Good talk - the change of basis to apply simple constraints in the spatial domain is simple but the details are clever and a bit mind blowing. And finally, the compression of 44:1 with little loss of detail was amazing. That's one where I'll need to read the papers. Could that technique also apply to image compression ? I wonder how it would perform.
I still remember baking lighting in a command prompt for custom maps in Call of Duty 1. Good old Radiant days.
It's quite amazing how far things have come.
Awesome learned so much!
Good talk - the change of basis to apply simple constraints in the spatial domain is simple but the details are clever and a bit mind blowing. And finally, the compression of 44:1 with little loss of detail was amazing. That's one where I'll need to read the papers. Could that technique also apply to image compression ? I wonder how it would perform.
why shoot many visibility rays from sample point in all directions instead of just one towards camera ?
bounce light?
indirect lighting can only be computed if light bounces are taken into account.
Light leaking. If you don't calculate visibility in many directions you get light leaking through thin walls and so on.
1.5gb is a memory problem? I would gladly pay 8gb memory requirement for those tasty tasty shadows and indirect lighting.