Subtitles vs Closed Captions: What's the Difference
Subtitles give viewers the words; closed captions add the sounds and let viewers switch them off. Here is how open captions and SDH fit in, and which to use for your videos.
Subtitles and closed captions both put text on screen, but they answer different questions. Subtitles assume the viewer can hear the video and give them the words, usually in another language. Closed captions assume the viewer may not be able to hear it and add the non-speech audio (music, sound effects, who is speaking), and the viewer can switch them on or off.
The terms get mixed up constantly, and the mix-up is partly regional: in the UK, "subtitles" often covers both. This guide separates the terms, explains open captions and SDH (the two labels that confuse people most), and ends with a short rule for which one to use on YouTube, TikTok, Instagram and other places your video lives.
Pick the right one by what you need
| If you need... | Use | Why |
|---|---|---|
| Your video understood in another language | Subtitles | They carry a translation of the dialogue for viewers who can hear but do not speak the language. |
| Deaf and hard-of-hearing viewers to get the full audio | Closed captions or SDH | They include speaker labels, sound effects and music cues, not just dialogue. |
| Text that always shows, on any player | Open captions (burned in) | The text is part of the picture, so no player has to support a caption track. |
| Viewers to choose whether to see text | Closed captions (a caption file) | The viewer toggles the track with the CC button. |
| Captions on a streaming service or disc that has no broadcast caption data | SDH | SDH is a subtitle track written to do the job of captions. |
Upload a file or paste a YouTube link into Short.now's subtitle tool, edit the transcript, then burn the text into the video or export an SRT/VTT file. 3 free videos a month, no watermark, no credit card.
What subtitles are
Subtitles are a text version of the dialogue. According to the W3C Web Accessibility Initiative, subtitles are for spoken audio translated into another language. The viewer hears the original audio and reads the translation.
Because the viewer can hear, subtitles normally leave out everything that is not speech. 3Play Media, a captioning vendor, describes subtitles as dialogue only, which usually omits sound effects and other audio. A foreign-language film with English subtitles is the classic example: you hear the footsteps and the music yourself, so nobody writes them down.
The regional wrinkle matters when you search or brief a freelancer. The W3C notes that some regions use "subtitles" for both same-language and translated text, and 3Play says the same about the UK. In the US and Canada the two words are kept apart. If a UK client asks for "subtitles" on an English video with English speech, they may mean what a US client would call captions.
What closed captions are
The W3C defines captions as a text version of the speech and the non-speech audio information needed to understand the content. The FCC's description is similar: closed captioning displays the audio portions of television and video programming as printed words, and captions can also identify speakers, sound effects, background music and laughter.
So the practical test is simple. If the text tells you that a door slammed, music started or "Maria" is talking off-screen, it is a caption, not a plain subtitle.
Why "closed"?
The word refers to visibility, not content. The FCC explains that closed captions are hidden as encoded data in the TV signal and appear only when the viewer activates them. Open captions, in contrast, always appear on the screen. The same logic carried over to online video: a caption track you can toggle with a CC button is "closed", even though nothing is literally encoded in a broadcast signal.
What good captions contain
The W3C is candid that the non-speech part is a judgment call: it is not always clear which non-speech audio to include, and it calls high-quality caption writing more art than science. In practice, caption writers cover:
- Speech, word for word, synchronized with the audio.
- Speaker changes, especially when the speaker is off-screen.
- Sounds that carry meaning, such as [applause], [door slams] or [phone rings]. YouTube's own help page uses [applause] as its example of a sound cue to type in.
- Music that sets the mood or is part of the story.
Open captions versus closed captions
Both of these are captions. The difference is whether the viewer controls them.
| Closed captions | Open captions | |
|---|---|---|
| Viewer can turn them off | Yes | No |
| Where the text lives | A separate track or file (for example SRT or VTT) | Rendered into the video frames |
| Needs player support | Yes | No |
| Styling | Mostly decided by the player, though some formats carry position and style | Fully decided by you |
| Changing a typo later | Edit the file and re-upload it | Re-export the video |
| Good fit | YouTube long-form, web players, anything where viewers need control | Social clips that may autoplay muted, or reposts to places without a caption track |
The W3C notes that most captions are closed captions, and that most web videos use the player's default caption style, usually white text in a black box. Authors can position and style captions, but support across browsers and media players is inconsistent and sometimes unreliable. That inconsistency is the main reason short-form creators burn their captions in: it is the only way to guarantee the text looks the way you designed it.
