Three generations of optical design: Berek, Mandler, Kolsch
Intro: the person behind the glass.
In this article I will digress from my usual analytical approach to lens evaluation and adopt a more philosophical perspective. Lens design is as much an art as it is a science and while an MTF graph gives you an insightful understanding of the level of aberration correction, it does only indicate how the lens will represent the object, the user is trying to capture on film.
Since Berek’s days any lens is characterized by (to list a few items) its contrast/resolution, aperture, field of view, color correction, weight, size, tolerances and cost. All of these elements are interrelated and a designer has to strike (find) a balance between these requirements.
Any optical design is concerned with balancing aberrations, and not removing them as the nature of spherical surfaces (of which most lenses are composed, wholly or partially (when using some aspherical surfaces)) and the laws of nature dictate that all images will have defects.
Mr. Berek knew this, as did Mr. Mandler and so does Mr Kölsch and does Mr Schröder (current head of the optical department and an optical wizard second to none).
All of them have different ways of specifying performance requirements as artistic judgement and scientific knowledge are on equal footing in a truly grand design.
If we are allowed for our purpose to divide the Leica lens evolution into three periods, we would propose the following ‘history’. The Berek era: the goal was to provide Leica users with a large array of lenses that exploited the characteristics of the new 35mm format and allowed picture taking in hitherto uncharted areas of human life. Berek had to confront and come to terms with limited manufacturing capabilities as far as mechanical tolerances are concerned, with a restricted choice of optical glass, the lack of anti-reflecting coatings, the need for high speed lenses and of course limited capabilities for optical calculations (actual methods for ray tracing and theoretical knowledge).
Berek was well aware of these restrictions and followed a practical course. He knew the demands of the Leica users for image quality, and he knew the limitations of the photographic material. We do not often reflect on the circumstances of the Leica photographer in the early decades after the introduction of the Leica. Getting decent results with a pocketable metal camera with a collapsible 3.5/50mm with an extended depth of field that needed to be focussed accurately and very thick slowspeed emulsions, that had to be enlarged at least 3 or 4 times, was for any user a challenge. The excitement to exploit these new possibilities of the format and technique had to be satisfied with results that were convincing. In his handbook from 1930 (Grundlagen der praktischen Optik”, Verlag Walter de Gruyter u. Co.,) he overwhelms the reader with formulae and strange old German signs, but behind this facade, you will find very, very interesting passages. When discussing the correction of aberrations, he notes that the Seidel aberrations are the most visible in an image, and should be corrected to a large degree. (von Seidel studied and classified the monochromatic aberrations: spherical aberration, coma, astigmatism, distortion and curvature of field and the chromatic aberrations).The higher order aberrations will however degrade the image whatever you do with the Seidel aberrations. The wider the angle of view or the aperture, the more these higher orders will have impact on the image. He then proposes, and this is the core of the matter, to leave a small amount of Seidel aberrations uncorrected for all these lenses where the Seidel approach is appropriate. It is better to balance the aberrations over the whole image field and aperture and to accept that there is a certain amount of uncorrected aberrations, which are so well balanced as a group, that the resulting image is most pleasing. In addition he notes that the location of the stop and the use of vignetting are powerful instruments to influence the residual aberrations in the desired way.
Berek knew the limitations of the calculations and adopted a two pronged approach. For the general design of a lens he used theory and his creativity and vast experience. With these he could pinpoint the weak spots in the design and select a limited number of rays, that he would calculate with great accuracy. These results gave him the feedback to see how close he was to the desired goal.
He had to accomodate his designs to the limited number of glasses available, and the lack of high refractive glass was quite restricting. If we study the image characteristics of the lenses that have been designed by Berek (most lenses from 1925 to 1948), we recognize this approach. Focal length and aperture are always within the limits of practical design ( a 4.5/135 and a 6.3/28) and when the aperture becomes larger (2/50 or 1.5/85) Berek will accept a slight loss of ‘pure’ performance at the wider openings to allow the overall design to be balanced and deliver harmonious imagery. Berek’s designs and approach were acknowledged in his days, but his accomplishments were not properly valued. In the many handbooks of those period, he is often mentioned, but Merte, Bertele, Tronnier, Lee and others were the opinion leaders.
