Triplet-69-3 40/3.5 (with tests at Zemax)

The lens presented in the review was designed and manufactured by Igor Porokh (aka lens_made_in_ussr). The material was prepared by Rodion Eshmakov.

Type of adapted lens.

Type of adapted lens. increase.

Triplet-69-3 4/40 is a non-removable regular lens of cheap Soviet scale cameras (prototype of soap dishes) produced by BelOMO (Viliya and others). Among analogues, the lens stands out for its large rear segment, due to which it can be adapted for use with SLR cameras, and without a factory aperture, the lens is also compatible with full-frame DSLRs.

On Radozhiv there are already as many as 2 reviews of Triplet-69-3:

But this article presents not just an ordinary remake, but a real product in the form of a 69mm F/40 Triplet-3.5 lens, qualitatively adapted for cameras with an EF mount through the use of a well-thought-out design from parts printed on a 3D printer.

Specifications of the adapted lens:

Optical design - Cook's triplet (3 lenses in 3 groups);
Focal length - 40 mm;
Relative aperture - f / 3.5 (due to additional aperture opening);
Aperture - 4 petals (square hole);
Aperture limits - F / 3.5-F / 16;
The minimum focusing distance is 0.45 m;
Hard-stop precise setting to infinity - no;
Thread diameter for filters - 52 mm;
Camera mount - EF mount;
Weight - less than 100 g.

Design features

The lens is made in the original housing made of thermoplastic polymer (probably PLA or ABS). All external parts, except for the factory title ring, are 3D printed. Thanks to this, the lens turned out to be almost weightless! At the same time, it does not feel flimsy or unreliable at all - there are no backlashes in the design.

Side view of the lens when focusing at infinity.

Side view of the lens when focusing at infinity.

I didn't take the lens apart (because I didn't want to ruin anything), but it appears to use a factory helicoid built into the new body for focusing. Focusing is performed by rotating the entire front of the lens together with the lens block at an angle of up to 180 °. In this case, the lens elongation is 4 mm, which provides a minimum focusing distance of 45 cm. It seemed to me that this was a bit too much: I am very used to lenses that allow me to shoot without additional devices at a scale of 1:3–1:2. But, apparently, the helicoid used did not allow for a smaller MDF.

Side view of the lens when focusing on MDF. Side view of the lens when focusing on MDF.

Side view of the lens when focusing on MDF.

The lens is equipped with MDF and infinity position stops. But the infinity stop is set inaccurately, which is why there is a slight overshoot. However, this puzzles photographers and all-metal modern lenses.

View of the lens aperture closed to F / 16 from the side of the front lens.

View of the lens aperture closed to F / 16 from the side of the front lens.

The lens from the factory has an unusual four-bladed aperture diaphragm, which forms a square hole when closed. In this adaptation, the aperture was preserved, but because of this, the lens is incompatible with full-frame SLR cameras due to mirror engagement.

View of the lens from the rear lens. This small “step” protruding beyond the bayonet plane does not allow you to mount the lens on a full-frame DSLR.

View of the lens from the rear lens. This small “step” protruding beyond the bayonet plane does not allow you to mount the lens on a full-frame DSLR.

The factory-made aperture of the Triplet-69, even at full relative aperture, is slightly covered and partially covers the rear lens (a photo), in this adapted version, it is completely open, which gave an increase of 1/3 of the aperture level. That is, the lens instead of the factory F / 4 turned into ~ F / 3.5. A lens without a factory aperture is still about 1/3 stop brighter (~F/3).

View of the lens from the side of the rear lens with the aperture closed to F / 16.

View of the lens from the side of the rear lens with the aperture closed to F / 16.

Aperture control is carried out using a ribbed ring located at the mount. The ring moves with clicks, roughly corresponding to F / 3.5, F / 4, F / 5.6, F / 8, F / 11, F / 16. The lens does not have aperture scale markings.

The lens block of the lens has not undergone any visual changes. In general, it is hardly possible to carry out any additional manipulations with him.

View of the lens through.

View of the lens through.

The lenses of the Triplet-69-3 are coated with the usual single-layer anti-reflective coating with a blue tint glare. The lens is a little yellow.

View of the lens through the lens with a closed aperture.

View of the lens through the lens with a closed aperture.

Triplet-69-3 40/3.5 by Igor Porokh leaves a very pleasant tactile impression. All mechanical parts work without any complaints, without jamming and extraneous sounds. The lens is very comfortable to use. A little missing aperture scale markings. And, of course, the big disadvantage is the plastic mount (just kidding!!!).

