Do instead of divising a float by a constant, multiply by its inverse