I use Dave Kirk's (NVIDIA effects browser v2.0, this technique is used in Microsoft Meltdown)
technique of shadow volumes generation (using only hardware with vertex shaders).
So I think to change this technique to traditional because of not very
good quality of volumes and some limitations, caused by using this
Does anybody know if the traditional shadow volumes generation
technique (using CPU with volumes capping) faster than Dave Kirk's
shadow volume technique?
Thanks a lot.