January 11th, 2009
I just started working on lens flares; I need to see the sun's location so I can test shadows! I am testing a new technique for dynamic lens flares based on the anatomy of the human eye, inspired by a paper by Tobias Ritschel.
The basic idea is to simulate the glare as a human eye would see it, instead of how a camera would see it. This means the glare has to be animated: the movements of the iris in response to bright light cause the diffraction streaks to change in subtle ways. Simulating an eye rather than a camera also results in a more grainy texture and a rainbow-colored lenticular halo, instead of the geometric internal reflections you normally see in camera flares.