Nikon's discussion of D800 vs D800E

[rjlittlefield]

I ran across this page today and thought it provided an interesting viewpoint on the choice of D800 vs D800E, that is, 36 megapixels in full-frame, with and without an anti-aliasing filter.

nikonusa.com: Moiré & False Color

I was particularly struck by their summary that:

For the vast majority of photographers who shoot a wide variety of subjects, shoot hand-held as well as with a tripod, use a selection of NIKKOR lenses and shoot at all aperture settings, the D800 and its 36.3MP using the OLPF will be the ideal choice.

For a few specific types of photographers—studio, commercial and still life—who have used medium or large format cameras, the D800E, for instance, will be the right camera for their needs. These photographers can control light, distance, aperture selection and their subjects to the degree where they can reduce the occurrence of moiré. This is because medium and large format digital cameras do not utilize an OLPF, and so photographers who shoot with them already have a working understanding of the added workflow needed to counteract or correct for moiré/false color. These photographers understand that extra time will be needed in the creation of each image, whether it is by solving any issues of moiré/false color during the shooting process, in processing the RAW image files, or using software to fix the image files.

We've had some discussion of this tradeoff before. I'm reminded of HERE and the surrounding thread.

Has anybody here studied this head-to-head for themselves?

--Rik

[DQE]

diglloyd.com has studied this in some detail through photographic tests with various subjects. I personally believe that most of his equipment tests and studies are well executed and provide useful information.

I'm reluctant to summarize his results here since he's explicitly asked that we subscribers not repost his findings elsewhere, and he works on an annual paid subscription model. He is very much an enthusiastic user of the Nikon 800E as long as one uses very high quality lenses and as long as one uses high quality photographic techniques.

Here's the Table of Contents, which AFAIK doesn't require a subscription:

http://diglloyd.com/prem/prot/DAP/index.html#NikonD800E

I hope this woefully incomplete information is of some use.

[lothman]

could a camera without AA-filter like the Nikon D800E or Pentax KIIs help for macro shooting when stepped down a bit further. So using diffraction as an AA-Filter and therfore beeing able to squeeze out a bit more od DOF without loosing image quality?

[lothman]

no opinions?

[rjlittlefield]

Here's the Table of Contents, which AFAIK doesn't require a subscription:

http://diglloyd.com/prem/prot/DAP/index.html#NikonD800E

Unfortunately, it does seem to require a subscription. On trying to visit that URL, I get an immediate "Authentication required" popup.

could a camera without AA-filter like the Nikon D800E or Pentax KIIs help for macro shooting when stepped down a bit further. So using diffraction as an AA-Filter and therfore beeing able to squeeze out a bit more od DOF without loosing image quality?
...
no opinions?

Beats me. I'm still unclear on how much improvement is provided by removing the AA filter if the development process is optimally tuned in both cases.

--Rik

[DQE]

Here's the Table of Contents, which AFAIK doesn't require a subscription:

http://diglloyd.com/prem/prot/DAP/index.html#NikonD800E

Unfortunately, it does seem to require a subscription. On trying to visit that URL, I get an immediate "Authentication required" popup.

could a camera without AA-filter like the Nikon D800E or Pentax KIIs help for macro shooting when stepped down a bit further. So using diffraction as an AA-Filter and therfore beeing able to squeeze out a bit more od DOF without loosing image quality?
...
no opinions?

Beats me. I'm still unclear on how much improvement is provided by removing the AA filter if the development process is optimally tuned in both cases.

--Rik

Rik,

I'm not sure what's going on with respect to free access to the site's TOC. I was sure that I had signed out and thus I thought that the TOC was not behind the paywall. Maybe there was a leftover cookie on my PC or something?

After confirming your finding that the link I provided is behind the paywall, I was able to access the TOC here, without logging in.

http://diglloyd.com/index-free.html#ReadingGuides

<EDIT>

and here:

http://diglloyd.com/index-dap.html

I subscribe to a couple of his modules and find them worthwhile and technically competent. Heck of a lot more informative than most of the standard photographic magazines I subscribe to!

[rjlittlefield]

Here is a publicly available report from dpreview:

http://www.dpreview.com/reviews/nikon-d800-d800e/27

The review covers both the D800 and D800E. Sections 27-29 of the review provide head-to-head comparisons of the two cameras.

Quick summary is that when processing from raw files, the D800E has a small increase in both sharpness and color artifacts.

