Ray you did not answer the question but went off on some other path. Bottom line is they sell to Nikon does it pay to show a P65 better. Makes you wonder. I never trusted DXO numbers first it is not real world and that is fine if you do . Not here to debate them but i can tell you I could not get this detail in the foreground with 35mm period. I would maybe even get it with a P25+ a P45+ I would say pretty much but after a lot of testing the P65 and P40 smoke those 2 backs on DR. This is a P40+ shot which without any work to it I don't believe shooting into a sunset this detail would show. This is the stuff I go by. If you want to go by DXO numbers great but I go by direct comparisons of known backs and real world situations. With that I have work to get done. But go out and test. I do it all the time with public results . If I wanted a D3X i would buy one without concern I know it is a good cam. That is not the point though. I can buy anything I want, it's all a write off anyway.
You also admit that cost is a concern and i hear that everyday and totally understand it but there is a lot to MF that many just want to blank out as well because they simply want to avoid paying. Which in turn makes there 35mm chose the best thing going. Sorry folks i own a forum too but i see this everyday in a Canon or Nikon forum. Defending it until the cows come home, which is fine but if your really interested in MF than you should go demo it and work with the files. There is a lot to it than just simple DR , noise levels. Honestly i don't care what folks buy all I care about is they have the correct information before they buy anything.
Guy, I thought I'd answered you pretty clearly. I tend to base my opinions on the evidence. If I have reason to doubt the impartiality or accuracy of the evidence presented, I'll do my own tests, if I'm able to without too much expense and inconvenience, and if I have the time.
When the 15mp Canon 50D was realeased over a year ago, an upgrade from the 10mp 40D, I decided the upgraded features were significant enough to justify the purchase of the camera even though I already owned a 40D.
One point that was raised by many folks on this forum and elswhere, was that the 50% increase in pixel count of the 50D meant that the camera would provide no resolution benefit (compared with the 40D) at apertures smaller than F8 because the effects of diffraction would negate any potential increase in resolution. Such opinions seemed to be based on theories relating to the diffraction spot size or Airey disc at F stops greater than F8, compared with the 50D pixel size.
Owning both cameras, I decided to check this issue for myself, using one of my best prime lenses, the Canon 50/1.4, using the LiveView feature on both cameras to achieve accurate focussing, using tripod remote cord and MLU enabled, as well as employing in all respects the usual flawless technique for which I am famous. I confirmed for myself that F22 was the stop at which no resolution benefit was gained from the 50D. At F16, the resolution of the 50D images is about the same as the resolution of the 40D at F11, and the resolution of the 50D images at F11 is about the same as that of the 40D at F8.
Now theories should always be tested against reality. DXO Mark's data are not theory. They are practical results flowing from highly specialised and carefully conducted test procedures. The idea that DXO Labs are in the pocket of Nikon, deliberately exagerating the impressiveness of the D3X test results, and/or deliberately diminishing the P65+ test results to make the D3X results appear more impressive, is just preposterous, unless you are into conspiracy theories.
DXO Labs produce a RAW converter for many brands of cameras, Pentax, Canon, Sony, Panasonic etc. How do you think those other manufacturers would feel, as well as DXO customers, if they thought Nikon was getting special treatment which was not deserved?
The D3X appears to be the only 35mm camera that exceeds the image quality of MFDB in some respects, but not in any significant way. A half stop increase in DR, compared with the P65+, on same size prints, is no big deal, but it's better than 1/2 a stop worse DR. Of course there are other areas in performance where the P65+ is at least slightly better than the D3X, such as marginally better tonal range, marginally better color sensitivity and marginally lower noise at 18% grey, not to mention the obvious improvement of increased resolution on very big prints. Anyone who owns an Epson Stylus 9900 or 11880 would find the additional pixel count of the P65+ very useful.
So, once again, I have no reason to suspect that the DXO test procedure is flawed. If you do have reason to suspect it is flawed with respect to the P65+, then please do us all a favour by demonstrating this with evidence.
If you do decide to do this, a few words of advice; make sure the lighting is the same for both shots and that both shots are truly and equally exposed-to-the-right; try to get the focussing exactly on the same spot in both shots, and always adjust focal length and aperture for equal FoV and DoF.
The P65+ sensor is slightly larger than most DBs which means there'll be slightly more than a stop difference for same DoF. F8 on the P65+ should be equivalent to F5 on the D3X, and F13 on the P65+ should be equivalent to F8 on the D3X.
Although your sunset shot with dark foreground is a lovely shot, it cannot demonstrate the qualities of MFDB. You need a comparison of the same scene. In any case, you must have read of Michael's comparison between the P45+ and the Canon G10 P&S. After showing A3+ prints of both images, experienced photographers found it impossible to tell which print was from which camera, until it was realised that the slightly shallower DoF in one of the prints was a clue that the print was probably made for the MFDB image. Your web image is smaller than A3.
I've also got a sunset shot with a dark foreground which is reasonably clean, taken many years ago with my first DSLR, the very limited 6mp cropped-format Canon D60.