Vinten Radamec Free-d - Manual (Page 4)

Vinten Broadcast Ltd free - d ™ Installation Manual V1.4.4
Page 4 of 46
1 Introduction
1.1 Virtual sets
A key component of any virtual studio system is a means of measuring the
precise position and orientation of each studio camera. This data is used to
render the virtual scene from the appropriate viewpoint.
A camera tracking system must allow unrestricted movement of many studio
cameras while panning through 360 degrees, work with a wide variety of
camera mountings including manual pedestals, cranes, and hand-held
cameras. The measurement of position and orientation of the camera must
be to an accuracy that ensures precise matching of the real and virtual worlds.
It should place no significant constraints on either the scene content or the
studio environment.
free-d™ satisfies all of these requirements.
1.2 free-d™
Developed by BBC R&D, this system uses a number of targets placed out of
shot above the studio lighting grid. The targets are composed of concentric
black and white retro-reflective rings forming a type of barcode. Each target
has a unique code number, and is mounted at one of two different heights for
optimum performance.
A small Target Seeking Camera (TSC) mounted on each studio camera views
these targets. The TSC is mounted rigidly with its optical axis at right angles to
the optical axis of the studio camera. A ring of LEDs surrounds the lens of the
TSC. They illuminate the retro-reflective ceiling mounted targets.
The data from the TSC is passed to a purpose-built hardware unit, (free-d™
processor ) which calculates the precise position and orientation of the studio
camera in real time. free-d™ identifies the targets, measures their position to
sub-pixel accuracy, and reads their barcodes. The position and orientation of
the camera can then be computed.
The studio camera’s zoom and focus are measured using conventional
mechanical sensors. These data are multiplexed with the signal from the
auxiliary camera.
free-d™ outputs an RS422 signal that contains camera X, Y,
height, pan, tilt, roll, zoom and focus data. Using this data output, the virtual
set system will then create a background very accurately matching the
position of the real camera, to within 1mm.