Closed Captions or Burned In Text: The Choice Decides Who Watches

Every video team eventually hits the same fork in the road. The captions can be a separate file the viewer switches on, or they can be baked into the picture so nobody has a choice. It sounds like a technical preference. It is actually a decision about who gets to watch your video, how it performs on each platform, and whether you will be able to reuse the footage in three years without starting over.

Closed captions are the version stored as a separate track. The player draws them on top of the video, the viewer can toggle them, change the size, sometimes change the font and background. Open captions, often called burned in or hardcoded, are part of the image itself. Once they are rendered, they are permanent.

Why the Separate File Usually Wins

Start with the boring reasons, because they are the ones that pay off. A caption file is text. Search engines can read it, which means the video becomes findable by what was said in it rather than only by the title and description you wrote in five minutes. Internal search systems index it. Journalists can quote from it. You can machine translate it into six languages in an afternoon and have a human revise the result, which is a fraction of the cost of redoing the render six times.

There is also the accessibility question, and it is not optional in as many contexts as people assume. Broadcasters and many public institutions operate under captioning rules with real enforcement behind them, and procurement teams increasingly ask for evidence before signing. A separate track is what auditors expect to see.

Then there is correction. Somebody spells a client's name wrong. With a caption file you fix one line and reupload the track. With burned in text you re-export the master, which in practice means the typo stays.

When Burning Them In Is the Right Call

All of that said, open captions exist for good reasons, and the people who insist they are always wrong have not spent much time on social platforms.

Autoplay feeds are the obvious case. On a vertical feed the viewer scrolls past in under two seconds, and a caption track that requires a tap is a caption track nobody sees. Burned in text starts working immediately. The same logic applies to video embedded in emails, to screens in shops and lobbies, and to any player you do not control, which includes most re-uploads by other people.

Burning in also lets you style the text properly. Position it away from a face, change colour when the speaker changes, move it up when a lower third appears. A caption track cannot do any of that reliably, because every player renders it differently.

The sensible answer for most teams is both. Keep a clean caption file as the master asset, and produce burned in versions for the platforms that need them. It is a rendering step, not a rewrite.

Captions, Subtitles and SDH Are Not Interchangeable

The terminology gets muddled constantly, and it matters when you are briefing a supplier. Subtitles assume the viewer can hear and translate the dialogue for a different language. Captions assume the viewer cannot hear and therefore carry the dialogue in the same language. SDH subtitles combine both ideas: same language, plus the sound cues, speaker identification and non verbal information that a deaf or hard of hearing viewer needs to follow what is happening.

Ordering the wrong one produces a file that technically works and fails the audience it was meant for. A supplier quoting you a low price for "subtitles" when you asked for accessibility compliance is usually quoting for the wrong deliverable. PoliLingua's guide to closed captions and subtitling covers the distinctions and the platform specifics in more detail than most briefs ever do.

Practical Standards Worth Holding To

Two lines maximum, thirty two to forty two characters per line depending on the language. A minimum of one second on screen and a maximum of about six. Reading speed no higher than seventeen characters per second for general audiences, lower for children's content or dense technical material.

Break lines where the sentence breaks, not where the speaker paused for breath. This single habit separates captions that feel invisible from captions that make people tired. Automatic tools get it wrong almost every time, because they segment on silence rather than syntax.

Keep the master in a plain format. SRT is universal and limited. VTT supports positioning and basic styling and is what you want if the video lives on the web. Whatever the supplier's internal tool is, insist on receiving one of those two, because it is the file you will still be able to open when the tool is gone.

What to Decide Before Filming

Most caption problems are production problems. Leave room in the lower third of the frame if text will sit there. Avoid busy backgrounds behind the safe area. Record clean audio, since every minute you save at the microphone costs three in the edit. And if the video is going to multiple markets, write the script knowing it will be read as text, not only heard.

None of this is exotic. It is the difference between a video that reaches everyone who tried to watch it and one that quietly loses a third of its audience to a choice made in an export dialogue box on a Friday afternoon.