Photomacrographic lenses, part 5
Macro-Nikkor 65 mm f/4.5 and 12 cm f/5.6,
Zeiss Luminar 63 mm f/4.5,
Minolta 100 mm f/4,
Tomioka Tominon 75 mm f/4.5,
Canon 35 mm f/2.8
This page compares the performance of a few photomacrographic lenses, ranging from the very best and expensive to obscure and cheap. In the first category are the Macro-Nikkor 65 mm and 12 cm, as well as the Zeiss Luminar 63 mm (of the first generation).
While in earlier tests I was concerned mostly with resolution, in the present test I also took into account contrast. For this purpose, I placed the test subject (a centipede) on a large, mostly white absorbent paper background. This was partly because I am aware, from personal experience, that a few photomacrographic lenses are quite bad in these conditions (my specimen of the Olympus OM 20 mm f/2, for instance, is quite unusable with this general setup because of low contrast and internal flare), and partly in order to verify assertions I read on other web sites about Zeiss Luminars exhibiting a poor contrast and internal flare with large white backgrounds, in spite of their very high resolution (which had not emerged from my earlier tests and practical use of these lenses). Unless indicated otherwise, no post-processing was used on the test images.
As a light source, I used a Nikon SB-800 in iTTL mode. The test camera was a Nikon D200. All lenses were mounted on Nikon PB-6 bellows at maximum extension (207 mm), except for the Canon 35 mm, which was tested with an extension of 128 mm because of its focal length much shorter than the other lenses tested here.
Macro-Nikkor 12 cm f/5.6
The Macro-Nikkor 12 cm f/5.6 (see also here) is the lens with the longest focal length of this batch. Contrast is good, with no visible internal flare (which usually shows as a washed-out central area of the picture, and tends to be especially visible when stopping down the aperture). Resolution, however, is disappointing for a lens supposed to be among the very best for this application. A careful examination of this lens showed what appears to be either a large detached area between two elements at the very front of the lens, or an oily substance that somehow leaked behind the front element. This will be investigated in detail when I have the time - I disassembled the lens (which gave me 15 separate parts, not counting a dozen of screws and the diaphragm, which is best not taken apart), albeit the defective front cell cannot be further disassembled and I left it alone for now. I found indeed traces of oil, as well as instant glue possibly left by a clumsy attempt at servicing the lens (unless this is an unusual thread sealant formerly used by Nikon), where none should have been present. At least, now I know that this type of fault causes an unacceptable degradation of the lens resolution.
This specimen is obviously defective, and certainly not typical of the results that can be expected with this lens. Nonetheless, the above are my results.
Some time after these tests, I disassembled the defective front cell of this lens (since I had nothing to lose), which involved filing away a thin aluminium sleeve surrounding the back of the doublet. Kudos to Nikon for not using adhesives to keep the lens in place in its metal mount. The two elements were solidly glued together, so I dumped the defective doublet in a small plastic jar containing a mixture of methyl and propyl alcohol, to test whether this had any success in dissolving the glue cementing the two elements together. After 3-4 weeks I took out the doublet, and its elements came easily apart. It took a bit of light rubbing with ethyl alcohol to remove some residues of the adhesive still adhering on the optical surfaces, but the process left me with two very clean elements, which I remounted in their metal sleeve and fastened in place with a little black silicone around the rear edge. This allows me to easily remove the elements, if I should decide to try and re-glue them with Canada balsam at a later time.
This is a textbook example of the results of not following a common recommendation: you should never clean the front element of a lens by applying to it a large amount of alcohol or other cleaning fluid. This is probably what an earlier owner of this lens did, causing the solvent to seep into the lens mount. Left there for an extended time, where it could not easily evaporate, it found its way to the joint between the two lens elements and started to dissolve the Canada balsam holding them together. Eventually, some of the mixture of adhesive and solvent moved out of the joint while slowly drying up, and left large air "bubbles" between the optical elements, with the consequences seen above. Since the effect of the wrong cleaning procedure was not immediate, the owner of the lens probably did not realize that he/she had ruined a high-class lens, and may have blamed the manufacturer or some other cause for the poor lens performance.
I no longer have the original test subject, but a quick test with a different subject in similar conditions shows that the performance of the repaired lens is greatly improved. Perhaps contrast would be slightly improved by re-cementing the two elements of the front doublet, which should reduce the internal reflections. However, this does not seem necessary from the results of the above test, and I now have a functioning lens instead of a lemon.
Minolta Auto Bellows Macro 100 mm f/4
The Minolta Auto Bellows Macro 100 mm f/4 is designed for use on bellows. I did not find the technical specifications for this lens, but I assume it can be used at the range of bellows extension of typical macro bellows.
This lens has a Minolta MD bayonet and a rear element that projects quite a bit behind the lens mount. Attaching it to Nikon equipment required an MD to Nikon adapter with a built-in optical group, modified by removing the optics and machining a larger hole to accommodate the rear of the lens barrel. This adapter also takes care of changing the automatic diaphragm to manual.
Resolution is good, albeit not on par with much more famous photomacrographic lenses. Contrast is also good, in spite of the fact that the front lens element is very exposed to ambient light.
