Leica Apo-Summicron-M 1:2/75mm ASPH
When Leica indicated that they were about to develop a new lens with a focal length of 75mm, I was a bit surprised. The 75mm seems to hold an uneasy place between the 50mm and the 90mm. There is a certain myth around this focal length, based on the excellent qualities of the Elcan 2.4/75mm, designed by the late Dr. Mandler. The physical specifications of this lens have been closely copied by the Voigtlander 1:2.5/75mm. Many photographers dismiss the 75mm as the lens seems to be too close to the 50mm focal length. If you make a few steps forward to your subject with a 50mm lens you can emulate the magnification factor of the 75mm, so the argument goes. And for the 90mm it is the reverse argument: just step back a few steps and you have the magnification of the 75mm lens. This type of reasoning ignores the finer points of the laws of perspective, about which I have to say more in this review.
To answer the basic question why the 75mm came into existence, we have to return to the high speed wars in the sixties and seventies. In those days, every respectful manufacturer tried to outdo the competition with ever-faster lens designs. The aperture 0f 1.2became the norm and the number of lenses with the specifications 1.2/50, 1.2/55, 1.2/58, 1.2/85 grew rapidly. Leica could not stay behind and produced the 1.4/80 for the R-series and the 1.4/75 for the M-series. The choice for these parameters was simply dictated by size and weight considerations. A 1.2/90 for R would be too big and heavy and a 1.2/90mm for M would obscure the rangefinder window too much. And the aperture of 1.2 could not be designed on the basis of the Leica equation for image quality.
So the 1.4/75 was a solution that satisfied the competing demands for size, weight and optical performance.
The rangefinder had to be redesigned with an additional frame for the 75mm and this change was partly responsible for the flare issues that have plagued the rangefinder window after the introduction of the M4P, the first body that could accept the 75mm lens. In fact both were introduced at the same time.
The Summilux 1:1.4/75mm has been designed more than a quarter century ago as a derivation of the classical Double-Gauss design. It offers excellent image quality, even according to current state of the art references. Its wide-open performance is quite good, but lacks the punch of lenses with smaller apertures. The effect of the residual aberrations is too strong at this wide aperture and cannot be countered by the means then available. The wide open performance sits between the more painterly drawing of the Noctilux at 1/50mm and the scientific drawing of the Summicron at 2/50mm, but is improved upon the Summilux 1.4/50mm, due to the smaller cone of light it has to transmit to the film plane.
The Summilux-M 1.4/75mm is a true workhorse: you need the lens and then you use it or you do not need it and then you do not buy it. The lens has no glamour or myth attached to it and is not a collector's item. And when you need it, you use it indefinitely and that may be the reason why so few Summilux 75mm lenses are offered on the second hand market.
The new Apo-Summicron-M 1:2/75mm ASPH is a derivation of the recently introduced Summilux-M 1:1.4/50mm. It has the same design characteristics: floating element, aspherical surface, a copious use of exotic glass types, the variety and cost of which explain in part the price. The design uses seven lens elements as analysis showed that element eight did not contribute to the image quality and had no added value.
The philosophy behind this design can be read in my review of the Summilux-M 1.4/50mm ASPH.
In a recent talk about the design considerations behind this lens, the designer drew attention to the main problems when trying to achieve superior performance with high-speed lenses. We are familiar with the main types of image aberrations, like spherical aberration, coma, astigmatism etc. But once these are sufficiently corrected the designer faces serious trouble in the reduction of the Petzval sum and the secondary spectrum (by means of achromatization). These aberrations can only be reduced by the use of the combination of many different glass types. If we are able to reduce the primary aberrations of spherical, coma and astigmatism, we get a perfectly sharp image point, but on a curved surface. The surface is curved because the lens elements have a curved shape. But we need a plane surface at the film gate. In the past, the designers had to introduce controlled amounts of astigmatism to offset the Petzval curvature. Leica lenses, especially the M lenses, used this technique, partly based on the nature of the Leica photography for reportage and documentary purposes, where good definition from centre to corner was not the prime directive.
Most aberrations are corrected by the method of lens bending: changing the radius of curvature. But lens bending cannot influence the Petzval sum and the chromatic aberrations. Only a change in the power of the lens element and the spacing of the lens element can do the trick. Changing the power implies often a different type of glass (different glass index) and the matching of the glass types can hardly be optimized by use of the computer. That may be one of the reasons why in the range from 35 to 90mm, the designs are all based on the Double-Gauss design type.
Leitz did experiment with exotic glass types in the past and even added layers of different glass together in one element to come to terms with these problems. An example of this glass can be found in my book. A theoretical study to break out of the grip of the classical designs was conducted at Leica several years ago and the result was a lens of breathtaking performance (an eight element 50mm design with GRIN lenses), but too expensive to manufacture. The insights gained however, were instrumental in designing the new Summilux 1.4/50mm ASPH.
