Original Post
When lighting in object space, I transform the world space light position into object space by multiplying by the object's inverse world matrix. However, I read that when lighting in view space, you should transform the world space light position by only the rotational part of the view matrix. Why is this?