Video & Audio

Anyone who creates new or uses existing video or audio content for university courses, websites, services, presentations, etc.

Start with Basics: 

• Complete the Multimedia Accessibility Training (U-M Canvas Course | Estimated time to complete: 60-90 minutes).
• Review Best Practices to Make Video & Audio Accessible (U-M Resource | Estimated time to complete: 10 minutes).
• Choose media players that support accessibility (U-M Resource | Estimated time to complete: 5 minutes).

For instructors using lecture capture, contact your unit IT group for information about the default web player and accessibility features.

• Individuals leading digital accessibility efforts for units or teams
• Instructors who use video and audio content in their courses
• Web content and social media content managers

Use the video and audio files accessibility inventory template (or create your own system) to assist with the review process. 

Step 1: Inventory your videos and audio files

Note the name, format, location, and length of each file.

Methods for identifying video and audio files may include

• For website content, using the digital asset management system to filter files with video and audio file formats.
• For Canvas course content, checking the "Media Gallery" for a list of uploaded video and audio files. Then do a manual check through each page for: files embedded into a course page from external platforms like MiVideo or YouTube as well as links to video and audio content on external websites or stored in external folders (like Drive or DropBox).
• For content stored in shared cloud drives, using the advanced search feature to locate video and audio files. 

Step 2: Assess use

Is the video or audio file still current/in use? Gather usage data such as downloads, views, or date last accessed to help determine remediation priority as well as whether the file could be considered for archiving. If you are the file owner and the file is no longer used or is outdated, consider if the file qualifies for archiving under Title II of the ADA. Eligible, properly archived video and audio files do not require proactive remediation.

Step 3: Assess accessibility

Assess and document the following:

• Do videos have captions?
  • If yes, were the captions auto-generated or human-edited?
    • If unsure, document that the source is unknown.
• Do audio files have an associated transcript (or caption file if uploaded to a video player)?
  • If yes, is the transcript (or caption file) auto-generated or human-edited?
    • If unsure, document that the source is unknown.
• Do videos contain visual content that is key to a viewer's understanding of the topic? (Examples of videos to which this may apply include lab demonstrations, tours through architectural or historic landmarks, dance performances, etc.)
  • If yes, is there adequate audio description of the visual content?
    • To determine if audio description is adequate, ask yourself: Could an individual understand important details and/or follow along with what was happening on screen based on the audio alone?

Step 4: Prioritize for remediation

Video and audio files don’t require remediation if they: 

• Are removed or archived (refer to archiving guidance in Step 2).
• Have an accurate, human-edited captions file or transcript.
• Have important visual information adequately described in the existing audio (video only).

All other video and audio files will require remediation. 

Prioritize remediation as follows: 

High priority: Video or audio files that meet any of the following criteria:

• Located on a main webpage, and/or
• Required to perform well in a course, and/or
• Required to access a program, service, or benefit at the university, and/or
• Communicate critical information (e.g., unit policies, application deadlines, evaluation results, etc.), and/or
• Frequently viewed/listened to, and/or
• Have been reported for accessibility issues.

Lower priority: All remaining video and audio files. This may include videos and audio files where the source of the captions and/or transcript is auto-generated or unknown, or where the captions and/or transcript require editing to reach 99% accuracy.

Units and teams need to determine who is responsible for remediating video and audio files. Some units may prefer to allocate employee time to correct automated captioning and transcripts, as well as add audio description where needed. Others may prefer to add a budget line to outsource video and audio remediation to professional suppliers.

If U-M employees are doing remediation consider:

• Identifying people fluent in the language(s) and any technical jargon within the files.
• Getting support from student employees.
• The amount of time needed to complete work (time estimates are noted in an accordion below).
• The complexity of the file (captioning and transcription is difficult to do well if individuals are speaking over one another or audio quality is poor, and audio description can be difficult to do well when it is added in post-production).