The Summilux is not so fully corrected as the Apo-Summicron-M 1:2/75mm ASPH, which is logical as the field of view and the aperture are both smaller, easing the design problem. But the liberal use of the high index glasses with anomalous dispersion, the use of the floating element (the spacing problem!) and the use of an aspherical element add up to the solution for the reduction of the Petzval sum and the chromatic errors. The aspherics in both the 1.4/50 and the new 2/75 are both the moulded version of the aspherical surfaces, not the grinded versions. One of the arguments to use the aperture of 2 is the size of the aspherical lens, which would be too large if the aperture were to be stretched to 1.4.
The floating element is of a special construction. To get a focusing range from 70cm, wider than the customary 1 meter, the throw of the rangefinder curve had to be widened by 5mm to accommodate the range from 70cm to 100cm, which is mechanically complicated. And the relative movement of both groups is not linear as in most designs.
The Summicron 75mm is a normal long lens construction and not a telephoto version. This supports the performance at close ranges. Still the image quality in the close range is not as good as what is delivered in the range from 1.5m tot infinity. For really critical work one needs to stop down to improve micro contrast and the overall contrast of the image.
At full aperture the lens delivers a very high contrast image with even performance from centre to the extreme corners. The resolution of 80 linepairs/mm from centre to edge is unsurpassed at this time of writing. Of greater importance is the edge contrast that is sharply delineated without the smallest amount of colour fringing. The lens is not perfect wide open as we can see a small band of lower contrast (a dip) in the zone from image height 9 to 15mm. Here the contrast of the horizontal line patterns suffers a bit, but the lines can be seen quite clearly, but with some fuzziness. Coma, often a problem in high-speed lenses is fully corrected, as is astigmatism.
Stopping down to f/2.8 improves contrast and now the dip is gone completely. Resolving power reaches a value of more than 100 to 125 lp/mm over the whole image area. The image is of exceptional clarity and punch and the definition of extremely fine detail is outstandingly good. Stopping down further increases depth of field, and some crispening of the finest recordable detail, but this will be hardly visible outside the lab situation.
In practical terms we may say that the lens can be used wide open with the utmost confidence and the images can be blown up to whatever size one wishes without fear for blurring the details. Pictures made at f/2 at medium distances around 4 to 6 meters deliver excellent imagery and when studying the level of recorded details, one cannot believe that these are made at an aperture f/2.
The colour rendering is quite neutral and on the rich side as far as saturation is concerned. The particular strong point of this lens is its stunning propensity to show the depth of the subject with a three dimensional quality that is seldom encountered.
Veiling glare is totally absent and the lens produces deep black shadows with clean separation of subtle shadow detail. The internal blackening of the mount and the black paint on the rims of the lens elements effectively reduces secondary reflections and halos around specular highlight spots are not detected. On the other hand, we cannot shoot deliberately against the sun without causing some reflections and the lens is not fee of secondary images under all circumstances. It would be quite unrealistic to assume that a lens is totally immune for flare. The Apo-Summicron 75 is quite good in this discipline.
This lens is a primary choice whenever the realistic rendition of solid objects is required. The definition of the unsharpness blur is quite smooth and lacks the harshess of some very high speed lenses.
I made comparison pictures with the 2/50, 2/75 and 2/90mm Summicron lenses and the same distances and at the same magnification and of course wide open and stopped down. The conclusion of this exercise will not please the bo-ke fans. I did not see any visibly significant differences in the reproduction of the background blur. The famous words that beauty is in the eye of the beholder may be transposed to the bo-ke discussion. The subjective qualities of the background are often a major element of the composition, but one should not focus too much on this aspect. The quality of the sharpness area is still the defining property of a picture.
The Summicron 75mm operates with the solid smoothness that is the defining property of the current generation of Leica lenses. The lens is heavy for its size, but the weight of the special glass types asks its toll.
High contrast lenses
There is some confusion about the meaning of high contrast in the context of lens testing. One reads very often about the possibility that a high contrast lens should be matched to a low contrast subject and a low contrast lens should be a good choice for a high contrast subject. The match of a high contrast negative to a low contrast printing paper is well known and presumably one assumes that the same kind of match works for the contrast properties of a lens. A high (low) contrast subject or scene is one where the contrast range from deep shadows to specular highlights is wide (narrow). Or in other words the tone reproduction curve is steep or extended. What works for the tone reproduction cycle in the subject-negative-positive process cannot be migrated to the properties of a lens.
A high contrast lens does reproduce the spatial frequencies with good edge contrast. The low spatial frequencies (around 10 to 20 lp/mm) define the main outlines of the subject. But we need the high spatial frequencies to get good edge sharpness. See the figure below, which shows a bar line with steep edges. To reproduce these edges we need the high frequencies! In the diagram the rounded smoothed curve represents the reproduction of the bar line when the high spatial frequencies have been filtered. The resulting bar line has soft edges and will be seen in the picture as an outline with a fuzzy edge.
