Micro-Nikkor 70 mm f/5
The Micro-Nikkor 70 mm f/5 was produced in the early and mid-60's by Nippon Kogaku (, Nikon's earlier name until 1988) in a relatively small number of specimens, perhaps not exceeding 500. It seems to have preceded the Macro-Nikkors, as well as the Ultra-Micro-Nikkors.
This lens was designed primarily for microfilming documents. Nippon Kogaku started designing and producing copying lenses in the mid-50's, because lenses designed in the West for the same purpose, although adequate for copying printed materials in Western alphabets, had an insufficient resolution to reproduce the complicated Japanese kanji characters. Subsequently, this technology was refined to produce the first lenses used to print the photomasks of early integrated circuits, which became the primary purpose of Ultra-Micro-Nikkors. The Macro-Nikkors, on the other hand, were a series developed for macro photography and photomacrography with the Nikon Multiphot apparatus, but also sold separately.
This lens has a metal barrel, apparently built entirely of brass alloy. There is no focusing helicoid, and the lens attachment is a standard 39 mm thread. The design of the aperture ring is similar to the Ultra-Micro-Nikkors it precedes, and unlike the scalloped aperture ring of the pre-N series of EL-Nikkor enlarger lenses, which is thinner and has more rounded contours. This lens was sold in a stained wood box with chromed hinges, internally lined in red velvet, quite similar to that of most Ultra-Micro-Nikkors.
In addition to the aperture diaphragm, this lens has a fixed internal stop immediately at the rear of the diaphragm. The diameter of the front element is almost enough for an f/2.8 maximum aperture, but the diaphragm and fixed stop limit the maximum effective aperture to f/5. This construction almost entirely eliminates vignetting and light fall-off at the edges of the picture, and is found also in other high-end lenses (e.g., the Zeiss S-Planar 74 mm f/4).
The black baffle at the rear of the lens unscrews to reveal a double adjustment ring, used to finely calibrate and lock the rear lens element during factory adjustment. This is a feature rarely found in consumer lenses, and reserved to high-end optics. The rear baffle is identical to the one of Ultra Micro-Nikkors. The baffle protrudes rearward when this lens is mounted on a DSLR, but it is not necessary to remove it in order to focus to infinity (the camera mirror may hit the baffle if this lens is mounted directly onto a DSLR - albeit, the lens cannot be used at all in this case, because its focus would be beyond infinity). By using specially-made bellows, I was able to focus to infinity with this lens and still have several mm of bellows extension to spare.
As far as I know, there are only four or five Micro-Nikkor models without a focusing helicoid (there are a plethora of EL-Nikkor, Apo-Nikkor, Apo-Macro-Nikkor, Macro-Nikkor, Ultra-Micro-Nikkor, COM-Nikkor, Nikkor-P, Nikkor-Q , Fax-Nikkor, Fax Ortho Nikkor, Xero-Nikkor and other series of Nikon lenses without a focusing helicoid, but they are not marked Micro-Nikkor):
The serial numbers of the black-barrel Micro-Nikkor 70 mm f/5 are 705 (i.e., focal length + aperture) followed by three more digits, probably starting with 001. I have not seen any specimen with the last three digits reaching or exceeding 500.
The principal technical data, as reported on web sites (I have no Nikon documentation of this lens), are as follows:
The standard magnification and image circle are appropriate for copying a single typewritten page onto a 36x24mm frame of 35 mm film.
If the resolution reported above is indeed expressed in line-pairs per mm and applies to the projected image (which is the simplest and most common standard measurement for resolution), it means that this lens has a resolution roughly twice the linear pixel density of a Nikon D200, or roughly four times the actual linear resolution of this camera (as computed from theoretical considerations). A Bayer sensor actually uses four adjacent photosites to record a pixel, spreading and averaging the light across these photosites with an anti-aliasing filter. The camera subsequently interpolates detail smaller than this area, based on the luminance values of the component colour of each photosite, and other information. Therefore, the linear density of pixels in a Bayer sensor does not directly correspond to the sensor resolution. As discussed below, however, the final image does contain some detail at a smaller scale than this group of photosites.
In practice, a lens with the resolution of the Micro-Nikkor 70 mm would suffice to bring out all the detail that can be recorded realistically by a 24 megapixel DX sensor, or a 48 megapixel FX sensor. In fact, this lens can cover a sensor area almost twice as large as an FX sensor, with a total resolution corresponding to about 90 megapixel. With a computed total of 6,900 line-pairs across the image circle, the Micro-Nikkor 70 mm places well within the resolution range of Ultra-Micro-Nikkors (4,800 - 24,000 total line-pairs across their respective image circles, as computed from data reported in http://akiroom.com/redbook/nhs2004/lens%20spec.pdf), and way above what is produced by virtually all modern "pro" macro lenses.
Distortion is relatively large for a precision lens, although less than that of some Ultra Micro-Nikkors, and very hard to spot in actual pictures.
