Creator tools / 001
PauseCut
Keep the voice. Lose the wait.
PauseCutEditable video pause removalConcept artwork. PauseCut is a command-line tool that exports an editable timeline.
Your first cut, without the dead-air cleanup.
You’ve finished recording. Now comes the slow part: finding the gaps, trimming the pauses, and checking that you haven’t taken half a word with them.
PauseCut handles that first pass. It turns your recording into an editable timeline with the dead air trimmed away, ready for you to shape into the finished video. It’s the tool I use in my own editing workflow, refined around the recordings that exposed the rough edges.
Why the details matter
Quiet isn’t always silence.
The start of “this.” The tail of “yes.” A breath between thoughts. Speech can dip below a silence detector’s cutoff while there’s still something worth hearing. Cut every quiet patch and you can lose consonants, clip word endings, or make a calm explanation feel rushed.
Finding a pause is one decision.
Choosing where to cut is another.
PauseCut pays attention to the sound around the edit, the level of the recording, and the space speech needs to feel natural. The result is a more considered starting point—with every edit still under your control.
Small tool. Thoughtful details.
More care around every word
Cut placement accounts for quieter starts and trailing sounds, with breathing room around the speech you keep.
Adapts to your recording
Compensates for low recording levels and takes room noise into account. Fine-tune the pacing when your footage needs it.
An editable timeline. Original media.
Exports an FCPXML cut list without re-encoding your video. Adjust the cuts in your editor and render when you’re ready.
A second look at the cuts
Checks cut boundaries and reports ones that may need attention, so your final review has a useful starting point.
Retake claps become edit points
Adds a cut after detected retake claps. The clap and take stay available for you to review and remove yourself.
Long recordings, smaller timelines
Optionally divide the result into manageable parts, favoring longer pauses as break points. Cached analysis makes repeat runs easier.
From recording to a timeline you can edit.
- 01
Set up once.
Follow the included guide to install the local prerequisites. Choose your recording.
- 02
Let PauseCut make the first pass.
Run the script to analyze the audio and create the cut timeline beside your source file.
- 03
Import. Review. Make it yours.
Open the FCPXML timeline in DaVinci Resolve, check the joins, and finish the edit. Your original media stays intact.
Before you buy
- Comfort using a terminal; Python 3.10+ and FFmpeg/ffprobe on your PATH. The guide covers setup with uv.
- A video editor that imports FCPXML 1.8 and supports your source media. DaVinci Resolve is the documented target.
- One source video with attached audio. Keep the source accessible for timeline relinking.
- You’re buying a Python script and documentation. Review the generated cuts before your final export.
A few things to know.
Does it work outside DaVinci Resolve?
PauseCut exports FCPXML 1.8, an interchange format rather than a Resolve project file. DaVinci Resolve is the established workflow for this script. Other editors, including Final Cut Pro, may be able to use the output when they support this FCPXML version and your source media. Those imports have not been verified for this product. Check your editor’s version and codec support before buying; generic XML support is not enough.
Is this an app, a plugin, or a script?
It is a downloadable Python command-line script with a setup guide. You run it on your own computer with Python 3.10+ and FFmpeg/ffprobe installed. There is no desktop interface or editor plugin. The guide includes a uv-based setup path and a Windows launcher.
Will it export a finished MP4?
It creates an editable FCPXML timeline that references your original recording. Import that timeline, review the cuts, make your creative edits, and export the final video from your editor. The pause-removal step does not re-encode or modify the source video. Processing time depends on the recording and your computer.
Does it remove every mistake or repeated phrase?
The main job is dead-air removal. Fillers, stumbles, and ordinary spoken retakes still need your judgment. An optional advanced mode can report or trim certain near-identical playback false starts, such as a clip opening captured twice. It is off by default and is not a general speech-rewrite feature.
What recordings is it best suited to?
Speech-led videos: talking-head recordings, tutorials, screencasts, and livestream recordings with audible gaps. This edition uses one video with its attached audio, not a separate multitrack capture setup. Constant background music, heavy noise, or unusually quiet speech can require tuning and manual review. Windows is the documented workflow; macOS and Linux have not been verified for this release.
Do I need an AI subscription or have to upload my footage?
No. Audio analysis runs locally, without transcription or a paid AI API. Your recording stays on your computer. Internet access may be needed to install the tools and dependencies.