Right. The computation reminds me a lot of the inverse of dithering algorithms for print. Plenty of tradeoffs to be made, different goals to be achieved, and of course, subjectivity.I understand what you're saying now. I didn't realize the demosaicing was so straightforward for these two modes.
