Audio and Video Accessibility: Captions, Transcripts and More

Video accessibility

High-quality recorded video with audio benefits everyone. 

Two types of video

• Prerecorded video with audio
  • A closed caption file (.WebVTT or .SRT formats) must be provided along with the video. 
    • Videos can be uploaded to the YuJa Enterprise Video platform - they are auto-captioned, and you can link to them within D2L.
• If the captions do not describe key visual information, then audio descriptions would be needed. 
  • Write these to describe that key visual information and manually add audio descriptions via the YuJa video platform. 
• Live video with audio streaming
  • Provide live captions for live video streaming events (called CART services: Communication Access Realtime Translation).
  • A sign-language interpreter can also be considered.
  • After the event, the recorded video can be uploaded to YuJa Enterprise Video platform. 
    • The video will be auto-captioned. Check for accuracy.

About captions and transcripts

Why are these important?

• Vital for people with hearing difficulties or deafness. They provide access to dialogue and explain non-speech sounds like music and laughter that occur in the media. 
• Also help identify speakers, overcome audio challenges like background noise or distinguishing people’s accents, and help to ensure overall access to media content. 

Captions

• There are two main types: 
• Closed Captions (CC) can be turned on or off in a media player when provided with the video.
• Open Captions (OC) are permanently embedded into a video.
• Captions must be accurate and properly synchronized with the video and audio.
  • Inaccurate captions can lead to miscomprehension of video content, especially with complex or specialized terminology like math and science equations, formulas, etc.
• Should include all dialogue, crucial sounds, and meaningful non-speech information like music.
• Captions need to clearly identify a speaker, especially when there are multiple speakers.
• Help to improve comprehension for individuals with cognitive impairments, and they can enhance focus, boost retention, and support diverse learning styles.
• Captions are different from subtitles: 

Subtitles are used for language translation and it is assumed the user can hear the audio but may not understand the language, so they typically only translate dialogue.

More about captions

• If a video is hosted on the YuJa Enterprise Video platform and embedded into D2L, then Panorama in D2L would check for the presence of captions.
• When uploading a video to the YuJa video platform it automatically captions it. The first screen shot below shows Media Details > Accessibility page. The second screen shot shows using the Video Editor to edit video captions for accuracy, spelling errors, and context.

AI tools and audio/video accessibility

• Generative AI tools, like Google Gemini, ChatGPT and others, can be used by students and instructors. Provide prompts to have the tool automatically analyze a transcript and list summaries and key takeaways, helping users to process and review content more efficiently.
• Also when given prompts, these tools can analyze video content, identify key visual elements, and automatically generate audio descriptions.

AI-generated captions and transcripts can pose accessibility challenges

• These text files will need people to review them and fine-tune text descriptions to ensure accuracy, spelling, proper context, and that nuances in emotional tone are explained.
• Another thing to check is timing of captions to make sure that they are not out of sync with the audio and video.

Audio descriptions

• Audio description (AD) is a secondary audio track that narrates key visual information for people who are blind or who have low vision.
• Narrates essential visual information not conveyed through dialogue that is captioned (e.g. scenes, actions, settings), ensuring full comprehension.
• After writing or generating an audio description, set these up manually in the YuJa video platform. Automatically pause a video to insert a description of a complex scene with key visual information.

Flashing content in video

• Avoid video content that flashes too quickly - more than 3 times per second - can trigger seizures in users with photosensitive epilepsy. Avoiding use of flashing content at all is best.
• Photosensitive epilepsy is a type of epilepsy where seizures are triggered by flashing lights, patterns, or other visual stimuli.
• When creating video content, avoid any rapid changes in brightness or color that might be perceived 
  as a flash.

Animated GIFs

• Avoid auto-playing animated gifs. Let the user choose to play/pause.
• The GIF must not contain anything that flashes more than three times in any one-second period.
• Descriptive GIFs: If the GIF is essential to understanding the content (e.g., showing a step-by-step process), the alt text must describe every meaningful frame or action.
• Illustrative GIFs (Memes): If the GIF is purely for illustrative or emotional effect, the alt text should convey that emotion.

Interactive Simulations

• For an interactive simulation that has an audio track - treat this like a video: captions must be provided to help people with hearing impairments interpret the content and context of the simulation.
• If a user can't use a mouse, and uses a keyboard to navigate, they must be able to navigate using the Tab, Enter, and Arrow keys.
  • When a keyboard user tabs to buttons or links, focus Indicators (a highlight) should display on any button or link within a simulation.
• Provide "Help" instructions that explain how to navigate the simulation with a keyboard.
• Avoid time limits, or allow the user to pause/extend the time.
• For an interactive simulation that has no audio track, provide a text version of what occurs within the simulation and what data is shared - adjacent to the simulation.
  • Example: "In this simulation, the user can drag a magnet toward a coil. As the magnet enters the coil, the voltmeter needle moves to the right, indicating a positive current."

Conclusion

• Providing captions and transcripts are core methods for achieving media accessibility.
• Providing these tools - is a requirement of the Web Content Accessibility Guidelines (WCAG 2.1 AA).
• Prerecorded audio-only content (e.g., podcasts) must be accompanied by a transcript. 
• Prerecorded videos with audio must include accurate closed captions and, if key visual information is not spoken, audio descriptions must also be provided.
• Both live audio and video streaming events require live captions.
• AI tools can be used to generate captions, transcripts, and audio descriptions, but faculty need to review them for accuracy, spelling, and proper context.
• If a video is hosted on the YuJa Enterprise Video platform and embedded into D2L, Panorama would check for the presence of captions.
• Media players must be fully operable by keyboard, not just with a mouse, and must support display of captions, transcripts, and audio descriptions.
• Avoid video content that flashes too quickly - can trigger seizures in users with photosensitive epilepsy.
• Avoid auto-playing animated gifs. Let the user choose to play/pause the content.
• Treat an interactive simulation that has an audio track like a video: captions must be provided to help people with hearing impairments interpret the content and context of the simulation.
• For an interactive simulation that has no audio track, provide a text version of what occurs within the simulation and what data is shared - adjacent to the simulation.