When looking up all the great things that Varaince Shadow Maps can achive I was wondering which downsides are comming with VSM. And well... it seems to have some light bleeding, other than that I wasnt able to find anything in my research.
So I was wondering. Does the depth percission decreese with VSM? Because we are storing depth and depth squared but have the same amout of bits as oposed to only store the depth, which means we store our values with less precision, correct? I know that when checking for shadows in VSM its not just comparing two depth values (I am honest and tell you, I couldnt fully understand the whole math going on with vsm ) but it still seems to me, that the more bits available to store the values, the better?!
Would be cool if someone could enlighten me here ; )