Use pi = 3.141592654

or like this:
float const pi = 4.f * atanf(1.f);

You don't need double precision to get AC

Edited by author 29.06.2012 18:44