Feature guides

Smart Pre-editing

Learn how pause and filler-word detection works, and how to keep edits sounding natural.

Updated 7/31/2026

Smart Pre-editing analyzes speech and pauses, then creates suggestions you can review on the timeline. It works best for spoken-word recordings, interviews, and podcasts.

What it can detect

  • Long pauses that do not add meaning
  • Obvious filler words and repeated phrases
  • Mistakes that the speaker immediately corrects
  • Silence at the beginning or end of a recording
  1. Let transcription and waveform analysis finish.
  2. Run detection once with the default sensitivity.
  3. Review suggestions around sentence boundaries first.
  4. Keep breaths and pauses that support the speaker’s delivery.
  5. Apply the edits only after listening.

Keep the result natural

Do not remove every pause. Short pauses help listeners understand ideas and preserve the speaker’s tone. In conversations, pay extra attention to interruptions, laughter, and fast handoffs.

Background noise can make voice detection less accurate. A consistent recording level and a quiet room will produce better results.