Review of Jupiter-3 1:1.5 F=5cm (1952, KMZ). How many different Jupiter-3s were there?

Material on the lens especially for Radozhiva prepared Rodion Eshmakov.

The lens block of the Jupiter-3 objective from 1952 in a new focusing mechanism.

The lens block of the Jupiter-3 objective from 1952 in a new focusing mechanism.


Jupiter-3 50/1.5 is a well-known old fast lens for rangefinder cameras, produced from the late 40s until the collapse of the USSR. The first lenses were called "ZK" - "Zonnar Krasnogorsky" - and were probably made using glass/lenses made in Germany. In 2016, KMZ organized the production of an updated version of the lens New Jupiter-3+ and optically identical Zenithar 50/1.5 E.

All reviews of Jupiter-3 lenses:

  1. Jupiter-3 1:1.5 F=5cm P (1952, KMZ). How many different Jupiter-3s were there?
  2. Jupiter-3 1:1.5 F=5cm P (1957, ZOMZ), M39
  3. Jupiter-3 1:1.5 F=5cm P (1963, ZOMZ), Contax RF
  4. Jupiter-3 1:1.5 F=5cm (1972, ZOMZ), Contax RF
  5. Lomography x Zenit New Jupiter-3+1.5/50 L39/M (KMZ, 2010s), M39 - detailed comparative review

Early Jupiter-3 lenses (before ~1955) and, especially, the "Sonnar Krasnogorsky" are now a collector's item and are rare, so the information concerning them is scattered and unreliable. In particular, one can often come across the opinion that early lenses differed significantly from later ones due to: different (usually German) glass, build quality (Stalin ate the children of a worker for an incorrectly machined lens), black magic - underline as appropriate. This article presents a rare example of a Jupiter-3 lens from 1952, manufactured at KMZ, and an analysis of the optical characteristics of the lens is performed, including in comparison with New Jupiter-3+, in order to finally answer the question: how many optically non-identical Jupiter-3 lenses were there in reality?

Navigation

Technical specifications

Optical design – 7 lenses in 3 groups, Sonnar;

Drawing of the optical scheme of the lens.

Drawing of the optical scheme of the lens.

Focal length - 50 mm;
Relative aperture - 1: 1.5;
Field of view - 45 °;
Frame format - 36 × 24 mm;
Aperture - 13 blades, without a preset mechanism;
Thread diameter for light filter – 40.5 mm.

Design and adaptation

I received the lens as a bare lens block without factory external decorative parts and other things - I would not have bothered to transplant the lens block if this rare lens had been included. In this case, it was necessary to come up with an optimal modification option to ensure maximum compatibility and ease of use - and therefore, the Canon EF bayonet mount was suddenly chosen as a mount to the camera.

Of course, the Jupiter-3 will not be compatible with Canon SLR cameras at all, but this bayonet mount is good because it allows you to easily attach the lens to any modern mirrorless camera via an adapter, and there are also adapters for Canon EF optics. tilt shift и shift adapters.

The focusing mechanism chosen was macrohelicoid M42-M42 17-31, providing a comfortable minimum focusing distance of about 25 cm. A special sleeve was custom-made to install the lens block in the helicoid. The lens aperture is controlled by rotating the lens block by the nose, onto which an empty filter frame was screwed for convenience. The appearance of the adapted lens is shown below.

Surprisingly, this old version of the Jupiter-3 does not have high-quality light protection: the first triple glue has no blackening on some surfaces and is noticeably glare, the aperture blades are shinier and lighter than those of the 1960s versions of the lens. There are also questions about the quality of the glass: even despite the lens coating with yellow glare, the lens is very yellow when viewed through.

Light transmission spectrum of the Jupiter-3 lens, 1952.

Light transmission spectrum of the Jupiter-3 lens, 1952.

A coating with a yellow glare color, which is logical, should lead to a shift in the peak of light transmission to the blue region, but in this case this is not observed, which indicates a very strong absorption by colored impurities (iron) included in the glass. In other words, the optical glass used in the lens has a very mediocre purity and too high absorption.

It turns out that in terms of build quality, the old Jupiter-3s have no advantages over the more common lenses of the 60s and 70s.

Optical properties: comparison of Jupiter-3 1952 and New Jupiter-3+

The New Jupiter-3+ lens is identical in all respects except for the anti-reflective coating and light protection. similar most frequently encountered lenses produced in the 1960s and 1970s, so the comparison with it is quite representative.

To assess the differences between the lenses, a series of paired photographs were prepared, taken under equal conditions on a full-frame camera. Sony A7sBelow are photos taken with the Jupiter-3 lens in 1952.

