Dual stage
This browser does not support immersive model presentation — you need Safari on visionOS. The controls below are disabled.
Experience
Why one video at a time?
Video docking rides on the Fullscreen API: the docked video is the
document’s fullscreen element, and a document can only have one. The
USDZ format annotation (RealityKit.CustomDockingRegion)
is also authored once per scene by Apple’s Blender add-on. So this
page ships one USDZ per docking target — cube or sphere — and swaps
the <model> source when you change target. The
“both” variant embeds two docking regions in one file as an
experiment; which region the player picks is not defined by the
current platform.
Static images have no such limit: the banded sphere on the centre plinth and the framed posters are ordinary textured geometry baked into the USDZ — images can wrap any mesh, video can only dock.