How Museums Review Audio Description and Captions for ADA-Compliant Exhibit Video

An accessibility coordinator's job on an exhibit video does not end when the captions are spelled correctly and synced to the frame, because ADA compliance is a much higher bar than basic caption accuracy, and it comes with a paper trail requirement that most creative teams are not used to producing. We see this constantly with museums preparing a new gallery video: the creative team considers the video finished once picture and sound are locked, and then the accessibility coordinator gets it four days before the deadline and has to review an audio description script against the visuals, check caption accuracy against a completely different standard than a general audience read, and somehow document who signed off on what in case anyone ever asks. That last part, the documentation, is the piece that trips up even well-run AV departments, because a museum can do everything right and still have no record proving it if the process lived entirely in email and hallway conversations.

Why ADA Review Is a Different Job Than General Caption QC

A general caption QC pass, the kind we cover in our caption QC checklist for museum exhibit videos, checks for sync, translation accuracy, and readability. ADA review checks for something else entirely: whether a visitor who cannot see the screen, or cannot hear the audio, gets functionally the same experience as a visitor who can. That means the audio description script has to convey visual information that a sighted visitor picks up automatically, artwork details, on-screen text, a demonstration happening in frame, and it has to do it inside the natural pauses in the existing narration without talking over anything important.

The accessibility coordinator is not proofreading at this point, they are evaluating equivalence, which is a fundamentally different review task and one that most video platforms built for general creative feedback were never designed to support well.

Reviewing an Audio Description Script Against the Actual Cut

The single biggest mistake we see is an audio description script getting approved as a standalone document, read and reviewed on paper, without ever being checked against the actual timed video it needs to fit inside. A script can read beautifully and still fail completely once you try to fit it into the eleven seconds of silence between two lines of narration in the real edit.

We saw this exact failure happen on a four-minute orientation film where the AD script writer, working entirely from a paper transcript, wrote a description of a painting's brushwork that ran fourteen seconds, aimed at a gap the writer assumed was roughly fifteen seconds long based on the timecodes printed in the transcript. The actual gap in the finished edit, once a music sting and a slight pause for emphasis were factored in, was closer to nine seconds, and nobody caught the mismatch until the accessibility coordinator played the finished mix and heard the description get cut off mid-sentence by the next line of narration.

This is exactly the kind of review that benefits from Audio Annotation tools built for exactly this purpose, letting a coordinator drop a comment on the precise waveform location where a described element needs to start or end, instead of describing the timing in a separate document that the editor then has to translate back into the timeline by hand.

When There's No Natural Pause for Description at All

Some exhibit videos, especially fast-cut orientation films or anything edited for a younger audience's attention span, simply don't leave enough silence anywhere in the runtime to fit meaningful audio description without talking over the existing narration. When that happens, the fix generally isn't cramming a shorter description into a smaller gap, because a rushed description that skips visual detail defeats the same purpose a missing description does, it's building what's usually called extended audio description, where the video itself briefly pauses at specific points so the description has room to play before the narration resumes. That's a bigger production decision than a script edit, since it means the edited runtime for the accessible version is genuinely longer than the standard cut, and it needs sign-off from whoever owns the exhibit's overall runtime and gallery flow, not just the accessibility coordinator. Reviewing this version well means checking two things at once, whether each pause is actually long enough for the description as written, and whether adding runtime at that specific point still makes sense next to whatever's happening in the gallery physically, since a video that pauses for eight extra seconds at a kiosk with a line of visitors waiting behind it creates its own separate problem that the accessibility coordinator and the AV lead need to weigh together before locking the extended version as final.

Caption Accuracy Against a Compliance Standard, Not a Reader's Ear

Caption review for general accuracy asks whether the words are spelled right and land on the correct frame. Caption review for ADA compliance asks a stricter set of questions on top of that: does the caption identify the speaker when there is more than one voice on screen, does it note relevant non-speech sound like a bell tolling in a historical demonstration video, and does it avoid paraphrasing dialogue in a way that would fail a formal accessibility audit later.

The catch here is that this second pass has to happen even when the first caption pass, the one covered in our general QC checklist, already came back clean, because the two reviews are checking for genuinely different things and one clean result does not imply the other.