Next are pictures from New Jupiter-3+.

To my surprise, these two lenses turned out to be different in their pictures not only in color rendition and contrast, but also in sharpness and bokeh, which is definitely not related to assembly errors.

Comparison of bokeh of the lenses Jupiter-3 1952 and New Jupiter-3+.

Comparison of bokeh of the lenses Jupiter-3 1952 and New Jupiter-3+.

100% crops from paired photographs taken with Jupiter-3 1952 and New Jupiter-3+ lenses to assess image quality in the center and corners of the frame.

100% crops from paired photographs taken with Jupiter-3 1952 and New Jupiter-3+ lenses to assess image quality in the center and corners of the frame.

You can see that the bokeh of the old version of the lens has a much less pronounced scale effect. The shape of the disks is significantly different. At the same time, the old lens turned out to be significantly better than the new Jupiter-3 in terms of image quality in the corners of the frame, although at an open aperture it was slightly inferior in the center.
Therefore, although the lenses have the same name, they are not optically identical: they must have at least different lens geometry, but they can also be different in the materials used in the design.

X-ray fluorescence analysis of the first, last and fifth lenses of the 3 Jupiter-1952 objective showed that the lens materials are similar but not identical. Thus, the spectrum of the front element of the old Jupiter-3 shows a larger strontium signal compared to the spectrum of the New Jupiter-3+ lens. The rear lens of the old version contains antimony, while the new one does not. The spectrum of the fifth lens of the old objective contains an unexpectedly strong lead signal.

If the differences in barium and antimony content can be associated with changes in the composition of glass from different years of production due to the development of technology (antimony oxide, like arsenic oxide, is a brightening agent, barium oxide can be balanced by zinc oxide and other oxides), then the presence of large amounts of lead in the fifth lens cannot be explained in this way. The fact is that it is reliably known that the fifth lens of the serial Jupiter-3 lens was made of a special flint OF1, the main signature of which is not lead at all, but antimony.

XRF spectrum of OF1 glass lens.

XRF spectrum of OF1 glass lens.

It turns out that the Jupiter-3 lens of 1952 differs not only in the geometry of the lenses, but also in the materials used in the optical scheme. In fact, it is simply a different lens.

Jupiter 3 Calculation – Old and New. Original by Ludwig Bertele

Strictly speaking, the fact that there were several versions of the Jupiter-3 calculations was not a secret to me a year ago. The surprise was that among the serial Jupiter-3 lenses, as it turned out, you can find truly very different optical versions, and not as it was for the lenses Industar-61.

The optical design of the most common and well-known version of the Jupiter-3 lens, including the New Jupiter-3+, as well as the scattering spot diagrams, aberration graphs and MTF curves are shown below.

This version of the lens can be recognized by the flat surface of the fifth lens, as well as by its characteristic signature in the X-ray fluorescence spectrum - glasses of the special flint type are noticeably different in composition from ordinary flints and crown flints. Among the disadvantages of this version of the lens, one can note a strong spherochromatism for apertures higher than F/2 and noticeable astigmatism for the edge and corners of the 36x24 frame, which is most likely the reason for the mediocre image quality across the field demonstrated earlier in the test. The calculated shape of the bokeh spots agrees well with the observed one: the disks have a clear border in the part facing the center of the image, which intensifies on the sides.

The calculation of the Jupiter-3 objective is also known, given in the library previously used by ITMO University for educational purposes. This version of the objective differs from the previous one precisely in the lens materials: the third lens uses LK8 glass, which is not included in the current catalog of LZOS materials, and the fifth and sixth lenses are made of F1 and BF16 glasses, respectively, and not OF1 and TK21, as in the large-scale version of the Jupiter-3 objective. The presence of a bright lead signal in the X-ray fluorescence spectrum of the fifth lens of the 3 Jupiter-1952 objective indicates that it is made of glass such as ordinary lead flint, to which F1 glass belongs. Moreover, the surface of the fifth lens in this objective is not flat, but slightly convex, which agrees with the data given in the calculation. Thus, the 3 Jupiter-1952 lens is indeed a different lens from the large-scale Jupiter-3 lens.

The peculiarities of this calculation are a different balance of field curvature (more) and astigmatism (less), as well as higher field aberrations. In this sense, this version of Jupiter-3 resembles Industar-26m. Although spherochromatic aberrations are better corrected in this lens, the correction of lateral chromatic aberration and distortion suffered. However, due to poor transmission in the blue region of the spectrum, the lens showed itself very well in the sharpness test, although according to the simulation results it should be somewhat inferior to the large-scale version. The appearance of the lens bokeh according to the simulation results agrees well with observations. In particular, the simulation shows the absence of a clear border on the side of the disk facing the center of the frame, which determines a less "scaly" appearance of the bokeh compared to the large-scale version.

