In the closing speech of DSC 2004 (Paris), it was remarked that, "No child would be impressed by a driving simulator because they are too used to the impressive graphics presented to them by the games industry." This is because the gaming industry is competitive about getting games with attractive graphics for people to buy; to this achieve this aim the market spends millions. Although aesthetically pleasing, the scenes presented in games are not physically realistic.
Traditionally, realistic image synthesis by techniques such as ray-tracing have been too computationally intensive to produce animations in real-time. However, advances in algorithms, traditional processors and more recently graphics cards, have helped bridge this gap. Although we are not there yet, these methods will mark the end of the bland OpenGL graphics, used in many simulators, in turn making simulators more realistic and believable.
This paper concentrates on the perceptual techniques that are being developed to reduce the gap further. In particular, it focuses on the techniques being researched and employed to get perceptual rendering working on a 6 degree of freedom motion simulator connected to an eight computer cluster.