Building a Paper Trail That Holds Up If Someone Asks

At the end of the day, the reason accessibility coordinators push so hard for documentation is that they are the ones who have to answer for it if a visitor or an advocacy group raises a concern after the video is already live in the gallery. "We reviewed it" is not an answer that holds up. "Here is the timestamped record showing the accessibility coordinator approved this specific version on this specific date, with these specific notes resolved" is.

• Every AD script reviewed against the actual timed video, not just on paper
• Caption file checked for speaker labels and non-speech sound notes
• Formal written sign-off from the named accessibility coordinator, not a verbal okay
• Timestamped record of exactly which version was approved
• Approval history retained for the life of the exhibit, not just through opening week

We built Approvals into PlayPause specifically because we kept hearing from institutions that had done the review work correctly but had no clean way to prove it after the fact. A locked approval record, tied to a specific version of the video and a specific reviewer's name, is basically an insurance policy that costs nothing extra to maintain if it is built into the review process from the start instead of bolted on afterward.

Who Actually Needs to See the File, and Who Doesn't

Accessibility review often involves people outside the immediate AV and curatorial team, an outside ADA compliance consultant, a disability advocacy liaison, sometimes legal counsel for larger institutions, and every one of those additional reviewers is another person who needs access to sensitive pre-release footage without needing a login to the museum's main production system. This is where Sharing Security and controlled review links matter more than people expect going in, because a video is not "released" until it is live in the gallery, and a leaked pre-release cut of an exhibit video, even a short docent-style piece, can undercut a carefully planned unveiling.

We tell museum accessibility teams who ask us this to treat every outside reviewer's access the same way they would treat a media embargo, generous enough to get real feedback, but closed the moment the review round ends.

Coordinating This Across a Touring Show or Multiple Reviewers

Institutions running a touring exhibit have an added wrinkle here, because ADA requirements and how strictly they get enforced can vary by venue and by state, which means the accessibility review has to happen fresh at each stop rather than being treated as a one-time approval that travels with the video. Our piece on how traveling exhibits track video caption localization across multiple museum venues covers the venue-by-venue coordination problem in more depth, and the same logic applies directly to accessibility sign-off: what clears at the originating institution does not automatically clear at every host venue down the line.

Keeping Feedback in One Thread Instead of Three

When an AD script writer, a caption vendor, and an in-house accessibility coordinator are all weighing in on the same video, their notes need to live in one place or the editor ends up reconciling three separate documents by hand, which is exactly how a described element gets fixed in the script but never actually makes it into the final audio mix. Routing everyone through the same Video Feedback thread, with comments pinned to timecode, keeps the AD writer's note about a missed visual detail sitting right next to the accessibility coordinator's note about speaker labeling, on the same frame, instead of scattered across a script document, an email, and a spreadsheet.

We built this into PlayPause after hearing the same complaint from enough accessibility coordinators, that they were being handed a "final" video with no visibility into whether their previous round of notes had actually been addressed, or just acknowledged in an email and quietly dropped. A shared thread with resolved and unresolved states solves that specific trust gap, because a coordinator can see exactly which of their notes are marked done versus still open, without having to re-watch the entire video from the start to check.

What We Tell Accessibility Coordinators Who Ask Us This

Basically, the museums that handle this well are the ones who stop treating accessibility review as a compliance checkbox tacked onto the end of production and start treating it as its own review discipline with its own reviewers, its own pass criteria, and its own documented sign-off, the same weight given to curatorial approval. The American Alliance of Museums has increasingly framed accessibility as core to the visitor experience rather than a legal minimum, and the institutions that internalize that framing early tend to have far calmer final weeks before an opening, because the accessibility pass is not competing for the same emergency attention as everything else that surfaces in the last 48 hours.

If your museum is still managing audio description scripts, compliance captions, and outside consultant access across email threads and a shared drive with no real record of who approved what, that gap is exactly what a proper review system closes. Talk to PlayPause about your exhibit review workflow and put a real, timestamped paper trail behind every accessibility sign-off before your next opening.