>>3814134>[CITATION NEEDED]It's physics dude. A weakened CFA allows more non-red light to hit the photosite dedicated to red and the same for the other two colors. The camera then has to extrapolate more based on that.
>I'll point out again that in Canon's lineup at least color quality improved over time. I'm pretty sure if I shoot a color chart with the 10D, which I still have, and the 5Ds, the 5Ds will be more accurate, not less. And both the 7D and the 5Ds can distinguish more tones/colors than the old 10D could. They're also more neutral under more severe lighting conditions. I had a harder time color balancing 10D and 20D RAWs to taste than I ever had with the 7D, 5D mark III, or 5Ds.1Ds III and 5D have better colors than modern Canons. Sure, the banding issues have greatly improved in the last years. 5D IV has pretty much no banding noise when pushed unlike 5D II and 5D III.
>Too many variables at play to make conclusions about hardware. There's also the implicit assumption about color at the scene. "I like CCD so this green must be right" is a far cry from "I still have the object and can hold it next to a monitor/print and judge."I kinda covered that already. It's not about liking or not. Just like chromacity is negatively affected by weaker CFAs, luminance is positively affected by it. It's a tradeoff.
I knew what green was right because one sensor was picking the difference and the other wasn't. I knew that pic related was the M240 because the color on the chairs looked manipulated, too reddish and almost matching the wall, while the lettering seemed accurate. In some pictures, there's little to no difference but when you have complex colors things change.
There's simply less information to work with.