The trade-off is the mirror image. A burned-in caption can never be turned off, never be translated by the viewer, and never be read by a system that only looks at caption tracks. If you want both a styled look and an accessible track, publish both: burn the text in for the social versions, and upload a caption file where the platform accepts one.
Where SDH fits
SDH stands for subtitles for the deaf and hard of hearing. It is a subtitle track that includes dialogue plus non-dialogue information such as sound effects, music and speaker identification. 3Play's description is that SDH assumes the end user cannot hear the dialogue.
It exists because not every platform carries caption data. 3Play notes that SDH often emulates closed captions on media that does not support them, and lists streaming, OTT, HDMI and disc platforms as typical homes for it. 3Play adds that SDH can be translated into other languages, while translation options for closed captions are limited.
Here is how the three compare on the points that matter:
| Subtitles | SDH | Closed captions | |
|---|---|---|---|
| Dialogue | Yes | Yes | Yes |
| Sound effects and speaker IDs | Usually no | Yes | Yes |
| Main audience (US/Canada) | Hearing viewers who do not understand the language | Deaf and hard-of-hearing viewers | Deaf and hard-of-hearing viewers |
| Common home | Foreign-language and translated content | Streaming, OTT, HDMI and disc | Broadcast TV and most video programming in the US and Canada |
3Play's own conclusion is worth keeping in mind: neither closed captions nor SDH fits neatly into one box, and when a title offers both, the visible difference is usually styling. Treat the categories as a guide to intent, not as a rigid standard.
The files behind captions and subtitles
If you upload a caption file rather than burning text in, the file format decides what you can control. YouTube's help page lists many formats and notes what each one supports:
- SRT (.srt) and SBV (.sbv): basic timing only. Style markup is not recognized, and the file needs to be plain UTF-8.
- WebVTT (.vtt): supports positioning. Styling is limited to bold, italic and underline.
- TTML and DFXP: support both styling and positioning, though YouTube's TTML support is partial.
- SCC (.scc): YouTube's preferred format when your captions are based on CEA-608 features, because it stores an exact representation of that data.
For most creators, SRT is the safe default and WebVTT is the upgrade when you need captions out of the way of on-screen text. Short.now's tool pages say you can export SRT/VTT files as well as burn the text in.
How captions work on YouTube
YouTube uses the terms "subtitles" and "captions" together on its help pages and does not distinguish them. What it does give you is several ways to add a track in YouTube Studio:
- Upload a file, either with timing or without timing.
- Auto-sync: paste or upload a transcript in the spoken language and YouTube syncs it to the video. YouTube does not recommend transcripts for videos over an hour long or with poor audio quality.
- Type along: type or paste the transcript while the video plays, and YouTube sets the timings. Include sound cues such as [applause].
YouTube also uses speech recognition to create captions automatically. When they are available, they are published on the video automatically, but only in the video's default language.
On the viewer side, on desktop you click the Closed captions (CC) button at the bottom right of the player to turn captions on, and click it again to turn them off. To change how they look, open the Settings gear, choose Subtitles/CC, then Options. You can change font, color, opacity and size, background and window color, and character edge style.
Which one to use for short-form video
For Shorts, Reels and TikTok, the video is usually watched in a feed where captions need to be there the moment the clip starts. That makes burned-in (open) captions the practical default: they show everywhere, they match your branding, and they do not depend on a viewer finding a menu. We walk through the steps in how to add captions to videos automatically, and the TikTok-specific controls in TikTok captions: how to add, edit and turn them on.
