Subtitles usually render spoken dialogue as text, often for viewers who can hear the audio but do not understand its language. Closed captions are designed to make audio information available in text, including dialogue and relevant sounds, for viewers who are deaf or hard of hearing. The labels and conventions vary by platform and region, but the central distinction is what information the text carries: dialogue alone, or the meaningful audio experience more broadly.
What each text track tells the viewer
A subtitle track commonly includes dialogue and may identify a speaker when needed. For a foreign-language film, it lets a viewer follow what characters say while hearing the original performances. Some subtitle styles also include sound cues, especially when those cues are important to the plot or emotion.
Closed captions generally include spoken words plus relevant non-speech audio, such as a door slamming, a phone ringing, or music changing. They can identify who is speaking and indicate delivery when that information matters. A caption track does not need to describe every sound. It should help viewers understand the scene’s meaningful audio.
The labels are useful, but not universal
Different broadcasters, platforms, and countries use terms and styling conventions differently. Some services use “captions” broadly; others distinguish open captions, burned into the image, from closed captions that can be switched on or off. Some subtitle tracks include sound cues, while some captions may focus heavily on dialogue. Check the actual track and intended audience instead of relying on its name alone.
The practical question is whether viewers can access the information they need. W3C describes localization as adapting content to language and locale requirements, which can include more than translating words [1]. A video intended for a multilingual audience may need subtitle translations, while an accessible version may need captions that represent important sounds. One track can sometimes serve both purposes, but only if it meets both needs.
Worked example: a quiet kitchen scene
Imagine an original scene: a character whispers, “The kettle is still on,” while a timer beeps twice and the lights go out. A subtitle for viewers who understand the audio might show only the spoken line. A caption track could render the whisper and add a concise cue such as “TIMER BEEPS” followed by “LIGHTS CLICK OFF,” if those sounds shape what the viewer understands. If the beeps reveal that food is ready or the blackout signals a power failure, omitting them changes the scene.
For a short exercise, take a 20-second clip and list every sound that changes the plot, identifies a speaker, or establishes a meaningful atmosphere. Draft a dialogue subtitle track, then a caption version. Compare them: which information belongs in both, and what audio detail would be invisible in dialogue-only text? Keep the cues brief enough to read alongside the image.
Timing and readability shape the result
Text must stay on screen long enough to read and should appear close to the speech or sound it represents. Break lines at natural phrase boundaries where possible. Avoid covering a face, a sign that matters, or another essential part of the frame. The viewer is dividing attention among text, image, and action, so a technically complete track can still be hard to use if it is crowded or mistimed.
Sound labels should be selective and clear. A caption such as “MUSIC” may be too vague when a shift from a lullaby to a warning tone matters. A concise description such as “MUSIC TURNS OMINOUS” can convey the change without trying to reproduce every note. Descriptions should report what the audio communicates, not invent a visual event or explain the story beyond the evidence.
Choose the track from the audience outward
For a translated film, ask which languages viewers need and whether they want to hear the original dialogue. For accessibility, ask whether the track communicates dialogue, speakers, and relevant sound. For a classroom or public presentation, check whether captions can be enabled and whether any onscreen text also needs translation. These are separate requirements that may call for separate tracks.
A useful production brief names the audience, the language, whether sound cues are required, how speakers are identified, and who will review timing and accuracy. That prevents a track labeled “subtitles” from being mistaken for a complete access version, or captions from being assumed to provide translation.
