Reference
Open captions or closed captions — which to use.
One is pixels and one is data. The difference decides reach, accessibility and whether your styling survives a repost.
Open captions are burned into the video picture and cannot be switched off; closed captions are a separate track the viewer can turn on, off, resize or restyle. Open captions guarantee the look and survive resharing; closed captions are searchable, translatable and respect a viewer’s accessibility settings.
The words get used loosely, so start with the distinction that actually matters. Open captions are part of the picture — burned in, always visible, impossible to switch off. Closed captions are a separate track that travels alongside the video and that the viewer controls.
Neither is better. They fail in opposite directions, which is why most people who work at any volume produce both from the same transcript: open captions for the short-form cuts, a closed-caption track for the long-form upload.
Open captions: the look is guaranteed
Font, colour, position, motion and per-word emphasis are all preserved, because they are pixels. Nothing renders them differently.
Open captions: survive the reshare
Downloaded, cross-posted, embedded, screen-recorded — the words go wherever the file goes. A caption track can be dropped in transit.
Closed captions: searchable and translatable
YouTube indexes an uploaded caption track and can auto-translate it. Pixels are invisible to both.
Closed captions: the viewer decides
Off, larger, a different colour, a different background — a viewer’s accessibility settings apply to a caption track and cannot apply to burned-in text.
The accessibility question, answered honestly
It is tempting to treat burned-in captions as the accessible choice because they are always on. That is half right. They do guarantee that a deaf or hard-of-hearing viewer sees the words, on any platform, with no setup — which is a real and substantial benefit on feeds where caption tracks are unreliable.
But they also override every preference a viewer has set. Someone who needs 200% caption size, or a specific colour combination for low vision, or captions moved away from the bottom of the frame, gets none of that from burned-in text; they get your choices at your size. A caption track respects their settings and a picture cannot.
The practical position most accessibility guidance lands on: supply a real caption track wherever the platform supports one, and burn captions in as well where the platform’s caption support is weak or where the video will be reshared out of that platform. Doing both is not redundant — it covers two different failure modes.
- Open captions are better when
- the video is short-form, gets reshared and downloaded, needs styling, or lands on a surface with no caption-track support at all — Reels, TikTok, Status, embedded loops.
- Closed captions are better when
- the video is long-form on a platform that indexes and translates the track, or the audience includes viewers with configured accessibility settings.
- Both, when
- it is a flagship piece. There is no penalty for uploading a caption track to a video that also has burned-in text.
- Neither is a subtitle, strictly
- "Subtitles" traditionally assume the viewer can hear and render dialogue only; "captions" include speaker changes and relevant sound. In practice the words are used interchangeably online.
Producing both from one edit
The reason this is usually framed as a choice is that most workflows make it one — you either style captions in an editor or you generate a subtitle file, and doing both means doing the transcription twice.
It does not have to. One transcription pass, one round of correcting names and jargon, then a burned-in export in whichever of 240+ styles suits the cut, and an SRT or WebVTT download from the same corrected words. They cannot disagree with each other, because they came from the same source.
Open vs closed captions
| Open (burned-in) | Closed (caption track) | |
|---|---|---|
| What it is | Pixels in the video | A separate timed-text file |
| Viewer can switch off | No | Yes |
| Respects viewer’s a11y settings | No | Yes |
| Styling you control | Complete | Little to none |
| Survives download / repost | Yes | Often not |
| Indexed for search | No | Yes, where the platform indexes |
| Auto-translation | No | Yes, where the platform offers it |
| Fix after publishing | Re-render and re-upload | Replace the file |
| Works on Reels / TikTok / Status | Yes | Limited or absent |
How it works
Transcribe the video once
Word-level timings from your audio, in English, Hindi or Hinglish. The file is never uploaded.
Correct the words once
Names, products and jargon. Both outputs inherit the same corrections.
Take both outputs
Burned-in export for the short-form cuts, SRT or VTT for the platform that reads files. Both free on every plan.
Common questions
What is the difference between open and closed captions?
Open captions are burned into the video picture and cannot be turned off; closed captions are a separate timed-text track the viewer can switch on, off, resize or recolour. Open captions guarantee the styling and survive resharing; closed captions are searchable, translatable and respect a viewer’s own accessibility settings.
Which is better for accessibility?
A caption track, where the platform supports one, because it honours the size, colour and placement a viewer has configured — burned-in text overrides all of that. Burned-in captions are still valuable on surfaces where caption tracks are unreliable or absent, so the strongest answer is usually both.
Are subtitles and captions the same thing?
Traditionally no: subtitles assume the viewer can hear and render dialogue only, while captions also convey speaker changes and relevant non-speech sound for deaf and hard-of-hearing viewers. Online the two words are used interchangeably, and most tools — including this one — produce dialogue text either way.
Can I change burned-in captions after publishing?
Only by re-rendering the video and re-uploading it, which on most platforms resets the URL and the engagement on the post. A caption track can be replaced without touching the video. That asymmetry is a good reason to fix names and jargon in the transcript before you export rather than after.
Do I have to choose?
No. One transcription pass produces both: a burned-in export in any of 240+ styles, and a free SRT or WebVTT download from the same corrected words. Exports and subtitle downloads are unlimited on every plan, so producing both costs nothing extra.