Use a caption or subtitle file when the platform supports one and the viewer benefits from control: long YouTube videos, courses, webinars and anything hosted on your own site. A separate track also keeps the text editable, so a typo fix does not mean re-exporting the video.
Whichever route you take, remember the audience. If part of your audience is deaf or hard of hearing, plain dialogue subtitles leave out information they need. Add the sound cues that change the meaning of a scene, and label speakers when more than one person talks.
Check the safe zone before you burn text in
Burned-in text sits inside the video frame, and each app places its own interface elements over parts of the frame. Before you commit to a caption position, check it against the platform's safe zone. Short.now has safe-zone pages for TikTok, Instagram Reels and YouTube Shorts.
A workflow that covers both
You do not have to pick one for good. This is a workable routine for a creator who posts the same video in several places:
- Get a transcript. Generate it automatically, then read it start to finish. The W3C is blunt that automatic captions are not sufficient unless they are confirmed to be fully accurate, and that they usually need significant editing. Its example: leaving out one word such as "not" can make the caption contradict the audio.
- Fix names and numbers first. Proper nouns, brand names and figures are where automatic transcripts go wrong most often.
- Decide which sounds matter. If a sound changes the meaning of a moment, add a cue for it. If it is background texture, leave it out.
- Export two versions. One with open captions for feeds and reposts, and one caption file (SRT or VTT) for YouTube and your own site.
- Check on a phone. Watch the final export with the sound off and confirm the text is readable and not covered by interface elements.
The slow part is the editing pass, and it is where a tool earns its place. Short.now's AI caption generator creates word-by-word captions with word-level timestamps, lets you edit every line, and exports burned-in video or SRT/VTT. If you only need the text, video-to-text transcription gives you a transcript to start from.
One honest limit: Short.now's subtitle tool page says it generates subtitles focused on speech, which is what most social videos need. If you need full closed captions with sound cues, use the editor to add them as you review each line, and read your result against the checklist above.
Common mistakes
- Calling burned-in text "closed captions". If the viewer cannot turn it off, it is open, whatever the tool's menu calls it.
- Publishing automatic captions unchecked. Accuracy problems change meaning, not just spelling.
- Using a translation where captions are needed. A translated subtitle track will not tell a deaf viewer about the music or the off-screen voice.
- Burning text in without keeping a source file. Once text is part of the picture, a typo means a new export. Keep the editable transcript or caption file.
- Ignoring position. The W3C notes that position and style support differs across players, so test the track in the player your audience actually uses.
- Assuming the default style is accessible. The default white-on-black is common, but contrast, size and placement are still your responsibility when you control the look.
Generate captions in Short.now, fix any line in the editor, and export a burned-in video for social feeds plus an SRT/VTT file for YouTube. Free to start: 3 videos a month, no watermark, no credit card.
Frequently Asked Questions
Open captions are captions that are always displayed on the screen and cannot be turned off. In digital video they are usually burned into the picture, which means they show on any player and in any feed, but you cannot change their language or hide them without re-exporting the video.
Normal subtitles give the dialogue, usually translated, for viewers who can hear the audio. Closed captions give the dialogue plus the non-speech audio, such as sound effects, music and speaker labels, and the viewer can turn them on or off. In the UK the word subtitles is often used for both.
On YouTube for desktop, click the Closed captions (CC) button at the bottom right of the player, and click it again to turn captions off. To change their look, open the Settings gear, choose Subtitles/CC, then Options. Other apps and TVs put the control in different menus, so check the player's settings.
Because the text is hidden until the viewer activates it. The FCC explains that closed captions are encoded data in the TV signal that appear only when switched on, while open captions are always on screen. Online players keep the same idea with a CC button that shows or hides a caption track.
Not without checking them. The W3C says automatic captions are not sufficient unless they are confirmed to be fully accurate, and that they usually need significant editing. A single missing word such as not can reverse the meaning, so read the whole transcript before you publish.
Ready to stop editing and start publishing?
Short.now turns your long videos into scroll-stopping shorts automatically — in under 2 minutes.
Get started free 3 free videos every month · No credit card needed