Does using acceleration structures for scenes with very large number of objects result in some form of loss of quality?
Staffemilyma53
@phoebeli23 using these structures do not result in any loss of quality. Since we only rule out the objects that have no chance of intersecting, we guarantee that if there is an intersection, it will be correctly identified.
ML2000-LT
Isn't the #pixels much greater in magnitude? Or is it that these accelerations would yield effective in multi-object sences such that we have 500+ objects. I'm wondering how effective this is since log() on a small number is not signifcant
Does using acceleration structures for scenes with very large number of objects result in some form of loss of quality?
@phoebeli23 using these structures do not result in any loss of quality. Since we only rule out the objects that have no chance of intersecting, we guarantee that if there is an intersection, it will be correctly identified.
Isn't the #pixels much greater in magnitude? Or is it that these accelerations would yield effective in multi-object sences such that we have 500+ objects. I'm wondering how effective this is since log() on a small number is not signifcant