If outsourcing remediation work, check the U-M supplier portal for preferred suppliers. Request pricing and budget appropriately.

Captions and transcripts should be 99% accurate. Audio description should enable someone to understand what is visually happening on screen just by listening (or by reading an enhanced transcript). 

Follow caption best practices, transcript best practices, and audio description best practices.

Start by reviewing best practices for captioning & transcription and best practices for audio description

Video & audio files you didn't create/don't own

1. Do a search for the file online to see if any versions exist with accurate captions or an accurate transcript. If found, utilize that version instead. If not, move to step 2.
2. Generate a transcript using an automated tool (refer to the tools linked below).*** You usually can't edit caption files for videos you don't own, so a transcript posted with the video may be the only option for creating access.
3. Edit captions/transcript for accuracy.
4. If a video contains visual content that needs to be described, include a description of visual content in the transcript.
5. Post the transcript to where the video or audio file is hosted.

Video & audio files you created/own

1. For videos with visual content that needs to be described, decide how to add a visual content description. If you decide to add to the audio track, this will affect your caption file so finalize the audio track first. If this doesn't apply, skip straight to step 2.
2. Generate automated captions or a transcript (refer to the tools linked below).***
3. Edit captions/transcript for accuracy (and add visual content descriptions to transcripts if selected in step 1).
4. Ensure the old video file is replaced with the captioned version OR that the transcript is added to where audio file is hosted.

***Videos & audio files containing languages other than English: 
If captioning/transcribing in the original language(s), select an automated captioning tool that supports that language. If files contain more than one language you may need to generate two files and then edit them together, as automated tools can typically only manage one language at a time. If you wish to provide translated captions, you will need to use an automated tool that also supports translation (remember to edit!) or send to a professional translator.

Review Media Player Accessibility Features for a list of the captioning, transcription, and audio description features available through commonly used media players.

If creating a caption file or transcript from scratch, it will take a professional transcriptionist about 4–5 hours per 1 hour of content. Novice transcribers typically take about 8–10 hours per 1 hour of content. This is why using automated captioning with human editing is strongly recommended.

Automated captioning tends to achieve 60–85% accuracy (assuming there is good audio quality and that speakers do not have strong accents). Expect to spend 2–4 hours editing every hour of automated caption/transcript content, depending on quality.

The time required for post-production audio description varies dramatically depending on factors such as the length of the video, the amount and complexity of the visual content depicted, the experience level of the describer, and the method used to fit audio description in with existing audio tracks.

• Individuals with access to reporting dashboards (integrate into monitoring metrics)
• Content creators/managers (integrate into workflow when selecting, creating, and/or preparing to post videos)

Your inventory of video and audio files should be periodically updated and reviewed to determine if further archiving or remediation is needed. Content that is no longer current can be considered archival under archived web content exception. Please refer to our guidance page for understanding suggested archiving practices.This can be added as a step in existing review processes for websites, applications, courses, etc. or you can establish a review timeline. We recommend revisiting your inventory 2-3 times per year and placing recurring reminders on your calendar or in project/task management applications you use.

Video & audio files you didn't create/don't own

• Factor in time for reviewing captions and transcripts for accuracy and for generating transcripts where accurate captions or transcripts do not exist.
• Aim to select videos and audio files that already have accurate captions and/or transcripts available.

Video & audio files you create/own

• [VIDEOS ONLY] Prompt speakers in videos to verbally describe any important visual content, to avoid needing to add described video tracks or transcripts later on.
• Use quality recording equipment and do sound checks to ensure clarity and reduce background noise. Direct speakers to talk into/toward microphones, to slow down speech if talking too quickly, and to avoid talking over one another where possible.
• Factor in time for generating automated caption or transcript files, and then editing them for accuracy.

Commit to not posting video or audio content until accurate captions and/or an accurate transcript is also available. If possible, update unit or team video and audio file inventories as you post new content, to save time in the future.