Not surprisingly, the oversized front lens element and internal stop mean that vignetting is practically absent.
Initially, I shot an apple tree in available daylight from a distance of approximately 8 m. As a comparison, I used also an AF Nikkor 50 mm f/1.4 D, and an AF-S Nikkor 18-70 mm f/3.5-4.5 G at 70 mm. Figure 2 (left column) shows crops of the centre of the image for all lenses at f/8. The two latter lenses are regarded as quite good. Aside for a small dark spot that mysteriously appeared on the apple during the test of the 18-70 mm (perhaps an insect?), and the obvious differences in scale due to differences of focal length, I can see no significant difference in resolution, contrast o colour rendition. The only interesting difference is that the 18-70 mm at maximum focal length is not a 70 mm, but actually closer to 65 mm. It is a common practice among manufacturers of current zoom lenses to overstate the zoom range of their lenses by as much as 15%, so Nikon is not an exception. In conclusion, this test shows that the Micro-Nikkor 70 mm f/5 is at least as good as other good Nikkor lenses when used at reproduction ratios far lower than the declared usability range. If there is any difference, it is too small for a D200 to pick up.
For close-up tests, I used a simple test procedure I described elsewhere. Focusing is not easy with the Micro-Nikkor 70 mm, because of its narrow f/5 opening. I had to rely on the focus-indicating dot in the viewfinder, rather than on my eyes. Fortunately, f/5 is still sufficient to provide enough light for this indicator to work reliably.
My first test in the close-up range was at 1:12, i.e., the optimal reproduction ratio. I am not showing results for this, because the resolution was very high up to f/8, with a very slight drop at f/11 and a gradually higher one from f/16. This is perfectly normal and due to diffraction, and no lens can avoid it. In any case, at this magnification the Micro-Nikkor 70 mm performs better than any of my other lenses, and quite a bit better than the Micro-Nikkor AF 60 mm f/2.8 D at all apertures. In the table of apertures and magnifications I computed from theoretical considerations, I actually stated that all apertures, up to and including f/22, are usable without a visible loss of resolution in this magnification range.
The main reason why the Micro-Nikkor 70 mm actually does degrade past f/11 is mainly due to the fact that this lens is so much better than any other I have in this range. It is also interesting to note that the D200 interpolation algorithm seems to be able to extract at least some real detail from each photosite, rather than just averaging four adjacent photosites and running a sharpening algorithm on the results. If the lens did not provide real resolution at the photosite level, no algorithm would be able to extract any detail. This also means that the anti-aliasing filter does not completely average light onto a group of photosites, but partly retains some of the original differences. Therefore, my generalization in the above discussion is a bit excessive, and a lens with very high resolution does make a difference even at the individual pixel level of the final image (as opposed to the level of the groups of photosites sharing the light diffused by the anti-aliasing filter).
In order to see any faults of this lens, I subsequently tested it at 0.25x, near the higher limit of the magnification range stated as usable. Figure 3 shows some differences in colour balance, which are artifacts of the testing process (incandescent lighting used to frame and focus has mixed with flash light, in different proportions when shooting at different apertures). There is a slight loss in resolution beyond f/5.6, and a noticeable one from f/16. At this magnification, the lens is still working at an optimal level, and exceeding the resolution of the camera when used fully open.
At 0.5x, the Micro-Nikkor 70 mm still performs relatively well, albeit not at the level of the best macro lenses, and no longer exceeds the resolution of the camera (Figure 4). The highest resolution is actually reached only at f/8 to f/11. Below f/8-f/11, the image is degraded mainly by chromatic aberration, and above this value by diffraction.
When reversed with an EL-Nikkor reversing ring, the Micro-Nikkor 70 mm is very good for photomacrography. Figure 5 (left) shows the results at 2.5x. A higher magnification would require a bellows extension higher than that provided by Nikon PB-6 bellows. The results are comparable with those obtained with a Zeiss Luminar 63 mm (Figure 5, right).
Finally, I compared the Micro-Nikkor 70 mm with the Zeiss S-Planar 74 mm f/4, which, in previous tests, gave me the the best results around the 1x magnification range. I tested both lenses at 0.25x (Figure 5), because this range is acceptable for both lenses. I am displaying only the results at f/8 (Figure 6), because the two lenses give absolutely identical results (yes, I did verify that the two images are from different lenses - I also believed at first that I had used the same picture twice).
The Micro-Nikkor 70 mm f/5 is a top performer, on par with the Zeiss S-Planar 74 mm f/4 where the usable magnification ranges of these two lenses overlap. In the close-up range and for copying documents, the Micro-Nikkor 70 mm is unrivalled among my lenses. The only minus of this lens is the difficulty in focusing due to the narrow aperture. On the other hand, the Micro-Nikkor 70 mm offers no visible advantages to other good lenses, when used for general photography outside the close-up range.