We have to remember that there were no methods or equipment for lens testing: one had to take actual photographs or one hadto calculate the aberrations from the lens descriptions and base a conclusion on the resulting graphs.
The smooth and pleasant shapes and tones that can be recorded with the Berek lenses are part of the Leica heritage. The Bauhaus generation was focussed on the “New Vision” and the grand masters of the snapshot were concerned with the surrealistic juxtaposition of image elements. This appraoch evolved into the universal pursuit of the candid photography. In both types the balanced flow of the light added to the image. Maybe the fact that Berek was a musician influenced his view on handling aberrations.
When stopping down Berek’s designs performed remarkably well. At the smaller apertures only the central portion of the lens is being used and the small aperture only allows a narrow bundle of lightrays. In this area only the primary Seidel aberrations are effectively degrading the image quality. But Berek had a tight grip on these errors and had them very well balanced in the overall design. The resulting performance is very commendable. The wellknown atmosphere in the images by Eisenstaedt, Doisneau and many others, this almost velvety depiction of the human condition, the sympathetic and even emphatic rapport between the camera and the subject, is most certainly and decisively influenced by the characteristics of the Berek designs.
The next stage is the period from about 1960 to 1980: this is the Mandler era. Leitz allowed themselves the luxury of two centers of competence for optical design (Wetzlar and Midland). It is interesting to note that the Wetzlar people studied apochromatics and the aspherics and the Midland designers concentrated on retrofocus designs. Still it can not be denied that there was a tense of competition between both groups. After 1960 new glasses with high refractive indices became available. The classical glass catalogue was filled with what we now call the “old glass types”, essentially developed by Ernst Abbe. As these glasses had low refractive indices, one had to bend the lens elements quite strongly to get the type of correction needed. But such curved glass is sensitive to manufacturing tolerances and one could not accomplish all. So the search for new glasses started with the additions of rare earth elements. Contrary to the wellknown assumptions, it was Kodak that invented the rare earth glass with lanthanum (La glasses) in the 1930s. Many of these chemicals were radio-active and could not be used. So the search went on to melt glass with the required properties and without radio-active elements. Leitz glass lab succeeded and with the famous LaK9 a new and important instrument became available to the designer. This type of glass has a high refractive index and a low dispersion. In those days this glass was really needed as the theory of aberration correction was not so advanced. Later one was able to show that the same image quality could be produced with a clever use of the old glasses. Most importantly the computer started to assist the designer in his tedious calculations. Now actual rays could be calculated at larger numbers, and so could remove the partial guesswork of the previous generation. To the importance of the glass selection was now added the importance of the design calculation.
The secret revolution however, was the concept of the merit function, where the designer would define a numerical value, composed of a number of aberration functions and related values, that represented the desired image quality and could instruct the computer to find a lens design that will suit this function. We read in Mandler’s book (On the design of basic Double-Gauss lenses) about his approach. The creation of a lens design has been superceded by the optimization of a design, as the classical basic design forms (as the triplet, double-gauss, retrofocus and symmetrical wide angle) were thoroughly investigated by the great designers of the previous generation. The fine tuning of these basic designs was the task of the new generation with the help of the computer. But the fundamental problem of balancing the aberrations was unchanged. The computer however allowed the designer to optimize the effect of the residual aberrations that would be allowed to operate in specific areas of the image. As an example we could accept a specific zone of spherical aberration to offset the curvature of field and check the effect by a change of the best location of the focal plane, but would introduce a controlled amount of focus shift when stopping down. If we now can distribute the residual astigmatism to the outer zones of the lens, where the undercorrected higher order aberrations can be checked by this astigmatism, the overall result is a lens that represents the state of the art and can be designed in a short period and can be cost efectively manufactured. The crucial part is here the remark about the “state of the art”. It was Mandlers goal to design lenses that were as good as could be possibly calculated in a fixed amount of time, when the merit function was appraoched.
Lens coating will not improve the basic characteristics of a design. So prewar designs that were coated after the war, will exhibit essentially the same optical characteristics. But with this technique flare could be reduced and most importantly the number of free elements in a design could be increased without loss of contrast introduced by the many single lens surfaces. For the classical Elmar and Summicron designs the use of coating had only a limited advantage. The reduction of loss of transmission is of course a very practical advantage. In the past an f/2 design transmitted the light only partially, making it effectively a 2.5 lens. With coating a true f/2 lens could be constructed.
