Pupil densification is a technique to improve contrast in a direct imaging interferometer. By enlarging the size of each aperture in a sparse array, the fraction of energy contained in the central point spread function peak can be increased. However, the effective field of view is reduced as the contrast improves. Our work investigates the optimal strategy for pupil densification in the context of extended source imaging. Analytical expressions are derived to optimize the signal-to-noise ratio at the edge of the field of view for a variety of hypertelescope configurations. Free-space beam transport and fiber-optic beam transport are both considered. The relative advantages and disadvantages between the considered hypertelescope configurations are discussed.