Choose a moment with its own beginning and ending

Start with a recording you own or have permission to edit. An interview, tutorial or webinar can contain several useful clips, but the most interesting sentence is not always a complete story. Look for a question, an explanation and a payoff that make sense together.

For a first pass, choose a focused passage around 15 to 60 seconds long. This is an editorial starting point, not a platform limit. If the speaker needs another sentence to explain the point, keep it. A complete idea matters more than hitting an arbitrary duration.

Use the transcript to find candidates

Open the ClippingShorts video clipping studio and import your original MP4, MOV or WebM file. Generate captions locally or import an SRT/VTT transcript. Browser support depends on the codecs inside your file, so try a short section first.

Choose Find highlights to request transcript-based AI suggestions. You can describe a topic such as practical advice or a surprising answer. Cloud mode sends transcript text to Cloudflare; local keyword suggestions avoid that request. Read each candidate in context before deciding to use it.

Trim for context, then for pace

Play a candidate from its first frame. If it begins with “that” or “as I said earlier,” move the start back until a new viewer can understand the subject. Remove a long lead-in only when the remaining opening still represents the speaker accurately.

Set the IN and OUT points, apply the trim and replay the result. Let the final thought finish. Avoid using a cut that changes the meaning of an answer or removes a qualification. Duplicate a clip when comparing two openings so you can keep both edits.

Reframe a horizontal video for a vertical feed

Use 9:16 for a portrait composition, 1:1 for a square version or 16:9 to retain a landscape layout. Check the full clip after reframing: a crop that works at the beginning can hide a person or a slide later.

Adjust horizontal framing and zoom, use fit mode when the entire frame matters, or try the two-person split for an interview. ClippingShorts can track the largest visible face; this does not identify the active speaker. Review automatic tracking and use manual framing where it misses the subject.

Read the captions and check the exported file

Correct names, numbers and specialist terms in the transcript. Choose a readable caption style and keep essential words away from the edges. See the caption editing guide for a more detailed review.

Export a short test as MP4 or WebM, then play the downloaded file with sound. Confirm that the crop, captions, opening and ending survived the export. Keep the source video and download your project JSON if you plan to return later; the project file stores edits, not footage.

ClippingShorts currently previews YouTube links but does not download their media. Import your original recording to edit it. The preview also exports files for you to post manually; social scheduling is not connected.

Put it into practice

Keep learning

How to add captions to a video in your browser ↗How to make podcast clips that stand on their own ↗