Verbatim captioning
Captioning that captures every spoken word exactly, including filler words, false starts, and disfluencies — as opposed to clean-read captioning that smooths the speech.
Verbatim captioning preserves the speaker's actual speech: 'um', 'uh', 'you know', repeated words, sentence restarts. It's preferred for legal, medical, and broadcast captioning where the exact spoken record matters.
Clean-read captioning (or 'edited captioning') removes filler and corrects obvious slips for readability. It's preferred for event captioning, where the transcript is reused as content and readers benefit from cleaned text.
Most captioning platforms produce something between these poles — automated ASR doesn't transcribe every 'um' but does preserve substantive content faithfully. Editing the transcript post-event to clean filler is straightforward when needed.