A viewer switches on text for a muted video and finds different menu labels. Subtitles vs closed captions comes down to audience: subtitles translate speech for people who can hear it, while captions also convey sounds for people who cannot.
Course designers and video teams can place this choice within wider planning for accessible online learning tools. The sections below correct the most common misunderstandings and then turn them into a decision.
Subtitles and captions are not the same thing
The W3C Web Accessibility Initiative separates the two by language. W3C WAI describes captions as text in the same language as the spoken audio, and subtitles as spoken audio translated into another language. Subtitles translate dialogue only and assume the viewer can hear the audio, according to Closed Caption Creator. They carry no background sounds or sound effects. A translated subtitle track therefore leaves out the sound information a Deaf viewer needs, however good the translation is.
SDH, short for subtitles for the deaf and hard of hearing, sits between the two. It puts caption content into a subtitle file, which is why people confuse all three terms.
| Feature | Subtitles | Closed captions | SDH |
|---|---|---|---|
| Language compared with the audio | Translated into another language | Same language as the audio | Caption content in a subtitle file; can be translated |
| Assumes the viewer can hear | Yes, dialogue only | No; made for viewers who cannot hear the audio | No; carries caption content |
| Background sounds and sound effects | Not included | Included | Included |
| Typical delivery formats | SRT, IMSC/TTML | SCC, MCC | IMSC/TTML, plus other subtitle formats such as SRT and WebVTT |
| Best fit | Viewers who hear the audio but do not know its language | Deaf and hard-of-hearing viewers, and people who read better than they listen | Deaf or hard-of-hearing viewers who also need a translation, or a subtitle-style delivery route |
Definitions and the audience split come from W3C WAI, while delivery formats and the SDH description come from Closed Caption Creator.
Captions are not only for Deaf viewers
Captions are often treated as a niche accommodation, but the audience is wider. W3C WAI says captions provide content to people who are Deaf and others who cannot hear the audio, and that they are also used by people who process written information better than audio. Three groups show up repeatedly:
- Deaf and hard-of-hearing viewers. They need dialogue and the meaningful sounds around it. Plain subtitles fall short because they leave out background sounds and sound effects.
- Text-first readers. Some people take in written information more easily than speech, so same-language captions help them even when they hear perfectly well.
- Muted viewers. A video production guide from Once Films advises assuming the device has been muted, so the text has to carry the message on its own.
When the audience is unknown, same-language captions tend to serve the widest range of needs. Add translated subtitles on top when viewers may not share the spoken language.
When to use SDH
Use SDH when viewers need caption-level detail but your delivery route works with subtitle files. According to Closed Caption Creator, SDH delivers caption content in a subtitle file format such as SRT, TTML, IMSC or WebVTT. It can also be translated, which a closed caption track normally cannot. The same source lists IMSC/TTML as the route for subtitles and SDH on streaming platforms and ATSC NextGen TV, while SCC and MCC carry traditional closed captions.
- Choose SDH for a Deaf or hard-of-hearing audience that reads a language different from the audio, since sound cues and translation are both needed.
- Choose SDH when a streaming or smart-TV workflow expects subtitle-format files but the audience still needs non-speech sound information.
- Choose plain subtitles when viewers hear the audio and only need the dialogue translated. SDH adds sound description they do not need.
- Choose same-language captions when the audio is already in the viewer’s language and the delivery route supports caption tracks.
Meeting WCAG and legal expectations
Captions can meet WCAG, but only when they exist for the right media and are accurate. W3C WAI cites success criterion 1.2.2 (Level A), which asks for captions on all prerecorded audio content in synchronized media, and 1.2.4 (Level AA), which asks for captions on all live audio content in synchronized media. A prerecorded training video with no caption track fails the Level A criterion, and a live webinar without live captions fails the Level AA one.
Accuracy is the usual pitfall. W3C WAI states that automatically generated captions do not meet user needs or accessibility requirements unless they are confirmed to be fully accurate. Treat machine output as a draft that a person must check against the audio.
A translation-only subtitle track also differs from the same-language captions W3C describes, so it is not a substitute where captions are required. Outside the web standard, Closed Caption Creator notes that FCC rules under 47 CFR Part 79 require captions on nearly all US television programming. Teams publishing to broadcast should check those rules in addition to WCAG.
Platform support and display
Captions are easiest to produce on platforms that generate them for you. Once Films notes that automatic closed captions are built into most platforms, YouTube among them, and also that they risk being wrong. Use the automatic version as a starting transcript, correct it, and upload the reviewed file.
Playback is the second pitfall. W3C WAI warns that support in browsers and other media players is inconsistent and sometimes unreliable, so a caption file that looks right in one player can lose its positioning or styling in another. Test on the browsers and devices your audience actually uses before you publish.
Line length and reading speed
Limits differ by format, so write to the one you will deliver. Closed Caption Creator reports that EIA-608 captions are capped at 32 characters per row across a maximum of four rows. Netflix and most streaming specifications permit up to 42 characters across two lines for Latin-script languages. Netflix caps adult content at 20 characters per second and children’s content at 17. These are ceilings, not targets. Shorter, well-broken lines are easier to read, particularly for younger learners.
Choosing by audience need
Work through the viewer, not the file type. Ask the following in order:
- Can the viewer hear the audio? If not, provide captions or SDH so sounds are described as well as speech.
- Does the viewer understand the spoken language? If not, add translated subtitles for that language.
- Is the audience mixed? Offer a caption track plus subtitle tracks, so each person turns on what they need.
- Is a legal or WCAG requirement involved? Provide accurate, human-checked captions for prerecorded and live media.
Choosing text options is part of the wider habits covered in Digital Literacy In Education: Essential Skills Every Student Needs In 2026 – Internetbuttons. Families supporting younger learners with on-screen reading and typing can also look at Typing Practice For Kids: The Best Free Ways To Build Speed & Confidence (2026 Guide) – Internetbuttons.
Frequently asked questions
| Question | Answer |
|---|---|
| Can viewers switch captions off? | It depends on the type. Closed captions can be toggled. Open captions are always displayed and cannot be turned off, because they are burned permanently into the video picture. |
| Which file format suits captions on the web? | W3C WAI says the most common format for captions on the web is WebVTT. SRT and TTML are also used. |
| What do captions look like by default, and can viewers change them? | W3C WAI says captions are usually white characters in a black box. Some media players let users set the style, size, colour and position, though support varies. |
| Why are more people turning on text for ordinary shows? | CBC reports that increasing numbers of young viewers turn on subtitles when they watch their favourite shows. The CBC piece also explains that subtitles translate foreign-language dialogue, while closed captions serve hard-of-hearing audiences. |