Of course, after this I wanted to take a look at the original lens patented in 1938 by Ludwig Bertele (US2186621) – the pre-war Sonnar 50/1.5. It turned out that in terms of the choice of optical materials, this lens is very similar to the large-scale Jupiter-3 – only the material of the fifth lens is noticeably different – ​​after all, there was no direct analogue of OF1 glass in the Schott catalog. In terms of the balance of aberrations and bokeh, this lens should also be more similar to the regular Jupiter-3.

It turns out that the most common large-scale Jupiter-3 is closer to the original L. Bertele lens, given in the patent, than the lens of 1952. Apparently, at the KMZ in the early 1950s, an attempt was made to recalculate the Bertele lens in order to reduce astigmatism and improve the image quality in general - and these lenses were even produced in some unknown quantity, but later this version was for some reason abandoned in favor of the one that is familiar to us all - where even the radii of curvature from those given in the patent do not differ much.

All this means that, most likely, the ZK (Zonnar Krasnogorsky) lenses and the large-scale Jupiter-3 from 1956 will be optically practically indistinguishable, but at some point before 1955, completely different lenses were produced under the name Jupiter-3.

So, there were at least three Jupiter-3 lenses: 1) A lens assembled using German lenses – Sonnar Krasnogorsk or ZK, 2) A recalculated version of Sonnar 50/1.5 with F1 and BF16 in the rear gluing – the lens presented in this article, 3) A version of Sonnar 50/1.5 with minimal recalculation with OF1 in the rear gluing – the large-scale Jupiter-3.

User experience

I liked the 3 Jupiter-1952 lens version optically even more than the new New Jupiter-3+ – despite the pronounced color distortions, low contrast and poor performance in backlighting. The old lens provides a more even distribution of sharpness across the frame, as well as a smoother, but still very characteristic “Sonnar” bokeh, reminiscent of the bokeh of a military 50/1.2 Zonnara from NVD TVNE-4B. Like other versions of the Jupiter-3 lens, this one also catches beautiful rainbow glare in backlight. Meanwhile, when stopping down to F/2-F/2.8, this lens provides a quite acceptable level of quality both in the center and across the field.

Below are some examples of photos taken with Jupiter 3 in 1952 and a full-frame camera. Sony A7s.

Then – photographs taken using shift adapter.


Conclusions

If this lens had a better quality execution, it would probably be the most interesting version of the Jupiter-3. Such a Jupiter-3, which is no longer just a copy of the German original, but a rethinking taking into account the experience of domestic optics, which is clearly visible through the spot diagrams in Zemax. But, as often happened and happens, initiative is not punishable, but rather ignored, and therefore, in the end, it was decided to make the Jupiter-3 almost exactly in accordance with the calculation of the author Ludwig Bertele. However, perhaps there were some other reasons for returning to the original source after the "creative impulse", part of which was the Jupiter-3 of 1952.

Add a comment:

 

 

Comments: 11, on the topic: Review of Jupiter-3 1:1.5 F=5cm P (1952, KMZ). How many different Jupiter-3s were there?

  • Boris

    I have several early Jupiters-3 51, 52 years for Kyiv and Zorki. Externally with different enlightenment. So it can be assumed that during this period at KMZ the lens was being refined for mass production. I also have in my collection Jupiters 55-63 years, 67-71, late 89. Sonnar 1,5/50 m39 48 years requiring maintenance. I like them all, the 60s Zagorsk were very high quality.

    • Rodion

      The coating of Soviet optics is completely random, because they used the spin-coating method, which does not allow achieving a reproducible result. The tolerances are very wide for the thickness of the applied layer. In this lens from the review, the coating was clearly selected based on the need for at least some compensation for the yellowness of the glass.

    • Nicholas

      Boris, our guys tried to bring a vacuum chamber for multilayer deposition from Germany, but they smashed it during a hard landing. After that, we only got MS in the 80s, and the experiments were done with single-layer films. Apparently, they were trying to find the optimum transmission curve, and the characteristics of the films changed several times. Orthochrome, panchrome, superpanchrome, then came mass color photography... "Yellowing" lenses are amazing for B&W, but completely unsuitable for slides. Then we adopted the GDR standard for lens color (and it was tied to the color formula of the "new" tessar - 11-00-00). I disagree about the irreproducibility, the optics-mechanics handbook gives clear instructions on this matter. And, in general, all Zagorsk lenses from the late 50s to the 70s had the same color.

