Daniel Amerson and Shaun Kime
In conventional, visual narrative media such as film, the placement and movement of the camera can be as important a device as the events of the narrative in conveying story and meaning. Hollywood has developed a set of common cinematographic techniques to convey meaning for different subject matters. For interactive narratives in 3D virtual worlds, the camera can be manipulated in the same way as traditional cinema. We propose a system for real-time camera control in interactive narratives called FILM, Film Idiom Language and Model, that uses these common cinematographic techniques to construct camera placements based on input from a narrative planner. Information about common film idioms is encoded in a scene tree using the FILM language. Objects within the FILM system use this knowledge in conjunction with the planner inputs to constrain the location and orientation of the camera for viewing a given action at execution time.