Why can we not build up intersecting transparent triangles in the same way as we do opaque and just account for alpha?
username
It seems like because we don't update the depth buffer, it is difficult to find which triangle is over another for the two parts divided by the intersection of two intersecting transparent triangles.
sarukkai
What is the best way to handle z-buffers and transparent objects if we were planning on using the z-buffer in a post-processing task (ex. incorporating mist into the scene with thickness depending on the distance from the camera)? Would we want to update the z-buffer with a form of weighted blending similar to what we use with alpha blending?
gracie
Does this work for supersampling?
kmc
I understand that semi-transparent triangles should be rendered either front-to-back or back-to-front, but I don't quite get why it has to be back-to-front? Why is that better than front-to-back?
Why can we not build up intersecting transparent triangles in the same way as we do opaque and just account for alpha?
It seems like because we don't update the depth buffer, it is difficult to find which triangle is over another for the two parts divided by the intersection of two intersecting transparent triangles.
What is the best way to handle z-buffers and transparent objects if we were planning on using the z-buffer in a post-processing task (ex. incorporating mist into the scene with thickness depending on the distance from the camera)? Would we want to update the z-buffer with a form of weighted blending similar to what we use with alpha blending?
Does this work for supersampling?
I understand that semi-transparent triangles should be rendered either front-to-back or back-to-front, but I don't quite get why it has to be back-to-front? Why is that better than front-to-back?