Lecture 9: Ray Tracing & Acceleration Structures (78)
kalebdawit
What is the cost tradeoff between minimizing bounding box area and the number of triangles per box? As Ren mentioned, we don't want our nodes to take up too much dead space because then too many rays will hit them even if they don't intersect triangles. But also, I'd imagine that we don't want to have too many triangles in a single box because then the cost of traversing that node and checking all traingles will be high. How do we manage this? Is the answer just to split the triangles into many sets?
What is the cost tradeoff between minimizing bounding box area and the number of triangles per box? As Ren mentioned, we don't want our nodes to take up too much dead space because then too many rays will hit them even if they don't intersect triangles. But also, I'd imagine that we don't want to have too many triangles in a single box because then the cost of traversing that node and checking all traingles will be high. How do we manage this? Is the answer just to split the triangles into many sets?