Why performance video breaks the usual habits
A board meeting is mostly people talking. A spring concert, a dance showcase or a high school musical is the opposite. Most of the sound is music. The words are sung, not spoken. Several voices overlap. Nobody on stage explains what the audience is seeing.
The habits that work for a meeting stop working here. You can't just caption each speaker and let the speech carry the meaning. Performance video gives you two separate jobs. Captions have to carry the music: what is sung, what is played, and how the audience reacts. Audio description has to carry the picture when the stage shows something the soundtrack never says.
This post takes each job in turn, using the guidelines published by the Described and Captioned Media Program (DCMP). If you want the event-day checklist for a specific show, the winter concert post covers that. This one is about how the captions and description should actually read.
What WCAG 2.1 AA asks of a recorded performance
Two success criteria do most of the work for a recorded performance. 1.2.2 Captions (Prerecorded), Level A, asks for captions on recorded audio. That covers the music and lyrics, not only the principal's welcome and the conductor's remarks. 1.2.5 Audio Description (Prerecorded), Level AA, asks for description when the picture carries meaning the audio doesn't. The guide on when audio description is required walks through that test in detail.
Some things sound related but are Level AAA, so the rules don't require them. That includes 1.2.6 (sign language), 1.2.7 (extended audio description, where the video pauses for description), 1.2.8 (a full media alternative) and 1.4.7 (low or no background audio). A sign language interpreter on stage is a good thing to offer. It isn't what WCAG 2.1 AA asks for.
The dates, as extended in 2026: under the DOJ's ADA Title II rule, public entities serving 50,000 people or more have until April 26, 2027, and smaller entities and special districts have until April 26, 2028. Under HHS's Section 504 rule, which covers HHS-funded health centers and others, recipients with 15 or more employees have until May 11, 2027, and those with fewer have until May 10, 2028. Both rules adopt WCAG 2.1 AA. Check which date applies to you.
This post explains what the rules and guidelines say. It isn't legal advice.
The DCMP Captioning Key on music
Lyrics
The DCMP Captioning Key says that if music contains lyrics, you caption the lyrics verbatim, marked with music icons (♪). Its lyrics guidance adds that you introduce the lyrics with the vocalist's or group's name, and add the song title in brackets when it is known or significant and there is time on screen to read it.
The Captioning Key also sets the placement of the icons: one ♪ at the start and end of each caption within a song, and two at the end of the song's last line. That double icon tells a viewer the song is over, which matters in a concert where one piece runs into the next.
Instrumental music
A band piece or an orchestral movement still gets captions. The Captioning Key calls for a bracketed description that names the performer or composer and the title when possible. The description should convey mood and stay objective. DCMP says to avoid words like “beautiful” or “melodic,” which judge the music instead of describing it. “Slow and quiet” tells a viewer something. “Lovely” doesn't.
Background music
Not every note needs a caption. The Captioning Key marks nonessential background music with a single ♪ in the upper right corner, and says not to caption background music that lasts under 5 seconds. In performance video, background music usually means the walk-in music before the show or the bed under a slideshow, not the performance itself.
Applause, laughter and silence
YouTube Help tells creators writing their own captions to add bracketed non-speech cues such as [applause], so viewers know what is happening. Do the same for laughter during a comedy scene or cheering for a soloist. When a pause is part of the piece, such as a long rest before the final chord, a cue like [silence] tells the viewer the sound stopped on purpose and the video didn't break.
CONDUCTOR: Our next piece is a traditional American folk song. [Concert Choir sings “Shenandoah”] ♪ Oh Shenandoah, I long to hear you ♪ ♪ Away, you rolling river ♪ ♪ Away, I'm bound away ♪ ♪ Across the wide Missouri ♪♪ [applause] [Symphonic Band plays “Chester” by William Schuman: slow, hymn-like brass opening] [tempo quickens, full band, loud and driving] [silence] [applause and cheering]
The song and names above are an example. Swap in your own program, ensemble names and lyrics.