High speed reportage lenses were in high demand by photojournalists and the Summicron trio (35, 50 and 90) by Mandler fullfilled the requests admirably. Higher overall contrast at the wider apertures was imperative and the character of the images changed quite a bit.
Scenes of war and other disasters were captured with an unflinching reality to bring home the brutality of the events.
The performance of the lenses changed to a generally higher overall contrast, As with Berek’s designs, the Mandler inspired lenses were a compromise between the corrections needed to get good optical quality and suit the demands of the users.
The Leitz lenses from the period 1960 to 1980 were very solidly build and may represent the best built lenses ever. But behind the solid mechanics, one could note a manufacturing technology that used the assembly stage of the lens to iron out some production tolerances. The bandwidth of the mechanical tolerances in this era was such that the optical quality was limited by the mechanical production tolerances and the glass types. The key word for the Mandler designs is “state-of-the -art”. The designs are first rate, even today, but the japanese competition came quite close to producing the same level of imagery. The Leitz designs had their own typical set of compromises that distinguished their lenses from the rest, but the best of Zeiss, Canon and Nikon and Topcon were measurably on the same level.
Typical for the Leitz designs of those days is a weak zone in the image at about two thirds of the negative area, which showed the compromise character. Curvature of field was not at the center of the design priorities and to offset this aberration, one had to accept this characteristic.
Why CoF was neglected is part of the Leica myth. As a fact we may note that Leitz designs were optimized to finetune the inherent traits of the double-gauss principle.
Multicoating was only used sparingly, partly because of its compex manufacture and partly because the designs did not benefit signifcantly when using this type of coating.
Generally it may be noted that multicoating when used as a design principle will enhance the quality of the result.
The Mandler era delivered first rate lenses, but they were mainstream lenses that fitted into the Zeitgeist of the 35mm style of photography. The innovations (zoomlenses) however were snatched away from Leitz by the Japanese companies.
The third period of Leica lenses may be characterized by Mr Kölsch, who masterminded the transition to new design concepts and new production technology. From 1980 to 1990, Leitz effectively had halted the design of lenses with a few remarkable exceptions like the Apo-Macro-Elmarit-R 2.8/100.
The Solms period can be characterized by the switch to new designs, new glasses, new design principles and new production methods.
It is a tribute to Mr Kölsch that he could break the traditional mold into which the Leitz designs had had to be designed. The most conspicious element is the adoption of moulded or grounded aspherical elements. The most important aspect is the narrowing of the mechanical tolerances to a fraction of what was the norm in the Leitz days.
The progress can be indicated when we look at the Apo-Summicron 2/90 asph. At full aperture this lens delivers better image quality than the Summarex 1.5/85 at an aperture of f11 and as good as if not better than the Summicron 2/90 at f/4.
By now the Leica designers attack every aberration and have the knowledge and means (glasses, machinery and precision) to no longer have to adopt a compromise. Naturally the ideal lens does not exist, but the balancing of aberrations is done at a much higher level and with much better precision than in the past.
If the Berek designs allowed the Leica photographer to capture images in light conditions where no-one had dared to walk and if the Mandler designs helped to define the role of the photojournalist as someone who could and would capture the minutiae of the human condition, where then would we locate the Solms designs?
If we study the classical pictures made with the Daguerre method we marvel at the exquisite detail that is captured in all its full glory and we even imagine the transcending of reality by the most accurate reproduction of events as seen by the camera eye. This is the same spirit we enjoy when we use Solms designs. We take pictures because of a psychological drive to transcend one’s own existence when freezing and fixing a moment in time that we do not want to forget or an emotion we want to cherish.
The early Leica photographers were at awe about what could be possible, the later generation documented its enviroment and its reaction to events in a straightforward, honest way and the current M6 user enjoys the thrill of capturing lights and shadow s with the technique of chiaroscuro: contrast, shape, line, form and texture are the medium, the accuracy and the information capacity of current Leica lenses are the tool .
One has to learn the tricks of the trade (about which I will talk later, if Bill allows me -:)), but the current state of the optical art, as defined by Solms, can redefine the joy of taking pictures as it did when Berek created his designs.
