“…A frequently used approach to the evaluation of interpolation kernels is to compare the spatial and spectral properties of these kernels to those of the sinc function, either by discussing their low-frequency band-pass and the high-frequency suppression capabilities [29,41], or by using such metrics as "sampling and reconstruction blur" [39,47], "smoothing", "post-aliasing", or "overshoot" [30], "truncation error", or "non-sinc error" [28], to mention but a few. These approaches are based on the fundamental assumption that in all cases, the sinc function is the optimal interpolation kernel.…”