Question concerning an ATI Depth Of Field paper


On page 13 they create a 1/16th size low resolution image
of the current MSAA framebuffer. After that they apply a
3x3 Gaussian blur:

When they create the low resolution image, I guess there
is no mipmapping involved? So I could just setup my big
framebuffer color texture attachment with bilinear
filtering and then render the content of this texture
into a small texture?