On his test page he noted that "This lens shows the highest center of the field sharpness and resolution that I have measured across this magnification range (except maybe at the highest mag (m=1.95)." His tests show that at 1:1 it has perfect field flatness, and it also has almost no lateral or longitudinal chromatic aberration, even wide open. In short, this is one of the best, if not THE best, lens for color macro photography at m=1 on a FF sensor.
This is a rare lens, but there is an even more scarce special Nikon lens that could be even sharper, judging from the specs. This is the incredible Repro-Nikkor 85mm f/1.0. Bjorn Rorslett says of this lens: "Another of those elusive industrial-type Nikkors from the heyday of the hippies, the 85/1 was optimised for 1:1 magnification and lacks any focusing means whatsoever...Images, even with the lens set wide open, have a crispness and bite that makes you wonder where the optical evolution is headed these days, since 30+ year old designs could give these stunning results....It makes a perfect macro setup for field use with or without a 1.4X TC added to it. Another application is deploying it as a relay lens for various non-retrofocus fisheye lenses (6, 7.5, 8, and 10 mm)." (http://www.naturfotograf.com/lens_spec.html)
Its wider aperture means that it will show less diffraction than the PN105A, but lens aberrations might mask that advantage. A head-to-head test is the only way to know for sure which of these great lenses is best. To my knowledge no such head-to-head test of these exotic lenses exists yet.
I recently bought a good copy of the Repro-Nikkor, with the intention of testing it against the PN105A and selling the one that was less perfect. These are my tests to decide the question.
Magnification is exactly 1.0 (to two decimal places) for all of the shots in these tests, for both lenses. I am using an Olympus PEN F MFT format sensor with 20Mp. Though some have reported that the RN85 is difficult to adapt to mirrored cameras, I found it is easy to adapt to Nikon cameras (and hence to mirrorless cameras using their Nikon adapters) via a 48-52mm Heliopan step-up ring and a standard Nikon BR2 reversing ring. By pure luck this gives exactly m=1.
Before we compare the two lenses, let's see how the RN85 performs wide open (f/1.0) versus f/2.0 vs f/2.8. Here is an unsharpened stack of unprocessed jpgs at f/1.0 (the shutter speed was 1/800s!):
Below is the same subject with the lens stopped down one stop, to f/1.4:
Below the lens is stopped down to f/2.8:
And here is a composite of 100% center crops at f/1.0, f/1.4, and f/2.8:
The f/1.0 result is slightly soft but has no color fringing at all. By f/1.4 it pretty good, and at f/2.8 it is about as sharp as it can get.
Note that this "100%" refers to tiny MFT pixels; if this were taken on a FF sensor with 42Mp, a 100% crop would look more like this (f/1.0, f/1.4, f/2.8):
Since the quality is so high, it is worth presenting the 400% crops of the center of the MFT image (f/1.0, f/1.4, f/2.8):
At f/2.8 the image is pixel-perfect. Even at f/1.4 it is very good.
Below is a composite of 100% corner crops of the the MFT image at f/1.0, f/1.4, and f/2.8:
On an MFT sensor, this lens produces similar resolution at center and corner.
In spite of the good performance, though, this series (together with the f/2.0 results, not shown) shows that the lens is not diffraction-limited below f/2.8. The quality improves as the lens is stopped down to f/2.8. This means that although the RN85 is a spectacular lens, the PN105A could still come out on top. In the next post to this thread I will compare the two lenses, using a more extreme target, with strong backlight and transparency. This should produce lots of CA, if anything can.
