Skip to content

How do I code a video with the AI Assistant?

1 min readAI Assistant

Overview

Open Code with AI → Single video (one video) or Multiple videos (a list with a picker to choose which videos are in scope). The tabs are Overview, Transcript, Time ranges, Frames, Regions, Description and Codes. The panel on the right follows the tab: the video, the playable transcript, the timeline with your time ranges, or the selected frame picture. A transcript is optional: frames and regions can be coded without one; a transcript adds the words around a frame and lets you code what is said.

What to do

  1. Transcript: transcribe the recording if you want to code what is said. Press a line to play it.
  2. Time ranges: drag on the timeline to select a stretch, then press Code this range.
  3. Frames: choose or capture frames, zoom the timeline, or press Capture current frame.
  4. Regions, Description and Codes: work on the captured frames with the same steps as for images. Describe a frame first; code suggestions rest on its visual description.
The Frames tab with Find frames
Frames: choose a rule (From ranges, Seconds, Count or Time), then Find frames.
The Ranges tab of a video
Ranges: Find ranges from the transcript, then Save, Play or Open each one.
The Audio tab of a video
Audio: the same rules find ranges for the sound.
The Sound panel of a video without a sound track
A video without a sound track shows this message instead of a waveform.

Terms in this answer