Captions vs subtitles: the difference that matters
Ask for “subtitles” in London and “captions” in New York and you may well get the same file. The words really do mean different things, though, and the difference decides what you put in the text.
The short answer
The W3C’s accessibility guidance puts it about as plainly as it can be put. Captions are “a text version of the speech and non-speech audio information needed to understand the content”. Subtitles are for “spoken audio translated into another language”.
So the split is about what the viewer is missing. Captions assume you cannot hear the soundtrack, so they carry everything in it you would need — who is speaking, the knock at the door, the music that tells you this is the sad part. Subtitles assume you can hear perfectly well and simply do not speak the language, so the dialogue is enough.
Side by side
| Captions | Subtitles | |
|---|---|---|
| Written for | Someone who cannot hear the audio | Someone who can hear it but does not speak the language |
| Language | Usually the same as the speech | A different one, translated |
| Dialogue | Yes | Yes |
| Who is speaking | Marked when it is not obvious | Not usually |
| Sounds that matter | Yes — the door, the siren, the music | No |
| Also used by | Anyone reading in a quiet office or a loud train | Anyone learning the language |
Why you keep seeing the words swapped
Because in much of the world they are not separate words at all. The same W3C page notes that “the terms ‘captions’ and ‘subtitles’ are used for the same thing in different regions”. Where no distinction is made, everything is a subtitle, and the split people reach for instead is intralingual — the same language as the speech — and interlingual, translated.
Which is worth knowing before you argue with a client about it. If someone in the UK asks for subtitles on an English-language video for a deaf audience, they are asking for what an American would call captions, and they are not using the word wrongly.
Closed and open: the other pair of words
This one is about delivery rather than content, and it is the distinction that changes what you actually export. Most captions are closed: they live in their own track or file and the viewer turns them on and off. Open captions are drawn into the picture and cannot be switched off by anyone, ever.
Neither is better. A platform with a player of its own — YouTube, Vimeo, a course site — wants a file, because a viewer who does not need the text can put it away. A feed that autoplays silently gives the viewer nothing to switch on, so the text has to be in the picture or it may as well not exist. A good deal of the time the answer is both: one version burned in for the feed, and a file for everywhere that will take one.
What it means for the file you make
Captioning is mostly a writing job rather than a technical one. An
.srt file holds lines of text and their timings
and nothing else, so a speaker label is just words at the front of a line, and a sound cue
is just words in brackets. The convention most people follow is a name and a colon for the
speaker, and square brackets for a sound — [door slams],
[soft piano].
Which means the same file can be either, depending on what you typed into it. Transcribe the speech and you have subtitles in the original language. Add who said it and what else could be heard, and you have captions.
Getting the words down first
Either way the tedious part is the same: writing out what was said, in time with the video. OpenSubs does that in the browser — the speech is transcribed in the tab, the video is never uploaded, and you edit any line that came out wrong. From there you can translate it, add the speaker labels and sound cues a caption needs, and leave with an .srt, .vtt or .ass file, or with the text burned into the picture.
FAQ
Are captions and subtitles the same thing?
Not by the strict definition, but the words are used interchangeably in much of the world. Captions are a text version of the speech and the non-speech audio you need in order to follow what is happening; subtitles are the spoken audio put into another language. Where the distinction is not made, people say subtitles for both and separate them as intralingual and interlingual instead.
Which one do I need for YouTube or a social video?
For a platform with its own player, upload a subtitle file and let the viewer switch it on: an .srt is the safe default. For a feed that autoplays without sound, burn the text into the picture instead, because there is nothing for a viewer to switch on. Many people do both — a burned-in version for the feed, a file for the platform that will take one.
Does OpenSubs make captions or subtitles?
It writes down the speech, which gives you the words either way, and it will translate them if you want subtitles in another language. What it does not do by itself is add speaker labels or note a sound that matters — those are captioning judgements, and you type them into the lines yourself in the editor.