How to make jump cuts: by hand, automatically, or from a transcript

To make a jump cut, remove a stretch from the middle of a single continuous shot and join what is left: the picture jumps forward in time. In a talking-head video you do that dozens of times, wherever there is a pause, a stumble or a repeated sentence. You can do it by hand on any timeline, let an automatic jump cut editor remove the silences for you, or cut from a transcript by deleting words. Which one is fastest depends on what you are removing.

If you want the definition and history first, see what is a jump cut and why the jump cut won.

Method 1: by hand on a timeline

Every editor can do this, from iMovie and CapCut on a phone to Premiere and DaVinci Resolve on a computer.

1. Play the take until you reach something you want to remove.
2. Move the playhead to the exact start of it and split (also called blade or razor).
3. Move to the end of it and split again.
4. Delete the piece between and close the gap (in many editors this is called a ripple delete).
5. Play across the cut to check it.

Upside: total control over every cut. Downside: time. You have to hear every problem in real time, and each cut needs two precise playhead positions. Zooming into the timeline and turning on the audio waveform helps you see where words start and stop. A ten-minute take with a cut every ten seconds is 60 rounds of this.

Method 2: automatic jump cut editors

Automatic tools detect silence, meaning audio below a volume threshold for longer than a minimum duration, and cut it out with a little padding on either side. Some also detect filler words or repeated takes.

Tool	Where it runs	What it removes
DaVinci Resolve	Desktop	Silence, via Ripple Delete Silence with threshold and padding settings (DaVinci Resolve Club)
Adobe Premiere	Desktop	Pauses and filler words through Text-Based Editing (Adobe)
AutoCut	Plugin for Premiere Pro, also DaVinci Resolve	Silence with adjustable minimum duration and sensitivity, plus other automation (Digital Production)
TimeBolt	Standalone desktop app	Silence and filler words; exports to Premiere, Final Cut, Resolve and Camtasia (AlternativeTo)
Gling	Desktop, for YouTubers	Bad takes, silences and filler words; integrates with Final Cut, Resolve and Premiere (Gling)
Descript	Desktop, browser	Filler words and gaps from the transcript (Descript Help)

Upside: speed. An hour of footage gets a first rough cut in minutes. Downside: every silence is treated the same. A pause before your punchline goes just like the pause where you lost your thread, and a fully automatic pass can make a calm explainer sound breathless. The settings that matter (threshold, minimum length, padding) are explained in how to remove silence from a video.

Silence detection also cannot see a repeated sentence. If you said something twice without a pause in between, both versions stay.

Method 3: from a transcript

The third approach treats the take as text. The software transcribes it, and deleting words from the transcript deletes the footage under them. Each deletion is a jump cut.

This is how Descript and Premiere’s Text-Based Editing work on the desktop, and how slicecut works on iPhone and Android:

1. Import the take. It becomes a transcript.
2. Tap a word to strike it, or drag across a sentence to strike the whole line. Restarts and repeated sentences are easy to spot because you can read them.
3. The cut lands right before the next word, so there is no frame to hunt for.
4. Set the pauses. One dial sets every pause to a target length, which gives the cuts a consistent rhythm. Pin the pauses that should stay long.
5. Export. Nothing is rendered until then, so every strike can be undone with a tap.

Upside: you see everything you said, so you can cut what silence detection misses, and you decide each cut yourself. Downside: it is built for speech. For cutting a silent action shot, use a timeline.

Which method for which job

What you want to remove	Fastest method
A handful of cuts in a short clip	By hand on a timeline
Dead air in long recordings, rough cut	Automatic silence removal, then review
Restarts, repeats and weak sentences	Transcript editing
Exact control over every pause	Transcript editing with a pause target

How to make jump cuts look intentional

A jump cut looks like a choice when the rhythm is consistent and the jumps are clean. A few rules:

• Cut between sentences or clauses, not in the middle of a word.
• Keep the gaps even. Random gap lengths make cuts feel accidental. The median pause between sentences in the talking-head videos we measured was about 0.35 seconds (How long should a pause be?).
• Keep the framing still. A locked-off phone makes the jump read as a time skip. A moving camera makes it read as a glitch.
• Cut on a closed mouth or a blink when you have the choice.
• Disguise the biggest jumps with a punch-in. Cutting to a tighter framing of the same shot at a jump turns a stutter into emphasis. Use it a few times per minute, not at every cut.
• Watch once with the sound off. If a cut looks jarring without audio, move it a few frames.

A practical workflow

For most talking-head videos, combine the methods: cut the content and pauses from the transcript, where you can see what was said, then open the clean cut in a timeline editor for punch-ins, b-roll and captions. The full order of operations is in how to edit a talking-head video.