Why Captions Matter
- 466 million peopleworldwide have disabling hearing loss
- 85% of Facebook videosare watched without sound
- Captions helpnon-native speakersunderstand content
- Captions improvecomprehension and retentionfor everyone
- Legal requirement:WCAG 1.2.2 requires captions for prerecorded video
Types of Text Alternatives
| Type | What It Is | When Required |
|---|---|---|
| Captions | Synchronized text of speech and sounds, displayed on video | All prerecorded videos with audio |
| Subtitles | Translation of dialogue to another language | When serving non-native speakers |
| Transcript | Text document of all spoken content | Audio-only content; recommended for all video |
| Audio Description | Narration of visual elements (inserted in pauses) | When important visual info isn't in dialogue |
Caption Quality Standards
Good captions are more than just words on screen. Aim for:
Accuracy
- 99% accuracy minimum— auto-captions typically achieve 70-80%
- Correct spelling of names, technical terms, and jargon
- Accurate grammar and punctuation
Synchronization
- Captions appear within 1 second of audio
- Stay on screen long enough to read (minimum 1 second per line)
- Don't overlap important on-screen text
Completeness
- All spoken words included
- Speaker identification when multiple speakers:
[Professor:] Today we'll discuss... - Sound effects that matter:
[door slams],[music playing] - Indicate music:
upbeat music
Readability
- Maximum 2 lines per caption frame
- Maximum 32 characters per line (42 for single-line)
- Line breaks at natural phrase boundaries
- Proper capitalization and punctuation
Auto-Captions vs. Human Captions
Auto-Captions Alone
- 70-80% accuracy (not sufficient)
- Struggles with accents, technical terms
- No speaker identification
- No sound effect descriptions
- Poor punctuation
Use as starting point only
Edited/Human Captions
- 99%+ accuracy
- Correct terminology
- Speaker identification
- Sound descriptions included
- Proper formatting
Required for compliance
How to Add Captions
YouTube
- Upload video to YouTube Studio
- Wait for auto-captions to generate (may take hours)
- Go toSubtitlesin the left menu
- Click on the auto-generated captions
- ClickDUPLICATE AND EDIT
- Review and correct every line
- Add speaker IDs and sound descriptions
- Publish the edited captions
Panopto
- Open video in Panopto editor
- ClickCaptionsin the left panel
- If no captions, clickImport captionsImport automatic captions
- Click on each caption segment to edit
- Use keyboard shortcuts: Tab to advance, Shift+Tab to go back
- Apply changes
Full Panopto Accessibility Guide
Microsoft Stream
- Upload video to Stream
- Go to video detailsTranscript and captions
- Auto-captions generate automatically
- Download the .vtt file, edit in a text editor, re-upload
Zoom Recordings
- Enable "Audio transcript" in Zoom settings before recording
- After recording, edit the transcript in the Zoom web portal
- Download .vtt file for use in other platforms
Caption File Formats
| Format | Extension | Use Case |
|---|---|---|
| WebVTT | .vtt | Web standard; HTML5 video; most platforms |
| SRT | .srt | Widely supported; YouTube, social media |
| TTML/DFXP | .ttml | Broadcast; streaming services |
Sample WebVTT File
WEBVTT
00:00:01.000 -->00:00:04.000
[Professor:] Welcome to today's lecture on accessibility.
00:00:04.500 -->00:00:08.000
We'll cover three main topics.
00:00:08.500 -->00:00:12.000
[slide clicking]
First, let's discuss color contrast.Audio Descriptions
Audio descriptions narrate important visual information for blind users.
When Required
- Visual content not described in the regular audio (charts, demonstrations)
- On-screen text that isn't read aloud
- Actions or expressions that convey meaning
When Not Required
- Talking head videos where visuals aren't essential
- Videos where the speaker describes what's on screen
- Audio-only content (podcasts)
Creating Audio Descriptions
- Identify visual content not covered by existing audio
- Write concise descriptions that fit in natural pauses
- Record narration or use text-to-speech
- Mix into a separate audio track or create an extended version
Live Captioning
For live events, webinars, and meetings:
Options
- Zoom Live Transcription— Auto-captions; enable in settings
- Microsoft Teams— Built-in live captions
- Google Meet— Turn on captions in meeting
- CART (Communication Access Realtime Translation)— Human captioner for high-stakes events
Best Practices for Live Events
- Enable captions before the event starts
- Speak clearly and at moderate pace
- Spell out unusual names or terms
- Pause for caption display during rapid dialogue
- For important events, hire a CART provider