Lensrentals: System MTF comparison, Canon vs Nikon

But that is exactly what I was doing, looking at the resolution values of the lens in conjunction with D7000 pixels and extrapolating that into values comparable to D3x's results. In effect, I was trying to see how much more a 36MP FX sensor would resolve compared to 24MP FX sensor. The problem was that according to my calculations the 36MP sensor would've resolved less than the 24MP sensor, which is simply ridiculous.

But perhaps goosel is right, that the small dent in the MTF figure is one of the culprits. However, the DX test shows that both DX 'border' and 'extreme' distances show equally poor resolution. It would be quite a stretch to assume that both the 'border' and 'extreme' sample points in the DX tests fell into the pit of the MTF figure while in the FX tests the corresponding sample points would hit the sweet spots.

Nikkor 28 f/1.8 MTF curve from dphotojournal.com

At the moment, I'm guessing they're different units and the one in the DX test is just unlucky.

PS: I'm curious because at the moment, 75% of my photos are taken with a 28mm focal length, and there are not too many alternatives. I'm 95% certain I'll pick up a Zeiss 28/2, though, so I'm not that worried.

PPS: Who was it that said that roughly 72.825% of all estimations were made on the spot?

--
regards
Janne Mankila, Finland