Nikon D810

Assumption is dangerous; so read on with that in mind. Sony has released all the exif codes so that 3rd party lens makers don't have to back engineer to produce lenses for sony cameras...I had assumed that they have done the same with the codes necessary to produce apps for the cameras

First coin photo I've seen taken with the D810! Very nice as well.

Is the f8 you quoted effective or nominal aperture?

Thanks! This is my previous shot with my D300 @ f7.1 using the 60mm micro--no bellows or any converters. Not focus-stacked; a single shot.

In this case, a little too much because I dialed in -2/3 EV exposure compensation to keep details from blowing out under those Jansjo lights. I discovered that's totally unnecessary on "flat control", but I can recover tonal range pretty well in ACR. What I should be doing is taking a few test shots, checking out the histogram and adjusting to cover the full tonal range.


Correction: both pics were shot at f8, but using different lenses. The first one was shot on the Nikkor 60mm micro; the later on the 105 micro. I've found that for coins, the 60 micro resolves more specular highlights on metal objects--it's actually a little distracting. I don't set sharpening in the camera. That's my very last filter in post-processing, so applying curves to luminance and color (I usually work in LAB) doesn't amplify sharpening artifacts.

Yep, significantly sharper on EFCS so it appears to be working well. Why f14 though? Is that effective or nominal? The 100% pixel sharpness is pretty good so I assume f14 effective. Have you calculated the DLA for the D810? Do you know the pixel pitch?

Since f14 is well into diffraction-limited territory for the D810 sensor, the actual improvement is probably quite a bit better than what you're showing. That is good news. To really show the improvement you need to use a lens that is diffraction-limited at f7.5 at the image magnification.

I'm curious why you went with the D810 if not to get the added sharpness due to removal of the AA filter and use of EFCS?

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.