When transforming surface normals from object space to world space, why can you not simply use the model matrix?
-
A
Normals use only 3 components
-
B
Non-uniform scaling would distort normal directions
-
C
Normals must remain in eye space
-
D
The model matrix only works on positions