[wim]

[quote name='Frank' timestamp='1328600493' post='15599']

I am curious how many Nikon lenses can match the 36MP resoluion on D800. My impression was that many (or some?) lenses already struggled a lot on the 24MP D3x.



Another question is on the lens diffraction. For a 36MP FX, at f5.6 the diffusion circle of red light is already larger than the sensor pixel size; while for the blue light the diffusion circle starts to become larger than the pixel size from around f10.



Frank

[/quote]

The Nyquist resolution is approximately 103 lp/mm, so I don't see any problems here. Same resoltuion as low iso (iso 25-50) "normal use" B&W film.



Of course, I have mentioned this before, the whole diffraction thing is totally out of context. When looking at the Nyquist frequency, and at good old film resolution, there is a severe mismatch in all of these discussions: we used to shoot quite happily with film, without worrying about emulsion diffraction. And it suffers from exactly the same problem - silver halide grain size in an emulsion is very uniform in size, it just isn't lined up as nicely as sensels are.



Anyway, based on my calculation, using the Rayleigh criterion, sensor diffraction limit for the sensor is F/16, which only means that from that aperture onwards, towards smaller apertures, lens resolution due to lens diffraction is less than the sensor resolution. As we still have the lens formula at F/16 we will still end up with a system resolution of 51 lp/mm, which is professional type resolution from a 35 mm size image on film, and 1.25 X to 2.5 X higher than an enthousiast would be able to capture. Note that on film we would capture these resolutions at much larger apertures, liek F/4 and F/5.6 (all due to system resolution and development and printing process): it is therefore quite a step forward



In short, no need to worry. If anything, it is to be applauded.



Kind regards, Wim