How Audio Description Brings Meetings to Life

What You'll Learn

• What gets lost when a meeting is heard but not seen, and why captions don't cover it

• What audio description actually is, in plain terms, and what it isn't

• Who it's for, described by what people need rather than what we assume about them

• What the rule asks for, and by when

• Where to start, and why the workflow comes before the backlog

The vote you cannot follow

Your captions are done. You're not.

Picture it. The chair says, “You've all seen the map.” Somebody laughs. A commissioner asks about the hatched area on the northeast corner, staff says the setback's thirty feet there, and the vote carries four to one.

Now imagine you're listening to that recording instead of watching it. You've just been told nothing. You don't know what the map showed, which parcel is hatched, where the northeast corner even is, or whose property just changed. And the captions are perfect. Every word is there, spelled right, timed right. The meeting is still unreadable, because the part that actually decided things never got said out loud.

That's the gap audio description fills. Not a legal gap first. A gap between what happened in the room and what made it home to the person listening.

So what is it, exactly

Audio description is a second narration track that says out loud what you'd otherwise have to see. It drops into the natural pauses in the existing audio, and a second voice tells you the essential visual stuff: the slide that just went up, the figure on the chart, the map, who just stepped to the podium, the vote tally on the screen.

It's worth being clear about what it isn't, because the term gets stretched a lot.

It isn't captions. Captions carry the audio to someone who can't hear it. Description carries the visuals to someone who can't see them. Different accommodation, different audience, different production work, and neither one is a substitute for the other.

It isn't a transcript, and it isn't a summary. It doesn't interpret the meeting, take a side on the decision, or fill silence with chatter. Good description is spare. It tells you what's on screen and then gets out of the way.

And it isn't everything. Video with no essential visual content doesn't need it. If the audio's already carrying the meaning, description has nothing to add. A meeting where a slide deck drives the whole discussion is  very different from two hours of people talking, and that second one might need almost nothing.

Who it's actually for

The direct answer is people who are blind or have low vision, for whom the visual half of your meeting is just missing. That's not a rare edge case, and it's not a group you can safely assume isn't watching.

The wider answer is anybody taking in your meeting without looking at it. The person listening in the car. The resident who put the stream on while making dinner. The staffer catching up on a hearing with their phone face-down on the desk. None of them are watching the screen either, so description makes the meeting work as audio, and that ends up helping a lot more people than it was built for.

There's also an argument here that has nothing to do with any individual viewer. Your meeting video is the public record of a public decision. When the thing that decided the vote was a map nobody named out loud, the record itself is incomplete, and it stays that way for everyone who comes looking later.

What the rule asks for

This part's shorter than most compliance material makes it sound, so here it is in one go.

The standard is Web Content Accessibility Guidelines (WCAG) 2.1 Level AA. For video, that means captions on prerecorded and live video, plus audio description on prerecorded video wherever the visuals aren't already carried in the audio. That last bit is the clause that saves you work, because it's a test you apply video by video instead of a blanket order to describe everything you've ever posted.

The dates come from the Department of Justice interim final rule of April 20, 2026: April 26, 2027 if you serve a population of 50,000 or more, and April 26, 2028 if you're under 50,000 or you're a special district government, whatever your size. If you're working from a memo with a 2026 deadline on it, that memo predates the extension, and plenty of them are still floating around.

Your state might get there first, too. Colorado's HB 21-1110 has applied the same WCAG 2.1 Level AA standard to state and local government since July 1, 2025, and unlike the federal rule, it lets individuals sue in state court. Worth checking yours rather than assuming the federal date is the one that matters to you.

Where to start

The thing that has to exist by your deadline is a workflow. A repeatable step in your publishing process that puts description on meeting video before it goes up, every time, without anybody having to remember to start it up. Build that first. Everything else is just sequencing inside it.

Once you've got the workflow, point it at whatever's already doing work: the meeting video sitting on your site right now, the recording tied to a high-profile initiative, the video that's part of an active service push. That's what a resident is most likely to open, and what a complaint is most likely to name, so it earns the first pass. That's a sequencing call inside a real requirement, not a way around one.

Then start thinking about the archive, even though it isn't the day-one job.

Archived video doesn't stay out of view forever. Somebody asks for a specific old recording, and if the person asking needs an accessible version, you've got to answer that under your effective communication obligations at 28 CFR 35.160, exception or no exception. One thing worth getting precise on, because it gets muddled constantly: a records request on its own is a public-records matter. What creates the accessibility duty is somebody asking for an accessible version. In practice, the two show up together.

The Department of Justice's own example is archived town meeting minutes from 2011, covered by an exception, where a person with low vision asks for an accessible version and the entity still has to deal with the request. How you meet it can vary with the facts, and the duty isn't unlimited. The planning point is simple: that request can land on any ordinary Tuesday, and the team that already decided what they'd provide answers it in an afternoon instead of starting from zero.

How this runs in practice

The part that surprises most teams is how little of this touches the meeting itself.

Description gets added after the recording exists. It's a post-production step on a finished file, which means nobody changes how the meeting runs, nobody narrates the room in real time, and nothing new goes on the dais. The workflow picks up right where yours already ends: meeting wraps, the file lands where it always lands, description happens to that file before it goes up.

That matters because of who's doing this work. Most of the teams we talk to have one person who runs the meetings, edits the recording, posts it, and does three other jobs that have nothing to do with video. A requirement that changed the meeting would be a genuine problem for that person. A step that attaches to a file they're already handling is a much smaller ask.

MediaScribe Narrate handles that step for prerecorded meeting video, and it's built to sit inside the publishing workflow rather than off to the side of it. Anything happening in real time is a separate question with a separate answer, and mixing those two up is the most common way this conversation goes wrong.

One more thing worth doing before any of this, and it costs nothing. Ask the people presenting at your meetings to describe what they're showing while they show it, so “as you can see on this chart” becomes “Public Safety is 42 percent, Infrastructure is 28.” Description that's planned into the room is easier and cheaper than description added later, and some of it stops being necessary at all.

What good looks like

You'll know it's working when somebody can listen to the recording start to finish and come away knowing what was decided and why, without ever seeing the screen.

That's a lower bar than perfection and a higher bar than compliance. It's also the only test that matters to the person it's for.

Captions made your meetings readable. Description makes them followable. Same meeting, same recording, one more track, and the whole room finally reaches everyone in it.

Hear It on One of Your Own Meetings

Reading about a description track isn't the same as hearing one on a meeting you already know. Bring a recording of your own to mediascribe.ai and listen to what that meeting sounds like once the visual half gets spoken out loud.