跳转到主要内容
Workflows 5 min read

Extract Frames: Turn a Clip into an Image Sequence

Screenshots pick one frame; frame extraction dumps a whole sequence at a set rate. Motion breakdowns, asset libraries, and cover-photo hunting all work better as batches. Here is how to pick the rate, the format, and how many images to expect.

The rate sets the rhythm

Six rates are available: 10 fps for motion analysis (swings, jumps, frame-by-frame faults); 5 and 2 fps for fast cuts; 1 fps as the default for most sifting; every 2 or 5 seconds for long-video triage — a 10-minute lecture at one frame per 5 seconds is 120 images, far faster to skim than scrubbing.

  • Motion breakdown: 10 fps
  • General sifting and cover hunting: 1 fps
  • Long-video triage: 1 frame per 2-5 seconds

JPG or PNG

JPG uses quality factor 2 — small and fast, right for sifting and archiving. PNG is lossless, right for further editing, cutouts, or archival. Sift in JPG, then re-extract the keepers.

Estimate the count first

Images = duration × rate. Thirty seconds at 5 fps is 150 images; at 1 fps, 30. One notch of rate changes the cleanup work several-fold. Triage long videos at a low rate, then re-extract the marked segment densely. All local, nothing uploaded.

Frequently asked questions