But interestingly, in-camera JPEGs shot with "standard settings" are markedly sharper and more detailed from the D800E. See http://www.dpreview.com/reviews/nikon-d800-d800e/29 for that comparison. As far as I see they do not evaluate whether there are different JPEG settings for the D800 that would make it be more like the D800E.

--Rik

[DQE]

The reasonably careful comparative tests provided at the subscription based camera and lens testing site I linked in my post in this thread showed that only *very* careful focusing at optimum apertures when using the best lenses will benefit from the extra pixels of these camera bodies. The author didn't do much testing for macro applications though.

[rjlittlefield]

...only *very* careful focusing at optimum apertures when using the best lenses will benefit from the extra pixels of these camera bodies.

Hhmm... And yet it's clear that a merely decent f/11 lens produces an optical image that outresolves the pixel density of a Nikon D800 by a factor of almost 2. f/5.6 is way beyond that. The test images are HERE and HERE. At the D800's pitch of 4.89 microns per pixel, even a theoretically perfect sensor can resolve only 102 line pairs per mm, and that's for detail that lines up ideally with the pixel positions. 30% MTF for an f/5.6 lens is around 195 line pairs per mm -- almost twice what the D800 could possibly do under ideal conditions, probably more like 3 times what it does in practice.

The author didn't do much testing for macro applications though.

Probably none for focus-stacked, which would have taken care of his concern about careful focusing.

But certainly the issue of careful focusing is important at these resolutions. The old standard CoC criterion of 30 microns for full frame translates to a whopping 6 pixels on the D800. If you really want to exploit those pixels, you have to throw away the old DOF scales and make new ones about 3 times finer.

--Rik

[Steve S]

I have no personal experience with the D800, but have made many thousands of exposures with the D800E, and have made comparisons between the D5200 and D7100; the latter two seem to have slightly different sensors but the same sensel pitch, the latter lacking an AA-filter. DXO assigns both sensors very nearly identical sensitometric characteristics. Dpreview says there may be very slight differences in resolution around f/3.5-f/4, but none at f/5.6 and beyond.

Examining raw files processed by Lightroom 4, of identical subjects taken with the same lighting, subjects with depth so differences in focus can be seen, I can see no difference whatever between the two APS-C cameras at f/5.6. Trying a Nikon 50mm f/1.4G at f/3.5, and a Nikon 60mm f/2.8G macro lens at f/4 on an array of USAF 1951 targets spaced 50mm apart at a distance of 8m, I have the indistinct impression that there is slightly more “microcontrast” with the D7100, and very slightly more color aliasing on barely-resolved line triplets. There is certainly no truly appreciable difference. Of course we do not know that Lightroom is handling the files in exactly the same way; I haven’t tried renaming them in the EXIF, etc.

If Nikon announced, “Surprise! There is no AA-filter in the D5200 either”, I would not be among those much surprised.

Nikon has also recently announced a fixed-lens APS-C compact, the “A”, which they say lacks an AA-filter. It would certainly appear that Nikon is on the road to abandoning such filters on high-Mpx cameras. It is said on the web that small-format digital cameras typically lack such filters, owing to the small sensel spacing.

It seems to me that a sort of consensus developed on the web about the differences between the D800 and the D800E that there was a detectible difference, but that different sharpening might largely overcome it.

By the way, I cannot recall an “operational” D800E exposure that was appreciably degraded by sensel-level color aliasing or moire.

I am a subscriber to some of the product of Lloyd Chambers, and strongly agree with him (and with the Englisman Tim Ashley) that in landscape photography most lenses are afflicted with curvature of field that degrades images on high-Mpx cameras. And of course phase-detect autofocus leaves much to be desired. Consider yourselves lucky, macro-stackers!

As to DOF tables, I’ve been stewing about whether to print and laminate some with CofC=10 or 15 microns; leaning towards 10.

Steve

Edited to correct 50cm to 50mm. Duh.

[DQE]

I really wish Canon would make some high-megapixel DSLRs, including some without the anti-aliasing filter.

Yet I also have the impression from online reviews that D800x cameras may also have lower pattern noise in the sensors. Also,some tests indicate that they may have lower noise at high ISO speeds. If so, these factors would also influence significantly image quality in dark areas in many cases. Some of these concerns may be less important in macro since we can usually use enough flash exposure to allow us to work at low ISO speeds and if necessary overexpose to benefit dark areas.

(I need to review Rik's links to dpreview.com tests - theirs seem reasonably well-controlled and carefully designed.)