What Is Face Tracking in Video Editing?

When you convert a wide video to a tall 9:16 frame, you throw away most of the width. Face tracking answers the crucial question: *which* slice of the frame do you keep? The answer is 'the one with the person in it' — even as that person moves.

How face tracking works

The software analyzes each frame, detects where faces are, and decides which subject matters. The crop window then follows that subject frame to frame. Good implementations add stabilization so the crop glides instead of jittering, and subject priority so the dominant speaker stays the focus in a crowded frame. Klypse's face tracking behaves like a camera operator, not a per-frame cropper.

Why it matters for short-form

• No cut-off subjects — the person stays in frame even as they pace or gesture.
• Stable, professional framing — no nauseating jitter between frames.
• Right subject in focus — in multi-person scenes, the speaker stays centered.
• Faster editing — you skip manual keyframing of the crop entirely.

Face tracking vs a static center crop

A static center crop keeps a fixed rectangle regardless of what's happening. It works only if your subject never leaves the center — rare in real footage. The moment someone steps sideways, they're half out of frame. Face tracking removes that failure mode entirely, which is why it's the backbone of good auto-reframing.

What about multiple people?

Advanced tracking handles multiple speakers — following whoever is talking, or splitting the frame so two people share the screen. That's essential for interviews and podcasts, where a single-subject crop can't tell the story. Read smart face tracking and cropping for how the decisions are made.

In short: face tracking is what lets a wide recording become a vertical clip without sacrificing the subject. It's the quiet feature that separates clips that look edited from clips that look cropped.