaker changes. Fix by adding speaker IDs or positioning cues.
• Controls that are not keyboard accessible. Fix by using native controls or adding ARIA and keyboard handlers.

Practical tips and priorities

Start with good captions and a transcript. Captions help the most people for the least effort. Add audio description for content where important visual information is not voiced. Offer sign language or signed versions for audiences who prefer it. Make the player accessible from day one to avoid rework later.

Publishing, metadata and SEO

Include a plain text transcript on the same page as the player. Add language tags to caption files. Provide a short accessibility statement near the video that lists available features such as captions, transcripts and described versions. Search engines and assistive technologies use this information.

Final checklist before publish

• Captions in WebVTT or SRT, reviewed and UTF-8 encoded
• Transcript available on the page
• Player keyboard and screen reader checks completed
• Audio description provided if key visual information is not spoken
• Accessibility notice and language labels present

Apply these practices to every new video. That reduces risk, improves reach and makes your content usable for more people.