Auto reframe is the feature that turns a horizontal video into a vertical one without you keyframing the crop by hand. Every clipping tool has some version of it. The gap between a good one and a bad one is entirely about how it decides where to point the crop.
Why a static centre crop fails
The simplest reframe takes the middle 9:16 slice of a 16:9 frame and holds it there. That works only if your subject sits dead centre and never moves. In reality:
- Two people on a podcast set means the crop lands in the gap between them.
- A speaker who gestures or walks drifts to the edge and out.
- A cut from a wide shot to a closeup isn't accounted for at all.
What a good auto reframe does instead
- Detects the subject — finds the face (or faces) in each frame.
- Moves the crop to follow — a virtual camera that eases toward the subject instead of locking to one spot.
- Respects your edits — detects shot changes in the source and re-frames each shot separately, rather than gliding the crop across a hard cut.
- Sizes the crop from the subject — close enough to frame the person, not so wide it catches the next person along.
The tell of a cheap reframe: the crop pans smoothly across a cut in the original video. A good one snaps its framing on the cut, the way a human editor would.
One, two and three people
The hardest case is a variable number of people. A single layout can't frame a solo closeup and a three-person wide shot well. Shot-aware reframing picks a one-, two- or three-pane layout per shot based on who's actually on screen. Gameplay plus facecam is another layout in the same family — a tracked face over a background clip.
How ClipSonic does it
ClipSonic's Auto Frame detects every shot, picks the layout per shot, tracks each face with a smoothed virtual camera, and cuts the framing on your edits — all rendered locally. Full detail: AI face tracking & auto reframe.
Download ClipSonic free and reframe one of your own clips to see the difference.