Optical properties

Triplet-69-3 40 / 3.5 is a very, very weak lens in terms of image quality. It is quite soft wide open even in the center of the frame (spherical aberration), and its edges even on APS-C do not bounce back up to f / 5.6-f / 8. On a full-frame sensor, the edges of the frame will not be sharp at any aperture - resolution is limited by the strongest lateral chromatic aberration and astigmatism. Like this!

Image contrast is low, somewhat worse than other similar lenses with single-coated optics. In hard backlight, the lens gives artifacts such as “sunshine”, but practically does not produce glare. A four-bladed diaphragm on point light sources forms a halo in the form of an asymmetrical four-beam star.

The bokeh of the lens within APS-C looks interesting, but the lens itself is too "dark" and has too long a minimum focusing distance for this bokeh to ever be possible. Square bokeh is completely unattainable, because along with aperture, background blur at available focusing distances also disappears. It makes sense to use the lens in artistic macro photography, but this will require additional devices (macro rings). Outside the APS-C frame, bokeh changes its appearance due to the strong influence of astigmatism: first it becomes very swirling, and then smeared into a mess. This triplet does not give characteristic “bubbles” in bokeh due to too low aperture ratio (and a small amount of spherical aberrations).

Below are sample photos on the Triplet-69 40 / 3.5 and the Sony A7s full-frame mirrorless camera. Some of the pictures were taken with the help of a shift adapter as a "shiftorama" (an increase in the frame area by ~1.5 times).

Conclusions

Of course, Triplet-69 40 / 3.5 is not at all the same “pancake” as smooth and sharp Canon EF 40 / 2.8 STM. This "fantastic plastic" is optically closer to lomographic lenses, although, ironically, even the notorious LOMO Minitar-1 is both more complicated and more perfect than this Triplet.

For me, the concept of the Triplet-69 as a compact full-frame lens for shooting at medium / long distances was not entirely clear. I would like to have such a lens more in the form factor of a compact macro lens, so that you can benefit from the unusual design of the aperture and better display the characteristic pattern. Nevertheless, the high quality of the lens and its thoughtful design are very impressive.

Bonus: the impact of optical design parameters on the quality of triplet 40 / 2.8 lenses

Let me remind you that the "triplet" (or "Cook's triplet") scheme is truly unique - it is the simplest anastigmat in terms of its optical design, providing a field of view angle of ~45°. Three spherical lenses is the bare minimum to adequately balance all monochromatic and chromatic aberration. But in such a lens there are very few correction parameters:

  • Optical glass parameters - refractive index and its dependence on the wavelength of light (dispersion), maximum 3 brands of different glasses in the lens (6 parameters);
  • Radii of curvature of optical surfaces - 6 pieces;
  • Lens thicknesses and lens distances are 6 more parameters.

In total - 18 variable parameters, of which 6 change discretely and involuntarily, as they are tied to real optical materials.

As a result of these limitations, the most typical triplet for SLR cameras of 36×24 mm format it has a field of view angle of ~45° (focal length 50-55 mm) and relative aperture not higher than f/2.8. There have been extreme attempts to make a 50/2.2 triplet that have spawned something very very scary. What happens if we now want to calculate an f/2.8 triplet with a 56° field of view (40mm focal length) for a SLR camera?

It turns out that this task is much more difficult than calculating the usual "fifty kopecks". The fact is that one of the most important characteristics of a lens is the value of its back focal length (BFO) in relation to the focal length (FR). Optics for full-frame SLR cameras should have a distance of at least 38 mm, which, at a focal length of 50 mm, gives a value of the PFD/FR ratio of ~0.76. For a lens with a focal length of 40 mm, this ratio will already be 0.95! In other words, the lens must be retrofocused within five minutes (ZFO > FR), and to fulfill this requirement, they usually use special optical schemes by the type of "television in reverse".

The "stretching" of the value of the rear focal segment acts as a kind of limitation of the range of adequate values ​​of interlens distances and lens thicknesses, which in turn leads to the need to put up with compromises in image quality.
Since there are very few correction parameters in the triplet in general, the correct choice of glass grades makes a decisive contribution to the success of the calculation. The main principles are as follows:
⦁ The refractive index of positive lenses should be as high as possible (field curvature and spherical aberration correction);
⦁ The refractive index of the negative lens should be relatively low (to correct field curvature);
⦁ Dispersions of positive lenses must be small enough, and those of negative lenses must be large enough to correct chromatic aberration.

Abbe diagram for the catalog of optical glasses GOST (IPZ, LZOS).

Abbe diagram for the catalog of optical glasses GOST (IPZ, LZOS).

