Glossary
What is a subtitle?
By the KeyMo team · Updated
A subtitle is text shown on screen that gives the spoken words of a video, most often as a translation for people who can hear the audio but do not understand the language. Captions go further and add sounds, and a transcript is the same words kept separate from the video.
Subtitle meaning in video
The HTML Standard defines subtitles as a transcription or translation of dialogue, for when sound is available but not understood. Captions are defined as dialogue plus sound effects, music cues and other audio information, for when sound is unavailable or unclear. The W3C accessibility pages use the same split: captions for same-language audio, subtitles for translated speech.
In daily use the words blur. W3C notes that some regions use 'subtitles' for both, and WCAG notes that captions are called subtitles in some countries. If you hear 'add subtitles' on a social video, people usually mean words on screen.
Subtitle vs caption vs transcript
| Subtitle | Caption | Transcript | |
|---|---|---|---|
| Gives | Spoken words, often translated | Words plus sounds and speaker cues | Words as a separate text |
| Assumes viewer | Can hear, does not follow the language | Cannot hear, or has sound off | Wants to read, search or quote |
| Lives | On the video, or as a track | On the video, or as a track | Beside the video |
| Accessibility | W3C: other-language subtitles are not directly an accommodation | Counts as an accessibility accommodation | WCAG's route for audio-only content |
For the longer comparison, see captions vs subtitles, caption and transcript.
Which to use on a podcast clip
- Same language as the speaker, watched with the sound off: captions, burned in, ideally word by word.
- Viewers in another language: a translated subtitle track, or a second version of the clip.
- Someone who needs the full text: a transcript, kept apart from the clip.
- A platform that wants a caption file: an SRT file or a VTT file.
KeyMo draws captions into the picture in 12 styles, word by word, or exports without them. That covers the first line above. It is not a translation workflow, and you should not describe it as one.
Making, editing and styling subtitles
A subtitle writer or subtitle editor does three jobs: get the words right, break them into short lines, and time each line to the speech. A subtitle generator does the first and third automatically with speech recognition. The result is a draft. W3C says automatic captions usually need significant editing, and YouTube says to review and edit them.
- Keep lines short. W3C's common-practice tip is one or two lines, generally under 32 characters each.
- Start a new sentence on a new line, and break long sentences at a logical phrase.
- Do not censor, adapt or add to the speech.
The subtitle background is the box or outline behind the text. In most web video the player default is white characters in a black box. A solid or semi-transparent box helps when the picture behind is busy. WCAG's general text rule asks for a 4.5:1 contrast ratio, though the pages we read do not say it applies to burned-in captions.
Common mistakes
- Calling a translation a caption and promising access that it does not give.
- Covering a speaker's mouth or key on-screen text with subtitle lines.
- Leaving machine-written subtitles unchecked.
- Burning in a language you may need to change later. A separate track is easier to swap.
Sources: HTML Standard: the track element (checked 11 October 2026); W3C WAI: Captions/Subtitles (checked 11 October 2026); W3C WAI: Transcribing Audio to Text (checked 11 October 2026); WCAG 2.2 (checked 11 October 2026); YouTube Help: Use automatic captioning (checked 11 October 2026).