      • Rodion

        The reproducibility of the chemical coating application was not good, as evidenced by the many lenses and eyepieces with completely different coating colors on different surfaces. Most likely, the color was adjusted by selecting lenses by the color of the glare, as, for example, in the I-26m or the same ZOMZ Yu3, but that is not a fact. In general, the experiments with coatings were not particularly successful, and, perhaps, a single-layer coating with pink-violet glare was more or less decent. Blue is terrible, yellow works poorly.

  • Yuriy

    Thank you for the most interesting review! It seems to me that only you and Arkady make such detailed reviews of lenses. And as for spectral analysis for mere mortals - here you are definitely the only one on the entire Internet! Good luck, spring mood... and new interesting discoveries!

  • Nicholas

    The analysis is very interesting. I'll put in my two cents. In the 80s, "Jupiters", "ZKs" and "Zonnars" were not uncommon in second-hand stores. Lens units were also sold separately. Therefore, we had the opportunity to test several dozen different samples. Then I returned to this topic. The conclusions were unambiguous. First. "Zonnar" and "ZK" give an almost identical picture, but "ZK" is usually somewhat worse, since it was most often assembled from different editions, which is reflected in the difference in the numbers of the front and rear lens units. Second. In terms of resolution, both of them at full aperture are inferior to "Jupiters" in the corners of the field in terms of sharpness. Third. The "beauty of the picture" of all these lenses is due to the fact that the coma in the corners ("birds") is compensated by geometric vignetting (or vice versa - whoever likes it better), as a result, the bokeh glare in the corners is as close as possible to those in the center. Fourth. None of the variants correspond to the first calculation by Bertele, in which the spherical aberration did not yet have an inflection. In all variants, the spherical part is corrected in the traditional way, as a result of which there is a focus shift between the values ​​of the relative apertures of 1,5 and 2,8. Fifth. The “Jupiters” have slightly stronger astigmatism in the corners, while the Sonnars and ZKs have coma and, it seems, curvature, so the picture is slightly different. Sixth. There were several design options: an exact copy of the Sonnar (“ears” with a small notch - ZK, “Ears” with a larger notch - “Jupiter-3 with ears”, knurling). Seventh. Sonnars already had a multi-layer (at least two-layer) coating, clearly visible on scratches and abrasions of the lenses, which determines their higher contrast compared to the Jupiters. At the same time, the quality of post-war Jena and Oberkochen Sonnars was the same. Well, and an addition. There can be no “iron” (rust, or what?!) in the lenses, the color is usually associated with the coating or with the quality of Canadian balsam and its substitutes. We were unable to reproduce the Zeiss coating, so we constantly experimented with single-layer coatings. The ugliest was the bright turquoise coating from KMZ, which gave a yellow-brown picture and bright blue highlights. The best in terms of contrast were apparently the Zagorsk “Jupiters” with “amethyst” coating, they also had very good blackening. The Valdai black “Jupiters” have good resolution and good color rendition, but poor blackening of the lens ends. According to Yarinovskaya, a good “sonnar” or “jupiter” with a working aperture of 5,6 has a resolution on Soviet KN-2 film of about 70 lines per mm.

    • Rodion

      It is wrong to talk about the aberrations of such a rather specific lens in terms of "comu", "astigmatism". The nature of the image at an open aperture is almost entirely determined by higher-order aberrations, which do not have names (or have, but not all).
      Bertele's original calculation was made, I believe, in 1934, and the article cites his 1938 patent. Post-war Opton Sonnars were recalculated using lanthanum glasses (Bertele's patents from the 1950s), while the triple bonds were retained.
      Iron in lenses can and does exist because it is a natural impurity to many oxides used in glass production. The higher the refractive index of glass, the stronger the iron impurity affects absorption due to the shift of the charge transfer band to the visible region. A striking example is the current glass LZOS TF10 and its Chinese analogue CDGM ZF7. Both are lead, but our yellow, Chinese is pure as a tear. With a yellow glare of enlightenment, there can be no yellowness of the lens when viewed through. Lens glue is also not the cause, because a single lens is also yellowish.

  • Sergei

    Dear author, did not mention Jupiter-3 produced by the Valdai plant “Jupiter”

    • Rodion

      In the context of what?

      • Sergei

        In the context of the title (heading) of this article

        • Rodion

          So, in terms of calculation, it is no different from ZOMZ and later KMZ.

Add a comment

Copyright © Radojuva.com. Blog author - Photographer Arkady Shapoval. 2009-2025

English-version of this article https://radojuva.com/en/2025/04/jupiter-3_kmz_1952_test_review/

Versión en español de este artículo https://radojuva.com/es/2025/04/jupiter-3_kmz_1952_test_review/