Subtitles, closed captions and SDH compared

“Subtitles”, “closed captions” and “SDH” get used as if they were synonyms, but they describe genuinely different things, made for different audiences, with different content in the file. The difference matters in practice. It decides whether your track is full of [door creaks] and speaker labels you did not want, and it explains why the same movie can offer both “English” and “English [CC]” as separate options.

Subtitles are for people who can hear but don’t know the language

Subtitles, in the traditional sense, are a translation or transcription of the dialogue only. They assume you can hear everything else, the music, the explosion, the phone ringing, and just need help with the words. A French film with English subtitles gives you the dialogue in English and nothing more, because you heard the door slam yourself.

This is why classic subtitle tracks contain no sound descriptions and rarely name the speaker. The assumption is that your ears are doing that work.

Closed captions are for people who can’t hear the audio

Captions were invented for a different audience, viewers who are deaf or hard of hearing, or anyone watching with the sound off. Because the audio may be completely inaccessible, captions carry everything meaningful in the soundtrack, not just dialogue. That includes

• sound effects that matter to the story, like [gunshot] and [phone buzzing];
• music cues such as ♪ ominous strings ♪, often with lyrics;
• speaker identification when it isn’t obvious, as in JIM: I'm not going.;
• audio direction like (whispering) and (muffled).

The word “closed” is a technical detail, not a content one. Closed captions can be turned on and off, as opposed to open captions, which are permanently burned into the picture. The term comes from US broadcast television, where captions were transmitted in a hidden part of the analogue signal (the famous “Line 21”) and decoded by the TV. That lineage lives on in the broadcast caption standards CEA-608 and CEA-708, which is why “CC” remains a mostly American term.

SDH means captions delivered as subtitles

SDH stands for “Subtitles for the Deaf and Hard of hearing”, and it is best understood as the modern hybrid. The content matches closed captions (sound effects, speaker labels, music), but it is delivered as a normal subtitle track, a text file or stream in the same language as the audio, styled and positioned like any other subtitle.

SDH exists because the old broadcast caption formats don’t travel. Blu-rays, streaming platforms and ordinary subtitle files needed a way to serve deaf and hard-of-hearing viewers without Line 21. So when you download an SRT and find it full of [dramatic music] and MAN: prefixes, you have an SDH track. The filename often literally says .sdh. or .HI. (for “hearing impaired”).

A quick comparison

	Subtitles	Closed captions	SDH
Made for	Hearing viewers who don’t know the language	Viewers who can’t access the audio	Viewers who can’t access the audio
Contains	Dialogue only	Dialogue + sounds + music + speakers	Dialogue + sounds + music + speakers
Language	Usually a translation	Same as the audio	Usually same as the audio
Delivery	Subtitle track/file (SRT, VTT, ASS…)	Broadcast standards (CEA-608/708)	Subtitle track/file (SRT, VTT, ASS…)

Which one do you actually want?

• Watching a foreign film? Plain subtitles. Sound descriptions would only add noise.
• Watching with the sound off, or hard of hearing? SDH or CC, since the sound descriptions are the point.
• Making a video for social media? Captions, effectively. Most feeds autoplay muted, so the text has to carry everything. Whether you burn them in (open) or attach a file (closed) depends on the platform.
• Publishing a video to the web? Ideally offer both, a dialogue-only track per language plus an SDH track for accessibility. If you can only ship one, ship SDH. Hearing viewers can ignore [music] far more easily than deaf viewers can imagine it.

Turning SDH into plain subtitles

Here is the practical problem this vocabulary lesson leads to. The only track you can find for your video is SDH, and you’d rather not have [clears throat] interrupting the dialogue. The fix is mechanical, because SDH extras follow strong conventions. Sound effects live in [square brackets] or (parentheses), music lines carry ♪ marks, and speakers are labelled in caps followed by a colon.

Anything that follows a convention can be stripped automatically. Our subtitle cleanup tool removes hearing-impaired tags in one pass. Brackets, music lines and speaker labels disappear, cues left empty are dropped, and the file is renumbered. The reverse, note, is not possible. A tool cannot invent sound descriptions that were never written, which is another argument for keeping the SDH version as your master copy and deriving the clean version from it.

The short version

Subtitles translate dialogue for people who can hear. Closed captions transcribe the whole soundtrack for people who can’t, using broadcast-era technology. SDH carries caption-style content in an ordinary subtitle file, the best of both, and the reason your downloaded SRT sometimes narrates every door creak. If you have SDH and want clean dialogue-only subtitles, strip the hearing-impaired text automatically rather than deleting five hundred brackets by hand.