Tomioka Tominon 75 mm f/4.5
Tomioka is a large company with a history of producing sophisticated optical systems. Since it mostly marketed them through other companies, it is not well known among photographers. Tomioka produced a series of Tominon photomacrographic lenses, sold with Polaroid MP-3 and MP-4 photomacrographic and copy systems.The lens barrels, as well as the original packaging, are unusually plain and boring. Nowhere on the barrel it says that these lenses are for photomacrography, and therefore several of their current owners don't know it (some actually believe that they are special-effects vignetting lenses for large-format cameras, others that they are large-format wideangles - the latter obviously never tested these lenses).
These Tominon lenses have an unusual (for modern equipment) 40 mm attachment thread, designed to fit in a shutter similar to those used on large-format cameras.
The tested model is the Tomioka Tominon 75 mm f/4.5. Performance is better than expected. Resolution is good (albeit not excellent), while contrast is lower than in most other lenses tested on this page. However, decreasing the gamma in post-processing restores most of the lost contrast.
This lens performs much better when stray light is prevented from entering the lens. A purpose-built narrow lens shade can conveniently be mounted on the filter mount. As a whole, this is a cheap but quite usable photomacrographic lens. You should not pay more than about 50€ for this lens (its second-hand value ranges from 20€ to fantasy prices above 100€).
Zeiss Luminar 63 mm f/4.5
The Zeiss Luminar 63 mm f/4.5 (see also here) shows a very high resolution across the whole frame at the current magnification (close to 5x). Contrast is good fully open, and slightly lower closed one stop. In spite of this, this is among the best of the lenses tested here. I can find no confirmation of rumors of it being prone to low contrast and internal flare (at least, not at the apertures that give the best resolution - it does get somewhat worse by stopping down further, but doing so also brings the lens past the optimal resolution limits imposed by diffraction, so this is not a defect that would become apparent in the work of a photographer who knows what he/she is doing). The tested specimen is from version 1 (the earliest from Zeiss), which should be the most sensitive to flare.
Macro-Nikkor 65 mm f/4.5
The Macro-Nikkor 65 mm (see here) is arguably the best photomacrographic lens ever made. It confirms its fame in this test. Resolution is very high across the whole frame (unlike the Luminar 63 mm, which shows signs of giving up at the edges when used at low magnification). Contrast of this and the preceding lenses is the highest among the ones tested here.
Canon Macro 35 mm f/2.8
The Canon Macro 35 mm f/2.8 is characterized by an unusually short focal length (not matched by a high maximum magnification - this lens is not recommended by Canon past 5x) and a conical front portion of the barrel that facilitates the positioning of light sources. This is especially desirable, because the working distance of this lens is quite short. The Minolta 100 mm discussed above has a similar barrel, which is quite unnecessary because of the much higher working distance (albeit it does not hurt).
This lens has an RMS attachment thread, like the Zeiss Luminar discussed above. The barrel of the Canon lens is much wider than typical for RMS lenses, and also has a relatively long aperture lever that facilitates turning the aperture ring. The lever sometimes gets in the way of other equipment. This lever is a screw (devoid of any screwdriver slits on its head), and can be removed by unscrewing it. It appears to be fastened to the lens barrel with a strong thread sealant, and is quite hard to remove the first time. The most effective way for doing so is clamping the body of the lever, padded by wood pieces on either side, into a vice, and turning the lens barrel appropriately by hand.
Resolution at 8x is still good, but contrast is definitely poor. This lens might perform better with a lens shade, but building one is problematic, and would reduce the already short working distance. For this reason, this lens is not a good choice for photomacrography. Its second-hand price also tends to be unrealistically high. You might spend somewhat more and get a much better Zeiss Luminar 40 mm f/4 (or a 25 mm if you need higher magnifications). Its only redeeming quality is the maximum f/2.8 aperture, which makes it easy to focus. Fortunately, the low-contrast area is uniformly spread across the frame instead of being concentrated in the middle like in the Olympus OM 20 mm, so it is quite possible to restore an acceptable contrast by manually adjusting the gamma in post-processing (above).
This lens has diaphragm blades with a V-shaped indentation in the visible portion of their edges, which produces a markedly zig-zag perimeter of the aperture. With the aperture closed down 2-3 stops (i.e., the optimal aperture for maximum resolution and depth of field), this leads to odd-looking out-of-focus highlights with the same shape, visible near the bottom of the test picture.
None of the above lenses are definitely poor (the main problem is the excessively low contrast of the Canon 35 mm, which, however, can be compensated for acceptably in post-processing). Predictably, the Macro-Nikkor 65 mm and 12 cm, and the Zeiss Luminar 63 mm are the best. If you want the very best and can afford it, these lenses are the obvious choice. Cheaper and more obscure lenses (and an expensive one from a famous brand - Canon) have contrast problems, but fortunately no concentrated flare hot-spot. The price-to-performance prize, however, is shared by the Tominon 75 mm f/4.5 and the Minolta 100 mm f/4. The first of these is much cheaper, the second has much better contrast than the Tominon, and both are overall quite good in terms of resolution.
Distribution of page hits (whole site) during the last month.
Provided by clustrmaps.com