Nowadays, virtual studio solutions are still using image-plane based displaying of the camera input, forcing an operator to manually position this plane in the virtual space according to the performer’s and the camera movement. Our solution realtime generates a 3D mesh in the virtual space, using the camera track and the depth informaton, allowing the performer and the camera to move / rotate freely. In this solution the positioning is fully automatic, no operators needed. The performer can walk behind / around virtual objects.