Dr. Walter Mandler died on april 21, 2005
Dr. Mandler died on April 21, 2005. Walter Mandler dominated the optical development within Leitz from about 1950 till about 1985. He was 'wissenschaftlicher Mitarbeiter' (science assistant) in the department of Max Berek and had intimate knowledge of the challenges and problems associated with the lenses for the small Leica format. When Leitz decided to set up a new company in Canada, he was asked to structure and manage the optical department there. Already in the fifties, the Leitz designers recognized the fundamental problems of small format and high-speed lenses and on both sides of the Atlantic solutions were created. In Wetzlar it was professor Marx who explored the first attempts of an aspherical design and in Midland it was Mandler who sought the service of the computer to speed up the design process. The period from 1950 to 1970 was one of the most exciting periods for optical designers as new approaches and insights could be explored without cost considerations, because of intense competition.
Optical design is partly pure mathematics and partly pure art. But the most important aspect is manufacturing quality and economy. One can design the most brilliant lenses, but when these creations are too expensive to produce, the design is ripe for the dustbin. In his 1979 dissertation Mandler explored in detail the limits of the Gauss design, carefully balancing optical quality and manufacturing cost. This book is still required reading for anyone who wants to understand the manifold aspects of computer assisted optical design.
Walter Mandler had a strong commitment to high-speed lenses that could be manufactured at reasonable costs. The Summilux 1.4/75 and 1.4/80 are examples as is the famous Noctilux 1/50mm. This lens is a masterpiece of optical design. It is very difficult, if not impossible to improve on this design without the use of exotic glass types and aspherical surfaces. He used the computer to explore all possible options and select the design that suited his requirements best. He was also a daring man: he created the Summilux 1.4/35mm when everybody assumed that such a lens was not possible. Mandler was a pragmatist more than a visionary. His designs are very competent, but he had to work within the restrictions of the Leitz philosophy of lens manufacture. It seems that lenses for 35mm photography were a bit boring for him, as they did not pose the challenge he wanted. This he found in the designs for the military and in the Elcan lenses he could explore more exciting optical limits.
It is a strange coincidence that the famous designers are of German origin, like Bertele and Rudolph and Berek, but all serious optical design programs are American. By location Walter Mandler was more exposed to these programs and their underlying design approach. The Germans on the other hand relied on a more fundamental and theoretical design method. The programs designed in-house by Leitz originated with Professor Marx, who was responsible for the first Noctilux with aspherical elements. Theoretically and from a performance perspective the Marx design was undoubtedly the best, but too expensive and the Mandler design used high index glass types to deliver (almost) the same performance at lower cost.
In the sixties and seventies the Japanese designers made big strides forward with novel techniques, like aspherical surfaces, floating elements and HD glass to push the lens limit to uncharted terrain. In many cases these designs were flawed as the Germans were quick to note, but the Japanese gained valuable knowledge and honed their designs to perfection. Leitz and to a lesser extent Zeiss, focussed on the other part of the lens equation. A good design is worthless, if not supported by adequate manufacturing techniques and here the German industry had definitely the advantage.
There are only a few letters by Mandler in the Leica archives and the contents give the impression of a passionate man, who feels slightly frustrated by the slow progress made by Leitz and the fast improvements made by the Japanese. He clearly saw that without fundamental changes Leica could not compete in the long run. But Leitz was in those days already living on borrowed money. His Elcan designs were by nature less cost oriented and the knowledge gained here could be transferred to the photographic department.
Walter Mandler must have been happy with the optical progress made in the last decade in the Solms design team. Many of his ideas have been developed further and current Leica designs employ the whole range of techniques now available: aspherics, floating elements, a wide selection of glasses, and a focus on the critical aberrations, like the Petzval sum and the correction of the secondary spectrum.
Mandler's study about the Double-Gauss designs is still the definitive analysis of the limits and potential of this class of lenses. The book was published in 1979 and represented the state of the art in optical design at that time. His achievement was the transfer of that theoretical framework to practical design. He did not develop really innovative designs, but his strong point was the exploration of existing limits and to find ways to implement the almost impossible. The Elcan 2.4/75mm and the Noctilux 1/50mm are examples of his craft and his passionate commitment to photographic lenses.
Dr Walter Mandler was 83 years old and lived from may 10, 1922 to april 21, 2005