Choir, band, dance, theater: what each one needs
Captions are needed for all of them. Audio description depends on how much of the meaning is visual. A seated orchestra and a dance recital are very different videos.
| Performance | Captions | Audio description |
|---|---|---|
| Choir | Lyrics verbatim with ♪, group and soloist names, song titles | Often not needed if the choir stands and sings. Needed for movement, props or on-screen text the audio doesn't mention. |
| Band or orchestra | Bracketed descriptions of each piece with composer and title, spoken remarks, soloist names | Rarely needed for a seated group. On-screen titles and soloist names that nobody says aloud still count as visual information. |
| Dance | Music cues with title and composer, applause, any spoken introductions | Usually needed. Movement, formations and costumes carry the meaning, and the soundtrack doesn't explain them. |
| Theater or musical | Dialogue with speaker names, lyrics with ♪, sound effects | Usually needed for setting, costumes, action and who is on stage. |
DCMP's Description Key gives describers some useful limits. It treats song lyrics like dialogue, so description shouldn't talk over them. If description is needed during a song, it should wait until verses repeat. That rule doesn't apply to background music with no contextual or educational purpose. More broadly, describers should talk over music only when necessary.
For dramatic productions, the Description Key says description should cover style, setting, focus, period and dress. It also notes that skill in describing live theater doesn't carry over to describing dance. If you write description in-house, don't assume the approach that worked for the spring play will work for the dance showcase.
Livestreams and old recital uploads
Livestreamed performances
If you stream a concert live, WCAG 1.2.4 (Level AA) asks for captions on the live audio. YouTube's live automatic captions alone usually fall short, and singing is harder for speech recognition than ordinary talking: held notes, harmonies and accompaniment all get in the way. YouTube Help says automatic captions can misrepresent content and that creators should review and correct them before relying on them. That page doesn't address music specifically, but the advice applies.
Once the stream ends, the replay is recorded video. It needs accurate captions under 1.2.2, and description under 1.2.5 if the picture calls for it. The livestream captions guide covers both halves.
Old concert and recital videos
Recital and concert videos posted to YouTube before your deadline are probably exempt (unclear). The rules exempt social media posts made before the deadline, but neither the DOJ nor HHS said whether YouTube counts as social media. A video embedded on a current page of your website is in scope regardless of when it was recorded.
The archived content exception doesn't apply just because a video is old. All four conditions have to hold: it was made before the deadline, it is kept only for reference, it hasn't changed since, and it sits in an area clearly labeled as an archive. The archived video guide explains each one. Either way, when someone asks for an accessible version of an old performance, you need to provide one.
A small-team workflow you can start this week
- 1
Find your performance videos
Scan your channel to see which concert, recital and play videos lack real captions, and run the website video check to find ones embedded on current pages. Those embedded videos come first.
- 2
Ask for the program and lyric sheets before the show
Get the printed program, song list and any lyric sheets from the music or drama teacher ahead of time. Captioning lyrics from a sheet is far faster and more accurate than transcribing them by ear from a room recording.
- 3
Build a name list
Write down ensemble names, soloists, the conductor, composers and song titles with correct spellings. Use it for speaker labels, lyric introductions and instrumental descriptions.
- 4
Caption by the conventions
Never publish auto-captions as they are. Edit them, or write captions from scratch: lyrics verbatim with ♪, the song title in brackets, an objective bracketed description for each instrumental piece, and cues for applause.
- 5
Flag dance and staged pieces for description
Mark every dance number, scene and segment with on-screen text that nobody reads aloud. Run each flagged video through the audio description check, and see how to add audio description to YouTube for publishing options.
- 6
Set your order and publish a request route
Confirm your date at /deadline. Then work through new performances first and older embedded ones next, and post a clear way for families and patients to ask for an accessible version of anything else.
Sources
- Described and Captioned Media Program (DCMP), Captioning Key
- Described and Captioned Media Program (DCMP)
- Described and Captioned Media Program (DCMP), Description Key
- YouTube Help
- YouTube Help
This is a reading of the rules, not legal advice. Something out of date? Tell us and we'll fix it.