Following the advice from paragraph 1 forces the use of heavy lanthanum crowns (STK) in the calculation. But how big is their influence? Consider a triplet lens model with parameters 40/2.8, obtained by scaling the famous Soviet projection triplet 78/2.8 https://radojuva.com/2021/09/triplet-78-2-8-2/ (1946). This lens does not use lanthanum glass in its design and uses conventional heavy crowns (TK).

Screenshot of the report from the Zemax program for the Soviet lens type Triplet, scaled up to 40 mm.

Screenshot of the report from the Zemax program for the Soviet lens type Triplet, scaled up to 40 mm.

The figure in the upper left corner shows the spots into which the lens turns a point source of white light (400-650 nm) at different distances from the optical axis (angles of view 0°, 30°, 45° and 56°). The bottom left shows the shape of the spots in prefocus (left) and out of focus (right) from the center of the frame (top) to the corners of the frame (bottom). At the bottom right, the diagrams show the shape of the field, taking into account curvature and astigmatism (left graph) and the amount of distortion for different values ​​of the field of view angle (right graph). On the principle diagram of the lens, the numbers indicate the refractive index of glass (n) and the Abbe number (v) of glass (the more - the lower the dispersion).

Note that the Soviet triplet has a very short rear focal length. The PFD/PR ratio is only 0.8, while the required 0.95. At the same time, the lens is completely hopeless beyond 45 ° - and recalculation is able to reduce the strongest astigmatism along the edge of the frame, only adding it to the central regions of the field. In addition, a strong slope of the curves on the field curvature plot to the left indicates a pronounced residual field curvature. This is not Petzval, of course, but it is too much for a 40 / 2.8 class lens. This triplet also has a significant coma, as evidenced by the asymmetry of off-axis spots and the asymmetric border on the bokeh disks.

Now let's turn to the lens described in the patent US3176582 (1960), which was calculated by Ernst Tronnier. But not the one who, on a dispute, was the first to consider a 50 / 1.8 lens with a concave front lens (the famous Ultron 50 / 1.8) - he was Albrecht Wilhelm Tronnier.

Screenshot of a report from Zemax software for a triplet lens from Ernst Tronnier's patent US3176582, scaled to 40mm.

Screenshot of a report from Zemax software for a triplet lens from Ernst Tronnier's patent US3176582, scaled to 40mm.

The value of the refractive index in the positive lenses of this lens has increased by 0.06, which is quite a lot. The rear lens of the objective is made of lanthanum crown. A good coma correction catches the eye - the off-axis spots are symmetrical, the bokeh discs have a symmetrical border, which is generally more pronounced than that of the Soviet lens. This means that the Tronnier lens would give much better characteristic “bubbles”. The curvature of the field, which greatly affects the image quality on covered apertures, has also decreased. But there is no huge increase in quality, because in this lens the rear segment has also been increased to 34 mm. FFD/FR is already 0.85. Thus, only 2 extra millimeters of working distance practically "swallowed" the entire effect of using high-refractive glasses - and did not choke.

But what happens if you still try to reach the desired rear segment of 38 mm? To answer this question, I recalculated the Ernst Tronnier lens using even more modern glasses (H-LAF50B crowns), focusing on astigmatism and field curvature correction. In general, there is nothing left of the Tronnier lens here.

Screenshot of a report from the Zemax program for a Triplet lens, calculated by me.

Screenshot of a report from the Zemax program for a Triplet lens, calculated by me.

It is easy to see how much a lens with an increased distance, even just outwardly, differs from a regular triplet - and this is precisely a trend, and not a special case. This lens has a rather low level of astigmatism (important for improving performance on covered apertures), relatively strong coma and undulating field curvature - increasing the index by 0.07–0.1 was not enough to overcome these distortions either. The image quality in the central area at an open aperture is comparable to the Tronnier lens, but beyond 20-25 ° is much worse due to the curvature of the field - a compromise. The calculated lens is close to the Tronnier lens in terms of the nature of the bokeh in the central region of the frame, but it gives scales along the edge of the frame due to the presence of coma.

Thus, in order to calculate a high-quality wide-angle triplet for a reflex camera, you need to use the most highly refractive glasses - like H-ZLaF4LA (n=1.91, v=35.25), for example. Just to be able to correct chromatic distortion. And if you use aspherics ...

History, however, did not follow the path of rapid development of the technology for melting optical materials with extreme parameters, and aspherical triplets are mainly used in cheap plastic projection systems. It turned out to increase even more aperture and image quality in the central area of ​​the frame is sufficient split positive lenses to correct spherical aberration. To improve field sharpness - single positive lenses can be replace with glues, in which not chromatic aberration is corrected (as in doublets of telescopes or televisions), but coma and astigmatism. In the case of the 40/2.8 Triplet, even a slight complication of the optical scheme gives an increase in optical quality that is much greater than the coolest optical glass provides. For example, the closest relative of the Triplet - the four-lens Tessar - has a much better image quality than the Triplet, which is similar in terms of technology and size of the WFD (compare with the Tronnier triplet).

