Burned-In Captions vs Subtitle Files
Burned-in captions, also called open captions or hardcoded subtitles, are text rendered permanently into a video's pixels, so every viewer sees them and nobody can turn them off. A subtitle file such as SRT or WebVTT, or a subtitle track inside the video container, keeps the same text separate and the player draws it at playback, so viewers can toggle it and editors can fix it without re-rendering. The difference is delivery, not content.
Side by Side
| Burned-in (open) | Subtitle file (closed) | |
|---|---|---|
| Viewer can turn it off | No | Yes |
| Viewer can resize or restyle it | No | Depends on the player |
| Survives downloads, reposts and cross-posting | Yes | No, the file is left behind |
| Styling (fonts, motion, word highlights) | Anything you can render | Whatever the player supports |
| Fixing a typo | Re-render the video | Edit a text file |
| A second language | A second render | A second file or track |
| Searchable, machine-readable text | No | Yes, for text formats |
| Encoding cost | A full re-encode | None; the file is attached |
Two Separate Questions
What the text contains. Captions are written for viewers who cannot hear, so they name speakers and describe sounds that matter. Subtitles carry dialogue for viewers who can hear but need the words, usually in another language.
How it is delivered. Open means burned into the picture; closed means carried separately and drawn by the player. Any combination exists. And closed does not always mean text: Blu-ray PGS and DVD VobSub tracks are switchable images of text that cannot be searched or restyled without OCR.
The Formats
SRT is the lowest common denominator, read almost everywhere:
1
00:00:04,120 --> 00:00:06,480
So here's the part nobody tells you.WebVTT is the W3C format for HTML5 video, with positioning and styling:
WEBVTT
00:00:04.120 --> 00:00:06.480 line:85% align:center
So here's the part nobody tells you.Note the comma in SRT and the period in WebVTT before the milliseconds: the most common reason a renamed .srt fails as a .vtt. SRT also cannot declare its character encoding, which is why accented SRT files sometimes arrive garbled. ASS adds named styles, motion and per-syllable karaoke timing, and is what most burn-in pipelines render from. Professional delivery also uses TTML/IMSC and broadcast formats such as SCC. Need to convert or retime an existing file? Use the free SRT to VTT converter or subtitle timing shifter.
When to Burn In and When to Ship a File
- TikTok, Reels and Shorts: burn in. Styled captions are part of the look, and reposts strip sidecar files.
- Long-form YouTube: prefer a file. Viewers can turn it off and YouTube's own automatic captions will not stack on top of burned text.
- Broadcast and streaming deliverables: a file. Delivery specs usually require a separate caption track.
- A video shared as a file, embedded or emailed: burn in. It is the only way to guarantee the text arrives.
Burning in has costs: every frame is re-encoded, text is frozen at the render resolution, a typo means re-rendering, and the order of operations matters. Reframe to 9:16 first and caption second, or the crop takes the ends of every line.
Common Misconceptions
- “Burned-in captions fail accessibility.” Open captions satisfy WCAG 1.2.2. No captions at all is the failure.
- “SRT supports styling.” Only text and timing travel reliably. Styling needs WebVTT, ASS or burning in.
- “Uploading an SRT captions the video everywhere.” A sidecar belongs to one upload on one platform. Download or repost the video and it is gone.
- “I can strip the burn later.” Only by inpainting. Keep a clean master if you might need one.
How Valmera Uses It
Valmera burns captions in. They are generated from a word-level transcript and timed against the finished edit, so cutting a minute of dead air never knocks them out of sync. Styles cover the short-form looks only burning in allows: word-by-word highlights, one word at a time, stacked hero words, and calm documentary subtitles, in bundled fonts with any colors. Reframing to 9:16, 1:1 or 4:5 happens in the same render, so captions are laid out for the frame they ship in, and the agent inspects rendered frames for collisions with faces, titles and platform buttons.
A typo costs one message and a re-render, because the edit is a non-destructive decision list and your upload is never modified. Valmera has no SRT or VTT import or export and adds no closed-caption track, so if your deliverable is a subtitle file, use a tool that exports one. In the other direction, Valmera can erase burned-in captions from footage before re-captioning it. See add captions to video.
Related Terms
- Word-level timestamps — The timing every caption is cut from, and what makes word highlights possible.
- Transcription and ASR — Where caption text comes from.
- Aspect ratio — Why a burn made for 16:9 loses its line ends when cropped to 9:16.
- Auto-reframe — Reframe before burning captions, or the crop takes them too.
- Text-based video editing — Editing through the transcript the captions come from.
Frequently Asked Questions
Made with Valmera: real exports
Caption a Video by Describing It
Word-timed burned-in captions, timed against the edit, styled in one sentence. Creating an account and uploading are free; editing requires a subscription.
Create an account →