Originally posted by falconeye From the bits of information gathered so far, I think I can summarize the video architecture now:
Great stuff, falconeye. That is a very impressive analysis.
I'm not a friend of suboptimal solutions such as the fringing causing subsampling approach, but your analysis shows us why we shouldn't care too much. After all, it seems that a different approach, while being awesome would have increased the cost.
I think it was a good move to basically just get still shooting FPS up to par and throw in whatever video quality that enables for free, rather than making a DSLR more expensive just in order to optimise a secondary function (video).
If we look at the video feature as I think we should, as a nice freebie, then any disappointment about still existing shortcomings fades away, AFAIC.