What is Subtitles vs Closed Captions?

Both are timed text tracks that ride alongside your video. The player reads the file, then paints the text at the start and end timestamps you wrote. The difference is what gets written into those timestamps.

Subtitles assume the viewer can hear the audio. They render spoken dialogue, plus optional forced narrative text for non-speaker context. Translated subtitles assume the viewer cannot understand the original language, so the text is a translation of the speech.

Closed captions assume the viewer cannot hear the audio at all. They include dialogue, speaker identifiers, music cues like [tense music], and sound effects like [door slams]. Captions are the legally recognized accessibility track in the US under FCC rules for OTT video.

On the file side, both use the same sidecar formats such as WebVTT and SRT. The .vtt or .srt extension does not decide the role; the operator does. Player APIs expose a text track list, and your CMS maps a language code plus a role label (subtitles or captions) to each file.

For live streams, captions arrive as a stream of cues from a human captioner or a workflow, not as a finished file. For VOD, you upload finished files with the video and switch them on at the player.

Two operator concerns sit behind this distinction, and they pull in different directions.

First is compliance. In the US, FCC rules require captions on most OTT programming delivered to consumers. Miss it and a show can be pulled, a complaint can land, or a distributor can demand a fix. Treat captions as a release blocker for any VOD title that originally aired on US TV.

Second is reach. Subtitles, especially translated ones, expand your addressable market. A Korean drama with English and Spanish subs opens in markets the audio track cannot serve alone. Subtitles also lift comprehension for second-language viewers on any title.

The practical split: captions protect you from compliance risk; subtitles grow your audience. Most mid-tier catalogs ship both, one caption file for the original language and 5 to 20 subtitle files for translations. Live programming needs an injected caption feed, not a file. Plan for both tracks in your CMS, not as an afterthought in the player.

Use this table when you are deciding which track to author for a given title or audience.

Treating them as the same thing in the CMS. They live in different role buckets, even if the file format is identical.

Forgetting the role label. Players expose a text track list. If you tag everything as "subtitles", deaf users may never find the captions track.

Captioning only the audio and dropping sound cues. A door slam or a gunshot is part of the story for a viewer who cannot hear it.

Translating captions instead of subtitles. Translating captions into ten tongues is a localization job, and it bloats the cost. Subtitle files are the translation surface, not captions.

Skipping forced narrative subtitles. Fantasy and sci-fi titles need untranslated terms like a place name rendered on screen. If you have no forced-subtitle track, the localization falls apart.

Ignoring live. VOD caption files do not help a 24x7 channel. You need an injected cue stream.