Files
hyperframes/skills/hyperframes-animation
miga-heygenandMiguel Ángel 6c353d89c8 docs(skills): teach shared cut constants and fill lights for Three.js scenes (#5077)
The three adapter skill had no guidance for camera cuts made inside one
canvas. A visibility window written as its own literal near a cut shows
the next shot's objects for the last frames of the previous shot, a
two-frame pop-in that a still per shot does not reveal. The skill now
says to declare each cut once, read it in both the shot selector and
every visibility window, and snapshot the last frame before each cut.

It also explains why a MeshStandardMaterial lit by a single key light
renders as a flat black cutout when the camera faces that light, and to
add a fill such as a HemisphereLight driven at a fixed ratio to the key.

Signed: Miga

Co-authored-by: Miguel Ángel <miguel.sierra@heygen.com>
2026-10-05 16:59:59 +00:00
..