Ray you aren't listening!
There are CHARTS that show how good the color is from each camera. You should not waste your time looking at actual images and judging the color for yourself. Just use the charts!
CHARTS!
Heh heh, very good Doug.
But you know, as a scientist I also have a healthy respect for charts, once they are based on data of good provenance and sound analysis, or realistic models. I see no reason (yet) not to place DxO results in the former category.
I will never understand the knee-jerk rejection of such results by some people. There's a tendency for a reaction like "that's not what I see with my eyes, so I refuse to believe it". Should we always "trust our eyes"? Well, my eyes tell me that clearly the world is flat and that the stars are in orbit around it. But science has shown such conclusions to be fallacies, not reality.
It's important also to remember that there are things that DxO doesn't test and does not claim to address. Colour bit depth is one thing; spectral response curves are entirely different, and they, I think more than anything else, determine our impression of colour quality.
I'd also love to see long exposure dark noise being tested and rated. If that were to happen, we would not be long in seeing the majority of the MFD sensors losing serious ground in the overall DxO ratings.
Now for the DR debate. I think that some of the controversy may be explained by the way that DxO presents DR as a single number at each ISO (a single number is of course how DR is defined). This encapsulates the range between min and max signals, but says nothing about the signal-to-noise in between, all the way along the range and up to the max. A camera with a low readout noise (modern Sony CMOS) will have the edge at the dark end, while a camera with a large pixel well depth (a characteristic of many Dalsa MF CCDs) will have the edge in the middle and at the bright end.
Another difference is that they will achieve these performances at different sensitivities. The MF CCD needs lots of light to fill that pixel well, so it will be say ISO 50 while the Sony CMOS will be around ISO 100-200 - it needs 1-2 stops less light or else it saturates. Personally, if DR was a tie, and neglecting for a moment all other factors but DR, I'd much rather have the camera which achieves it at a higher ISO.
I see very, very little discussion of signal-to-noise in these debates. Probably because it's harder to quantify (one must work with individual colour channels in the un-interpolated RAW file) and naturally most photographers can't be bothered. But it determines the "quality of the DR" (not quantity). It's like, one guy says he can fit 20 people into a Mini. I'm impressed, but he comes back with 20 5-year-olds. Hmmm. Another guy says he can also fit 20 people into the Mini (see, again 20 people...same DR!). He comes back with 20 prop-forwards. Now I'm seriously impressed!
Ray
PS love your B&W series...some are reminiscent of Ansel Adams' Oregon Dunes...you definitely succeeded in your "body as landscape" goal.