Screenshot of report from Zemax software for Tessar 80/2.8 lens (Hasselblad) scaled up to 40mm.

Screenshot of report from Zemax software for Tessar 80/2.8 lens (Hasselblad) scaled up to 40mm.

And if we apply the scheme of a five-lens double Gaussian (Biometar / Vega) to create a 40 / 2.8 56 ° lens, then a fairly high optical quality is achieved even without lanthanum glass.

Screenshot of the report from the Zemax program for the Vega-11U lens, scaled up to 40 mm.

Screenshot of the report from the Zemax program for the Vega-11U lens, scaled up to 40 mm.

Looking at these charts, you can't help but understand why Planar was named that way. Of course, the scaled Vega has a very short posterior segment, but it demonstrates the idea well. And by adding just one lens to the same Vega, you get a retrofocus lens Mir-1 37 / 2.8, which has a very good optical quality ...

The calculation of a triplet with characteristics that are not optimal for such a scheme is quite an entertaining task and is best suited for studying the effect of correction parameters on the resulting image and lens pattern, which can be useful for calculating more complex schemes. Fortunately, modern software packages for the calculation of optical systems are simple enough to be mastered by a person without specialized education and offer a wide range of options for calculating and studying various optical systems.

You will find more reviews from readers of Radozhiva here.

Add a comment: Sergei

 

 

Comments: 11, on the topic: Triplet-69-3 40/3.5 (with tests at Zemax)

  • Sergei

    The review of this Belarusian-made triplet turned out to be somewhat academic (with justification of the existing limitations of this optical scheme).
    But first of all, craftsmen-practitioners are interested in whether it makes sense to adapt the T-69 if there is a very similar T-43 from Smena LOMO (with a conventional aperture).
    And to the most likely use on crop (instead of the rare full frame).

    • Rodion

      Another “more interesting and correct”? The T69 is notable only for its greater aperture compared to the T43 (almost twice, if without a native aperture) and a larger segment. T43 cannot be pushed onto a full frame DSLR, but T69 is possible, albeit without a diaphragm. Although this is also debatable - if there is a desire, you can screw the diaphragm. It's only a question of necessity, but that's another story.
      In my opinion, T43 is better for crop. T69 is like that. It is either like making a makrik funny, or not at all. In the landscape, he is none; for portraits, too, there are more interesting options.

  • Victor

    I wonder if Igor can repeat such a design for a fee? T-69 is available…

    • Rodion

      He has a VK group with the address *lens_made_in_ussr - write to him, ask. I know for sure that he still made the lens for Pentax. And on the pentax, he does not catch the mirror even on ff.

  • i is glorious

    Such a question: how is the filter stack in front of the matrix taken into account in the software?

    • Rodion

      It can be taken into account if you know its thickness and the material from which it is made. Lenses for night vision devices are considered immediately based on the presence of glass in front of the photocathode, for example.

  • after all

    I had Willy. Shit. Seagull is way better

    • Rodion

      And no wonder. There is a tessar in a seagull, and without any compromises on the working length.

      • after all

        with the scheme it is clear, but how does the working length affect the quality?

        • Rodion

          I wrote about it in the "bonus". If specifically about the physical foundations, then the point is that the closer the lens is to the image, the better it is able to correct its quality. Mathematically, this is also justified as the fact that the lenses that are close to the image, the light falls at a small angle relative to the normal, so the aberrations show little.
          It turns out that if the rear segment is “stretched”, then the correction deteriorates sharply.

        • Rodion

          By the way, early lenses of the “planar” type had quality problems with an increase in the rear segment. Therefore, the FR of the “reflex” biotar was 58 mm, although rangefinder lenses such as summikron had a focal length of 50. Lenses that were “stretched out” despite the quality are Paknolar 50/2 https://radojuva.com/2019/06/carl-zeiss-jena-1q-pancolar-250-adaptirovannyj-dlya-nikon-obzor-chitatelya-radozhivy/ and Helios-65.

Add a comment

Copyright © Radojuva.com. Blog author - Photographer in Kiev Arkady Shapoval. 2009-2023

English version of this article https://radojuva.com/en/2023/03/triplet-69-3-40-3-5/?replytocom=585713

Versión en español de este artículo https://radojuva.com/es/2023/03/triplet-69-3-40-3-5/?replytocom=585713