How YouTube’s automatic captions work on Shorts
By the maker of Captions Cut. Facts checked on from the pages listed under Sources.
The short answer
YouTube makes automatic captions with speech recognition on long videos and Shorts, in 67 languages, from the video’s default language, and publishes them when they are ready, which may be some time after upload. Fix them in YouTube Studio under Subtitles, or upload your own SRT or VTT file. Viewers switch them on with the CC button.
How they work
- Speech recognition, in 67 languages from Afrikaans to Zulu; only the default audio track’s language.
- “Automatic captions may not be ready at the time that you upload a video,” YouTube says; they publish themselves when done.
- Possibly inappropriate words show as [ __ ] unless you turn that filter off in YouTube Studio.
- English gets “expressive captions”: loud speech in capitals, and sounds like sighs and gasps.
Why they fail
YouTube lists the reasons: an unsupported language, a very long video, poor audio, a long silence at the start, and overlapping speakers or mixed languages.
Fixing them, or bringing your own
- YouTube Studio › Subtitles, choose the video, then Duplicate and edit the automatic track.
- Or upload a file, with or without timing: SubRip (.srt), WebVTT (.vtt) and other formats.
- Captions Cut saves an SRT timed to its export, so the same captions can be burned into the Short and uploaded as a track.
Questions
How long do YouTube automatic captions take?
YouTube gives no figure: it depends on how complex the audio is, and captions may not be ready when the upload finishes.
Sources
- YouTube: automatic captions
- YouTube: edit or remove captions
- YouTube: add subtitles
- YouTube: supported caption files
- YouTube: viewer caption settings
Each page read on .