Fig. 5. Larger field of view scenario with two versions of the office scene.
Virtual lens vignetting effect without spatial sectioning (column a), with spatial sectioning (column b). Voxels at a given depth z should be computed for a time t when they are actually illuminated by part of the virtual illumination pulse. Using only one distance-time shift (L = 1 in Eq. (12) for all voxels results in large errors in voxel brightness that lead to a blurry appearance. Using two different distance-time shifts (L = 2) in different voxel regions results in crisper images. This motivates the spatial sectioning method to mimic a perfect virtual imaging system that is not limited to a small field of view.
