[=Zeblote;75044]
Um, why not? Using the hairs example as gravity vectors, all the hairs would be normals to the surface and the problem wouldn’t exist, wouldn’t it?
[/]
I also have to provide a forward vector (X-Vector, where gravity is the Z-Vector) for the camera rotation. In most cases, using the global x-axis would work, but where the x-axis gets very close to or equals the gravity vector, problems with floating point precision arise.