I would remove DR from that list of MF advantages: my prediction is that DR will favor the D800 over DMF CCDs with their far worse read noise levels (now about five to ten times worse than good CMOS sensors). Does that just leave MF the market for those who need even more resolution, and the dilettantes who blindly believe that "the sensor is bigger, so the results must be better"? Hopefully not; if only because MF lens systems offer some continuing IQ advantages for some uses.
True, but don't forget that there are MF CMOS backs allegedly coming over the hill any day now! If (big if) that happens, it levels the noise/ISO/DR playing field, and then sensor size matters again.
Agreed, as I said in another thread on this topic:http://www.luminous-landscape.com/forum/index.php?topic=62244.msg500981#msg500981Can you point to signs that such "big CMOS" is coming? I can vaguely see Fujiflm trying MF sensors again: the failure of its MF sized SuperCCD seemed to be due to "politics" like the inertia of MF back makers, despite its good promise of performance advantages over the traditional Full Frame CCD technology offers by Kodak and Dalsa. Maybe products like the D800 will wake DMF makers up to the commercial need to move on from CCDs.
Friend of mine is a sponsored Nikon user on the local scene. He has tested the new Nikon and says that the image quality is fantastic. Not great at high iso and pretty hard on lenses. He also told me that he didn't think it a match for the Leaf Aptus 7, a back he is very familiar with. All in all he thinks it a great camera and exceptional value.
I took a look at the photo gallery on Nikon's website as well. Underwhelming is indeed the word... I am sure the camera is capable of doing much better.
The sample photographs are not very convincing. Yes it's a terrific DSLR, no it's not medium format quality. Compared to a Pentax 645D those files looks smudgy, lacking detail and sharpness. I don't believe for a minute it will make quality 40x50" prints anything like that from a 645D.
There is in fact zero theoretical reason why it wouldn't be every bit as good but it is true that the current samples are not as good as expertly processed 645D files.