A clean cutout is only part of the picture
A presenter may have a clean key and still feel disconnected from the set. Viewers also read contact with the floor, depth relative to objects, shadows and reflective surfaces. Those relationships need to remain coherent as the person moves.
Camera movement and presenter movement are different problems
Camera tracking tells the virtual scene how the physical camera is moving. It does not, by itself, describe where the presenter is standing. A person can walk across a correctly tracked shot while a fixed shadow or reflection stays behind.
Talent tracking supplies position information about the people in the space. This lets the scene respond to their movement as well as the camera’s movement.
Let floor contact follow the person
Wyvern’s vision-based system tracks each presenter’s whole body as a 3D object and projects that representation into the virtual scene. This gives the scene information about the body’s spatial form, not only a single position. Shadows, reflections and depth relationships can respond coherently to movement and different camera angles. The system automatically updates shadows and reflections at the presenter’s feet as they move.
The aim is to remove the need to keep repositioning these effects manually during a live show. Their appearance still needs to fit the virtual environment and the shot.
Make the effect belong to the scene
A reflection should suit the surface it appears on, and a shadow should agree with the scene’s lighting. Stronger effects are not always more convincing. A subtle contact cue can do more for a presenter’s connection to the floor than a prominent effect that looks out of place.
Inspect movement, not only a still
Test the intended presenter positions, camera angles and movement. Look at floor contact, foreground crossings and reflective materials throughout the shot. A useful production demonstration shows what remains consistent and where the setup needs adjustment.