Nikon D810

The limit at which diffraction starts to soften the image on the sensor (DLA) is independent of the lens, and only depends on the pixel pitch. DLA is the aperture at which the first null in the diffraction pattern ("airy disk") lands at the edges of the closest adjacent pixels. For smaller apertures, the main lobes of the airy disks start to merge, and pixel-pixel contrast starts to degrade. At an aperture half of the DLA the airy disks are centered at twice the pixel pitch.

Adding a lens to the discussion is more complex. A perfect lens is easy to analyze. Since it has no aberrations it is by definition diffraction-limited at all apertures, and interactions between the lens aperture and camera are ideal. But if the lens is not diffraction-limited its optimum aperture becomes important. If the optimum aperture is larger than the DLA, then it is best to operate the lens at the DLA to simultaneously optimize depth of field and sharpness. This is often the case for lower magnifications, as the effective aperture is larger than for higher magnifications (fe=fn*(m+1)).

For the case of your 105VR, at around 1:1.1 you found the optimum aperture to be f14. This is an 87% reduction vs the DLA of the D810 (f7.5) so will result in enough diffraction blur to reduce the sensor linear resolution by 46%, from 36MP to less than 20MP. The bleed-over of the airy disks will reduce the sharpness of the image at 100% pixel level significantly.

If you re-take the images at f7, it may or may not improve the sharpness, and it actually might degrade it since the lens will be aberration-limited at f7.

Regarding the question about why you went for he D810, for some reason I thought you already had a D800 and were buying the D810 as an upgrade. I fully understand the upgrade from D300 even if you don't intend to use EFCS often.

One more note: in the case of the 105VR, I just finished a slew of test shots, with everything dialed-in to the test subject (including micro AF calibration). With EFCS enabled, image sharpness is considerably better at every aperture/shutter speed. In fact, f11 @ 1/30 w/EFCS is the sharpest of any shot from f7.1 @ 1/80 up to f20 @ 1/10. On my 60mm micro, the best result is f9 w/EFCS @ 1/30. Quite obviously, EFCS reduces shake even at 1/80s--that surprised me. These were shot ~1:1 ISO 100--your results may vary to mine.

These are the kind of 'rubber hits the road' tests that convince me I'll use EFCS for macro whenever possible.

....and that, in a nutshell, is why you hear us constantly harp on the subject of EFSC. It (the lack thereof) is one fundamental ill which has to be cured before you can begin to explore the actual limits of coin macro photography or even the lens you might already own.

I can't see the image.

I generally don't use AF as the critical focus points are rarely the ones chosen by the AF algorithm. If you're at f14 it may not matter much though since DOF is double what it would be at f7.

Comparing the f7 vs f20 images there is little difference, and indeed the f11 or possibly f16 look the same or slightly better. This means the lens aberrations of the 105VR pretty much degrade the image at f7 as much as diffraction does at f20. This is an interesting quality of the 105VR that I had not noticed before, and must have been a design goal, ie to make the image quality pretty much independent of aperture.

A quick update: I see that Nikon's Software Developer's Kit for the D810 was released in mid-Sept. Which means that tethering software can now be updated with D810 functionality. I'll post my results on the D810 and the EFSC when it becomes available.

A question( I don't have any idea of what the answer is). How much picture degradation is caused by the low bypass filter (also known as anti-aliasing)? Background to question My camera doesn't have a EFCS or a Low by pass filter. I did quick comparison with my other camera that has a smaller sensor EFCS and a low by pass filter. With the same lighting my camera with out the bypass filter produces better shots of coins
The Nikon D810 doesn't have a low by pass filter. It has the same sensor( sony made) as my camera.

Most of the tests I've read about compare the D810 to the D800e, which has an OLPF.
A lot of "tests" I've seen are essentially opinions from pixel-peeping--as are my results.
A few sites are pretty careful at focusing on measurable results, like here.
I'm waiting for DXO Labs to test Nikon's lenses on the D810, and see how that compares to their D800/E results.