If you want to shoot the same image with both FF and DX (i.e. FOV
and DOF both the same), then the maximum megapixel count
achievable before the diffraction limit becomes an issue is EXACTLY
THE SAME.
Yes, but you are ignoring the optical advantages of stopping a lens
down to f/4 on a DX to achieve the same DOF as a f/5.6 on a FF.
Having played with 17-55DX and 17-35FF I would say that 17-55 becomes very sharp at wider apertures than 17-35, i.e. its sweet spot is at wider apertures. Not having this kind of money myself and using Sigma 18-50 f2.8 DC instead, I would say the same thing: Sigma 18-50 f2.8 is sharp on wider apertures on average than Sigma 28-70 f2.8.
In other words, I don't
want to stop down more. First, because I choose aperture according to the amount of DOF I am after and second, because with modern DC/DX lenses it doesn't buy me much anyhting.
If you want an even narrower DOF this advantage really reduces
DX format to a point where even the best lenses are inadequate.
Yes, if I want to have really narrow DOF then indeed with DX I can't achieve it, because lens don't open up that wide. But this has nothing to do with diffraction. Yes, this is another DX limitation besides noise. I don't think it really affects too many people though.
As I said earlier (if you had taken the time to carefully read it),
the diffraction limitations of smaller photosites mean you can't
stop down a lens on a DX sensor much past f/11 and you can get to
f/16 on a FF sensor (of course, optics will once again come into
play).
Yes, but in order to get the same images with DX I
don't*want to stop down as far as with FF bodies. I can get the same
DEEP DOF for wideangle landscape shots at a stop wider apertures. Still remaining exactly as far from a diffraction limit as I would be had I shot the same landscape shot with a FF camera.
Finally, setting both cameras at f/8 and shooting and focusing to
infinity you can print considerably larger on a FF sensor than you
can on a DX sensor.
Why on earth would I want to do that? In that situation the cameras would render two different images (DOF of DX shot would be much deeper, i.e. nearby grass would be much sharper than on FF shot). Why would I try to compare image quality of two differently rendered images?
* * *
Anyway. I do agree that even though the sweet spot of DX lenses is at wider apertures than FF lenses usually have, it won't completely negate the slight absolute resolution advantage FF cameras will have at some point in time in the future.
Right now though, there are plenty of good lenses around which can eke every last drop of resolution out of a 12MP DX sensor. You just use different lenses and different apertures than you otherwise would with FF bodies.
After sleeping on it I have to say I am sorry about the way I said it myself.
But it is terribly frustrating that every week or so someone comes and tries to use bogus arguments to argue that FF is superior to DX and the problems Nikon faces. Two of those are: diffraction limit of DX and Nikon F mount somehow preventing Nikon from producing a good FF camera.
Both are factually wrong.
Diffraction limit just isn't any worse for DX because you always want to shoot at wider apertures with DX than you'd want with FF. Because of DOF. And as a result, you will be as far from diffraction limit as you would be with FF cameras.
As to F mount... On all comparable Nikon and Canon lenses, lens exit pupils are exactly the same size, i.e. the light hits sensors at exactly the same angles for both Nikon and Canon lenses. Even though Canon mount is wider, 99% of Canon lenses don't actually use it to any advantage.
Having read both of these arguments many a time, this is tiring, boring, and aggravating and I do get angry reading either one of them again.
There are reasons why FF is superior. Diffraction limit just isn't one of them.