Automatic pause removal
Remove dead air from screen recordings and talking videos
Scan the audio for long quiet sections, keep configurable padding around speech, and cut video and audio together so tutorials, lectures, meetings, and podcasts stay in sync.
- Cuts video and audio on the same speech ranges.
- Adjustable silence threshold, minimum pause, and edge padding.
- Runs locally with no upload, account, watermark, or time quota.
Tighten tutorials and screen recordings automatically
Screen recordings often contain pauses while a page loads, a presenter finds the next control, or a command finishes. Those gaps make a short explanation feel slow, but finding every pause manually takes longer than the recording itself.
AntiUpload detects quiet ranges from the audio track and applies the same cuts to the picture. The cursor movement, screen action, and narration remain synchronized in the exported MP4.
Use threshold, minimum silence, and padding together
The threshold decides how quiet audio must be before it counts as silence. The minimum duration protects normal conversational beats, and padding keeps a little sound on both sides of every cut so the beginning or end of a word is not clipped.
For a clean microphone, the defaults are a practical starting point. Fan noise, keyboard noise, and room ambience may require a higher threshold. Increase the minimum silence or padding when the result feels too aggressive.
Know when energy-based detection is the wrong tool
This is an energy-based detector, not a speech-understanding model. It works best for voice-led recordings without background music. A quiet music passage may be classified as silence if it falls below the threshold.
The output must be re-encoded because frames from multiple time ranges are being joined. For a single known start-and-end cut, use Video Trimmer instead; it can often preserve the original stream.
How to use this workflow
- 01.Choose the recordingAdd a screen recording, lecture, meeting, podcast, voice memo, or other spoken audio or video file.
- 02.Tune silence detectionStart with -30 dB, a 0.5-second minimum silence, and 0.15 seconds of padding. Use a higher threshold for noisy rooms and more padding if words sound clipped.
- 03.Remove pauses and reviewThe first pass finds quiet ranges and the second joins the remaining speech. Play the output to make sure the pacing still sounds natural.