How Snap Frame Works
From flat screenshot to perfect perspective in three steps — here's exactly what works, what doesn't, and why.
Select the screen area
Select a rectangle, an image, or a flattened 4-point vector covering the screen you want to fill. Rotated shapes are fully supported.
Upload your screenshot
PNG, JPG or WEBP. Your image is processed locally in Figma and never uploaded to a server.
Apply
Snap Frame calculates the perspective from the shape's four corner positions and warps your image into place as the fill.
What works
- RectanglesDraw one with R over the screen area, ready to go.
- ImagesExisting images work as targets — the screenshot replaces the fill.
- 4-point vectorsAny flattened shape with exactly 4 corners (Cmd/Ctrl+E). Great for angled screens.
- Rotated shapesRotation on the shape or any parent is handled automatically.
What doesn't work
- Frames and GroupsContainers can't be used as targets. Select the shape inside instead.
- Shapes without 4 cornersAnything with more or fewer than 4 corner points — the perspective math needs exactly 4 corners.
If your shape isn't accepted, trace the screen area with a rectangle — it takes two seconds and always works.
Under the hood
Snap Frame reads the four corner positions of your selected shape — including all rotation from the shape and its parents — then computes a perspective transform (homography) that maps your screenshot's corners exactly onto the shape's corners. Every pixel is remapped through that transform, so parts of the image that tilt away compress naturally, just like real perspective. The result is applied as a fill. All of it happens locally on your device.