Is everybody sleeping in this morning? Where's the ruckus I expected to find over the 5d2 samples I posted? ;D
In one more kick, at a camera that I think is junk with too many flaws, primarily its terribly noisy low ISO shadow performance, I ran a quick test on some files I had shot for this purpose just before I sold it to some poor sap.
I'd shot 1 EV steps from -5 to +3 of a smooth toned nearly neutral surface. Under wide-specrum flourescent, unfortunately, the sun was on the other side of the planet at the time so I couldn't use it.
Going from my real-world e.g. back on pg 8 of this thread, the ladies black pants seem to be responsible for the -5 EV hump in the histogram.
So I took my -5 EV sample shot into ACR, RAISED IT LESS THAN 2 STOPS AND THE FPN WAS ALREADY A PROBLEM.
If the 5d2 has nearly 4 stops above 0, and I'm being generous here, then the 5 stops below are only 9 EV worth of DR out of what it's claimed on DxO to be over 11 stops.
Well, if one wants to do ANY pushing in post, the DxOmark measurement for this camera's DR is
still misrepresented, it's still woefully
optimistic.
By my simple measurements, if you need to do a +2 EV push in the deep shadows, for
whatever reason, then
the 5D Mark II has a USEFUL DYNAMIC RANGE OF LESS THAN 9 STOPS before pattern noise becomes a problem.
= = = = = =
ADDENDUM 13-02-10 2320mst:
I knew it was likely even worse than this so I quickly checked a couple other test shots:
the -4 EV shot, raised less than 2 stops, also shows FPN
the -3 EV shot, will also show FPN if raised by 1 to 2 stops!
5D2 is now down to a 7 or 8 stop DR camera if any appearance of FPN is the cutoff point.
Little wonder I was not happy with it.
Anyone else want to do the same tests with their 5D2? It might be satisfying to know I had a lemon.
OTOH, you might not want to know the truth about your own camera. 
= = = = = =
That, my Canon-loving friends, is what I call a P-o-S camera and that's why I got rid of mine.
That, is why the 5D2 was the worst camera in my stable for my purposes and the most disappointing piece of Canon gear I'd ever purchased.
That, is the kind of useful information you can get from shooting dark frames and pushing them in post.
That, is why the 6D is looking like a major low ISO IQ improvement over the 5D2 and one main reason why I'd recommend it over the 5d2 for anyone who can afford the price difference, if they want to shoot Canon.
And finally, that is why I'd like to see a lot less moaning on this topic in general from people who haven't done any basic tests on this camera. Maybe I had a lemon, it was one of the earlier made ones. I doubt it. Altho it did meter with too much variability compared to my other, older bodies.
If you don't push in post, ever, for any reason, then your 5d2 will likely serve you well enough.
IF you do need to push in post, for creative reasons or merely to recover from an underexposure error, then the 5d2 could be a disappointing camera for you.
Since this is a 5d3 thread, sort of, shoot your own tests and see what you come up with. I don't have one, don't want one.
= = = = = =
another addendum - I-R agrees with me. See their DR results page, closer to the bottom
www.imaging-resource.com/PRODS/E5D2/E5D2IMATEST.HTM
and their sample images look to be cleaner in the shadows than shots from my camera
= = = = = =
