Skip to content
Klyp

Speaker reframe

Always on the person talking.

Klyp turns wide video into vertical, square or portrait clips and keeps the active speaker in frame — with four layouts to choose from on every clip.

How it works
Footage: ITU · AI for Good · CC BY. Rendered by Klyp.

Highlights

  • Speaker tracking

    Faces are matched with who is speaking, and the frame cuts to the active speaker.

  • Four sizes

    9:16, 4:5, 1:1 and 16:9 — set for the whole project or per clip.

  • Four layouts

    Follow speaker, split, fit or center — switch any clip in the Layout panel.

From wide to vertical

A 16:9 interview becomes a 9:16 clip.

Klyp splits the clip into shots, finds the faces in each one and works out who is talking from the audio. The crop follows the speaker with a steady dead-zone, so it does not twitch, and cuts to the other person only after they start talking.

  • Groups are framed together when everyone fits
  • Moving subjects are followed when there is no face
  • Vertical sources are left as they are
Footage: ITU · AI for Good · CC BY. Rendered by Klyp.

Layouts

Follow, split, fit or center.

Auto picks a framing for you. Follow speaker crops to whoever is talking. Split stacks two people, one on top of the other. Fit keeps the whole frame over a blurred background — nothing cut off. Center is a simple centre crop.

  • Change the layout per clip
  • Fit is the safe choice for slides and screen recordings
  • Split works best when two faces are visible side by side

When the crop misses

Change the layout, not the whole clip.

If tracking picks the wrong person on a shot, switch that clip to Fit to keep the whole frame, or to Center when the action is in the middle. Changing the size re-runs framing for the new shape. Dragging the crop by hand (manual keyframes) is built into the editor and coming soon.

  • Layout is set per clip
  • Fit never cuts anyone off
  • Manual crop keyframes: coming soon
Footage: ITU · AI for Good · CC BY. Rendered by Klyp.

The same moment rendered for feeds and for YouTube. Captions and framing adapt to each size.

  • 4:5 portrait

  • 1:1 square

  • 16:9 landscape

Footage: ITU · AI for Good · CC BY. Rendered by Klyp.

How to use it

Speaker reframe, step by step.

  1. Step 1: Choose a size

    Pick 9:16, 4:5, 1:1 or 16:9 when you add the video — or later, per clip.
  2. Step 2: Let Klyp frame it

    Speaker tracking runs automatically on every clip.
  3. Step 3: Check the layout

    Switch between follow, split, fit and center in the Layout panel.
  4. Step 4: Fix a miss

    If a shot is framed wrong, switch that clip to Fit or Center.

Especially useful for podcasters, gaming & streams and agencies.

FAQ

Questions, answered.

Yes. Klyp frames the group when everyone fits one window, otherwise it follows the active speaker. Very wide panels with many people often look best with Fit.

There is no dedicated facecam-over-gameplay layout yet. Use Fit to keep the whole screen, or Center when the action is in the middle.

Yes — each clip can have its own size; framing is recalculated for the new shape.

Vertical sources are kept as they are, with at most a tiny crop to hit the exact output size.

Keep every speaker in frame.

Turn your last long video into ready-to-post clips in minutes. Free to start, no card.

See pricing