Keep in mind, people have been complaining about banding noise in the 5200 when shadows are lifted by around 2 stops. I have not seen any super serious example of it so far, but the D5200 sensor does not seem to really fare any better than Canon sensors (which can usually only withstand about 2 stops shadow lifting). The numerous mentions of banding in the 5200, and the mentions about how similar it is to Canon 5D II banding, make me wonder even more about the objectivity of DXO.
D5200 is showing shadow banding?!?
That'll be seriously disappointing.
I haven't had a chance to research on it yet, was hoping it would be as smooth as the D5100 and D800.
The denser AF system would be handy for some action work.
That doesn't bode well then for higher end Nik DX bodies, not yet announced.
maybe I should buy a couple more tilted D5100s while they're still on the shelf at $400. FWIW, they're just as good up to a large enough print size.
One area on DxOmark where you might
find evidence of low ISO shadow banding is if you look at the full review for a camera\measurements\Full SNR\log display.
E.G. On the 5D2, low ISOs you see the curve suddenly hook down at the dark end.
This may not be the entire reason or consistent enough to use as a banding guideline because they don't show this hook for the nasty-bandy 7D and show hooks on the 40D, which I have and does not show up any low ISO banding.