From IR
For the D800...FF
Quote: Our laboratory resolution chart revealed sharp, distinct line patterns down to about 2,700 lines per picture height in the horizontal direction, and about 2,700 lines per picture height in the vertical direction in JPEGs. (Some might argue for higher, but aliasing artifacts begin to appear at those resolutions.) Complete extinction of the pattern didn't occur before the limits of our chart. We were able to extract a little more horizontal resolution with RAW files processed through Adobe Camera Raw 6.7, about 2,900
D7100 APS-c
Quote: Our laboratory resolution chart revealed sharp, distinct line patterns down to about 2,600 lines per picture height in both the horizontal and vertical directions. Extinction of the pattern didn't occur until around 3,600 lines in both directions.
Canon 5D mk 3
Quote: Our laboratory resolution chart shows strong detail with distinct line patterns down to about 2,400 lines per picture height horizontally and to about the same vertically from in-camera JPEGs. Extinction of the pattern occurred just past 3,400 lines horizontally and vertically.
K-5 II
Quote: Our laboratory resolution chart revealed sharp, distinct line patterns down to about 2,200 lines per picture height in the horizontal direction, and about 2,100 in the vertical direction in best quality JPEGs. Complete extinction didn't occur until about 2,700 lines in both directions. We were able to resolve a little more with an Adobe Camera Raw conversion, about 2,300 lines in the horizontal direction and about 2,200 in the vertical, and complete extinction of the pattern was extended to about 3,000 lines.
If the discussion is lw/ph, there is a small case to be made for the D800, if you can see it or not would have to be determined by examination of the two images.
If you were to put four K-5 images in pano mode next to a D800 image, you would get almost 10,000 by 6000 pixels and 4600 by 4400 lw/ph.. you can't get anywhere close to that with a D800. Once the K-3 comes out the difference at low ISO will be pretty negligible.
Also not as well as being darn close to a D800 the D7100 surpasses a canon 5d MK III in resolution. I'm sure you can get more resolution out of a D800, but only buy buying very expensive lenses. And even then, you'd have to use the same very expensive lenses on the D7100 to demonstrate the advantage.
In my mind, I have to see it in the images is still a valid point. Until someone can prove the human eye can see the difference between 2600 lw/ph and 2700 lw/ph I'm not sure you have a point.
I'm not advocating these numbers as accurate, I'm simply making a point. The point being that even in a lab setting, with a single set of tests, a K-3 should be darn close to a D800 some of the time, just as it happened with the D800 and D7100 in this set of tests.) The difference between FF and APS-c is barely noticeable until you go over 1600 ISO and 24x 36 print size. I've also seen tests where people have squeezed 3600 /ph out of a D800, useing an exceptional lens (as in over 2k) and over 2900 out of a K-5. As a general rule I've sort of come to the conclusion that with exceptional glass you can squeeze 70% to 85% of the actual number of pixels out of a sensor. So if you have a K-5 and at 3200 pixels deep getting 2900 lw/ph is phenomenal, and it's not a realistic expectation in the field. The number quoted from IR above is 71%.
A D800 with 4900 pixels deep, should be capable of 3500 lw/ph. (And that's more typical of the lens tests I've seen for high end lenses, I have no idea what happened to the IR test, maybe it's a typo) But , not everyone agrees that more lw/ph after a certain point makes a better image ( look at a DP2 image compared to a D800 image printed at 20x30 or even a K-5). You have to do a test comparison to see if the higher lw/ph is what you want.
My suggestion is still, without test images you can't decide what you want. You can read all the specs you want, you can speculate what those numbers mean to your eye, but until you actually see the difference you're just speculating.
It's pointless talking about lw/ph,if you can't see the whole image large enough to see a difference. And even the most generous tests 2600 to 3600 give a D800 only a 31% advantage over a D7100 in lw/ph. Once you get north of 2500 lw/ph what does a 31% increase in lw/ph look like in print? My best guess is it's not noticeable you view the images on a screen over 4000 pixels wide or until over 24x36 print size, and even then, you have to have the prints side by side to see the difference.
On images less than that, the image taken by the best photographer will be the best image, on images more than that, there is no conclusive evidence either way as far as I can tell.
Taken with a K-5 and Sigma 8-16, the level of detail is ridiculous, some might argue, un-necessary. The question is, would a D800 image look better, or even more ridiculous?