Audio/Video Overview

A common misconception is that a transcript alone is enough. A transcript covers the audio content, but a prerecorded video with meaningful visual action or representations still needs a real audio description track to fully meet the standard. A simple test: if you close your eyes but can still follow the program, such as a talking head interview, you probably don't need it. A recorded slide lecture where the instructor reads everything on the slide aloud usually doesn't need it either. A screen-recorded software demo, or a video where someone gestures at a chart without describing it, usually does.

When each one is actually required

• Captions: required on essentially all video with dialogue. 

• 1. 1. For pre-recorded videos, captions are required in WCAG at Level A

     2. Live videos (like a Zoom-recorded lecture) captions are required at Level AA 

• Transcripts: encouraged for video content, but not required for all your media. They're most important for audio-only files (like a podcast), where they're the only text alternative available.

• Audio description: only needed when visuals carry information the narration doesn't already say.

The auto-caption trap

This is the single most common mistake instructors make: turning on YouTube's or Zoom's automatic captions and assuming the job is done. 

Auto-captions alone, including YouTube's default captions, are not WCAG compliant. 

Speech recognition still regularly mishears names, technical terms, and accented speech, and it doesn't reliably capture non-speech sounds. Auto-captions are a good first draft, but someone still needs to review and correct them against the actual audio.