A text version of a video's audio — dialogue plus non-speech sound — on a separate track the viewer can toggle on or off, unlike burned-in open captions.
Last verified · 2026-09-17 · by Moe Ameen
Closed captions are a text version of everything audible in a video — the spoken dialogue plus the non-speech audio a deaf or hard-of-hearing viewer would otherwise miss, such as music cues, sound effects, and speaker identification. The defining word is "closed": the text rides on a separate track the viewer can turn on or off, almost always via a button labeled "CC." It is data the player overlays on demand, not pixels painted into the picture, which is exactly what makes it toggleable, restyleable, translatable, and readable by search engines.
That toggle is the whole distinction from open captions, which are burned directly into the video frame and cannot be removed. It is also what separates closed captions from subtitles: subtitles assume the viewer can hear but not follow the language, so they carry dialogue only; captions assume the viewer cannot hear at all, so they add the non-speech information. SDH — subtitles for the deaf and hard of hearing — is the hybrid you see on streaming services and discs: caption-style content delivered through the subtitle mechanism.
Technically, a closed-caption track is encoded and carried according to where the video runs. Broadcast and cable in North America use CEA-608 (the older line-21 analog standard) and CEA-708 (the digital-TV standard with fonts, colors, and flexible placement). The web uses timed sidecar files — WebVTT (.vtt) natively in HTML5, the ubiquitous SubRip (.srt), and TTML/IMSC in broadcast-derived streaming. When a platform "auto-captions," it runs the audio through automatic speech recognition and emits one of these tracks; that output is a genuine closed-caption file, and also frequently wrong until it is proofed.
For creators the term is a small trap. In broadcast and accessibility usage, "closed captions" means this toggleable track. But most short-form social video ships open captions — burned-in words in the vertical safe zone — because muted autoplay, cross-posting, and full styling control all favor pixels over a track the platform may not even display. The accessibility-standard name is closed captions; the thing most creators actually publish to the feed is open captions.
Closed captioning began on broadcast television, not on the internet. In the US it grew out of 1970s–80s accessibility work and was cemented by the Television Decoder Circuitry Act of 1990, which required caption-decoding hardware in TV sets, encoded on line 21 of the analog signal — the origin of the CEA-608 standard. The move to digital television brought CEA-708, which lifted the styling and positioning limits of the analog era.
The obligation followed content online. The Twenty-First Century Communications and Video Accessibility Act (CVAA), enforced through FCC rules, extended caption requirements to internet video that had previously aired on US television, so captions could not simply be stripped when a clip moved to the web. The Americans with Disabilities Act and Section 508 created separate obligations for businesses, public bodies, and federally funded organizations, and the FCC set quality rules requiring captions to be accurate, synchronized, complete, and properly placed. Outside the US, the European Accessibility Act became binding on 28 June 2025, widening accessible-media requirements — captions among them — across EU member states.
The creator-driven, burned-in aesthetic is a separate lineage. Animated, brand-fonted, word-synced captions emerged from TikTok's CapCut ecosystem around 2021 and hardened into a default expectation on every short-form platform — so the mainstream creator experience of "captions" is open captions, even as the underlying accessibility standard remains closed captions.
| Platform | Behavior |
|---|---|
| YouTube | Serves a genuine closed-caption track behind the CC button, auto-translates it, and indexes the text for search. Uploading an accurate .srt/.vtt is pure upside — and Shorts can still carry burned-in open captions on top. |
| TikTok | Has an auto-caption feature that produces a toggleable track, but in practice creators burn in open captions for styling and placement control. Muted autoplay makes visible-by-default text essential; safe-zone placement (middle band) keeps it clear of UI chrome. |
| Instagram Reels | Offers auto-generated captions, but burned-in open captions dominate for the same reasons as TikTok: muted feed, cross-posting durability, and identical rendering everywhere. Closed captions matter more on longer in-feed video. |
| Broadcast / OTT streaming | Uses CEA-608/708 or subtitle-delivered SDH, and is where legal caption requirements bite hardest. Quality rules (accurate, synchronized, complete) apply — raw ASR output generally does not meet them without human review. |
| LinkedIn / X | Video autoplays muted in-feed and rewards burned-in captions; native caption-track support is inconsistent, so open captions are the reliable choice for reach. |
The useful mental model is that "closed captions" is an accessibility and broadcast term that creators inherited and then quietly repurposed. The strict meaning — a toggleable, machine-readable track — is real and matters enormously for YouTube search, for translation, and for legal compliance. But the thing that actually drives short-form reach is the opposite mechanism: open, burned-in captions that play by default on a muted feed and survive every re-upload. The mistake I see most is treating this as an either/or. It is not. On a Short or a Reel you burn the captions in. On the YouTube upload of the same content you also attach a real caption file, because that track is indexable text and the burned-in pixels are not.
The part almost nobody gets right is that both jobs have to happen on every clip, at the volume a real posting cadence demands, and that is where hand-captioning in a desktop editor one file at a time falls apart. This is why an engine like Kompozy burns on-brand open captions by default into every video format that carries spoken dialogue — Persona Shorts, Clipped Shorts — in the vertical safe zone, in your brand font, rather than leaving it as a separate manual pass; the muted-feed case gets handled at generation time. The point of understanding the terminology is operational: know which caption form does which job, then automate the form your platforms actually reward instead of retyping subtitles into an editor after every render.
Closed captions are a text version of a video's audio — dialogue plus non-speech sound such as music, sound effects, and speaker identification — carried on a separate track the viewer can turn on or off, usually via a "CC" button. Because the text is data the player overlays rather than pixels burned into the picture, closed captions can be toggled, restyled, translated, and read by search engines.
Closed captions are a separate, toggleable track the viewer can switch on or off; open captions are burned directly into the video frame and are always visible. Open captions cannot be removed, which is exactly why short-form social creators favor them — they survive re-uploads, play by default on muted feeds, and render identically on every platform. Closed captions offer flexibility and search-indexability; open captions offer guaranteed visibility.
No. Subtitles assume the viewer can hear the audio but not understand the language, so they transcribe dialogue only, often translated. Closed captions assume the viewer cannot hear, so they also include non-speech information like sound effects and speaker labels. SDH (subtitles for the deaf and hard of hearing) is the hybrid — caption content delivered through the subtitle mechanism, common on streaming and discs.
They are the North American broadcast closed-captioning standards. CEA-608 is the older analog-era standard carried on line 21 of the TV signal, limited to a monospaced style and a few positions. CEA-708 is the digital-television standard, which adds fonts, colors, sizing, and flexible placement. Web video generally uses sidecar caption files like WebVTT (.vtt) or SubRip (.srt) instead.
Often, depending on the publisher and where the content runs. In the US, the CVAA and FCC rules require captions on internet video that previously aired on television, and the ADA and Section 508 create obligations for many businesses and public bodies. The EU's European Accessibility Act, binding since 28 June 2025, requires accessible audiovisual content. The rules also demand quality — accurate, synchronized, and complete — which raw auto-captions frequently are not.
On short-form social video — TikTok, Reels, Shorts — burned-in open captions almost always win, because viewers watch muted and the text must be visible by default and survive cross-posting. On YouTube and other long-form or on-demand surfaces, also supply a real closed-caption track, since it is served behind the CC button, translatable, and indexed for search. The strongest setup on important content is both.