Make a first cut

Markdown · Agent index

Turn a short spoken recording into a captioned MP4. Start with the Yapwing preview on an Apple Silicon Mac running macOS 26 or later, and a recording you have permission to edit.

1. Open Yapwing and choose File → New Project… (⌘N). Save the project in a folder you want to keep.
2. Choose File → Import Media… (⌘I) and select the recording. Let local transcription finish. The words appear beside the preview, and the clip appears on the timeline.
3. Play the recording and select a passage you want to remove in the transcript. Delete the selected words. Their recorded audio and picture leave the cut together; the original file stays unchanged.
4. Play across the edit. Use Edit → Undo to restore it if needed. To bring back removed words later, select them in the transcript and choose Edit → Restore Deleted Words.
5. Correct any caption spelling in the transcript. A wording correction changes the caption, not the recorded voice.
6. Choose File → Export… (⌘E), pick a new destination and export the video. The first export is free; later exports need an active subscription. Open the resulting MP4 and check the cut and captions.

The first time Yapwing runs, it downloads its speech model in the background and shows “Getting the speech model…”. A recording imported before the download finishes is transcribed by a smaller built-in model, and its captions update afterward.

The export’s size follows the project’s format (1080 × 1920 for the default Portrait 9:16; see Formats, silence and loudness), at 30 frames per second, with H.264 video and AAC audio in SDR. A preview build has had less testing than a stable release; inspect an export before publishing it.

For a project you will move to another Mac, use File → Collect Assets before copying its folder. Projects, originals and undo explains what to keep.

Shorten pauses

Choose Edit → Tighten Pauses to shorten every pause longer than 0.4 seconds to 0.4 seconds. With a selection that contains a pause, it changes only the selection. The change is one Undo step.

Formats, silence and loudness

• View → Format and the toolbar’s Format pop-up change the output to Portrait 9:16, Vertical 4:5 (1080 × 1350), Square 1:1 (1080 × 1080) or Landscape 16:9 (1920 × 1080).
• Edit → Include Silence, under Tighten Pauses, also removes silence measured in the recording’s sound.
• Export levels the sound to −14 LUFS and keeps its true peak at or under −1 dBTP.