A high contrast lens will reproduce the details of the subject with good edge contrast, whereas the low contrast lens will reproduce the details with fuzzy edges and the cut-off level where fine detail is totally blurred is reached quick quickly.
Low contrast lenses are not able to reproduce the high spatial frequencies faithfully and quite often have a high amount of residual aberrations, which will be seen in the picture as an elevated level of flare. This may be the cause for the confusion: the flare in the low contrast lens will redistribute some of the light of the highlight parts of the picture to the shadow areas, which will have more density at the expense of the highlights. It seems that the tonal range is extended in the shadows, but that is not the case. The tonal range is identical, and there is not more shadow detail. The shadows seem to be easier to print.
Comparison with other Leica lenses
The obvious candidates are the Summicron-M 2/50 and the Apo-Summicron-M 2/90 ASPH.
The MTF diagrams (50mm) and (90mm) are indicative of the state of the art. The Summicron 50mm is weaker in the corners and the critical frequency of 40 lp/mm is for a large part below a contrast transfer of 40%. In practice this translates in a slightly less punchy imagery. The Summicron 50 at full aperture draws with somewhat wider brushstrokes than the Summicron 75mm. One may reflect on the additional effort that has to be put into the 75 to get a visibly enhanced imagery compared to the relatively modest outlay of the 50mm design.
The Summicron 90mm is in the same class as the 75mm, but note the more pronounced curvature of field of the 90mm. At the most critical inspection of pictures made at aperture f/2 with the 75mm and the 90mm, one can note that the 75mm has a tighter grain structure, an indication that the residual aberrations are reduced to an even lower level.
The four lenses in the focal range from 50mm to 135mm represent the Olympic platform for the Leica M range: the Summilux-M 1:1.4/50mm ASPH, the Apo-Summicron-M 1:2/75mm ASPH, the Apo-Summicron-M 1:2/90mm ASPH and the Apo-Telyt 1:3.4/135mm are able to extract every possible image detail from today's film emulsions and one feels entitled to question the necessity to improve these lenses further as long as one is working with film. The Leica M camera system is often designated as predestined for wide-angle photography and while undoubtedly true, the qualities of the standard to medium focal length lenses are such that the joy of photography on the performance edge of the medium has to be experienced to become a true Leica master.
The Color-Heliar 2.5/75mm
This Voigtlander design is the obvious candidate for comparison with the new Leica 75mm lens. Overall the C-H has a very good definition of detail in the centre part of the image, but with quite soft rendition of detail in the outer zonal areas of the negative. There is also a visible colour fringing at the edges of the main subject outlines. Stopping down improves the contrast and the performance becomes more evenly distributed over the whole image area. In numbers: wide open the resolution in the centre is excellent with 125 lp/mm, but this drops to 30 lp/mm at the edges with quite low contrast and fuzzy edges. At f/5.6 the edges are improved to 50 lp/mm with some fuzziness and a faint amount of colour fringing.
The C-H is a very good design with an excellent price-performance relation and one can wonder how it is possible that Cosina can deliver such a performance for the modest price asked for the lens. The image quality of the lens will satisfy many users, it in self a sign that the relentless drive for the ultimate performance has a natural end.
Status of optical design
The Summilux-M 1:1.4/50mm ASPH and its sibling, the Apo-Summicron-M 1:2/75mm ASPH do represent the current state of the art in optical design. The combination of aspherical surfaces, floating elements and in particular the break through in glass manufacture with the new glass types may be the turning point in the history of optical design for film based photography. Looking at the MTF graphs one may see still room for improvement, compared the best designs in the R range (the long focal lenses), but for the M range one is inclined to see the current level as a platform.
To extract more performance out of the design would imply more complicated designs at a cost that no one will feel prepared to pay. On the other hand the technique required to use this high performance is so demanding that it makes the concept of a dynamic style of M photography obsolete. Mechanically the new lenses are at the bleeding edge of manufacture and quality assurance. These designs could never be made in large quantities and one may be very happy that Leica is not sitting on its laurels and wishes to exploit its niche properties to deliver outstanding quality for the discerning Leica user and aficionado.
The position of the 75mm in the M lens line.
I already own the Summilux 1.4/75mm and I am familiar with the fact that the perspective and field of view of the 75mm are often advantageous, compared to the 50mm and 90mm focal lengths. The 50mm angle of field is in many cases just too wide and you have the inclination to reduce the distance between the camera and the subject, but then some subtle distortion will mar the picture, especially when photographing persons. With the 90mm you need more distance than is required for fine focussing or a psychological rapport to your subject. When you change quickly between the 75mm and 90mm, you will be surprised how fast and accurate the focusing becomes with the 75mm, where the 90mm is slower due to its bigger throw and the reduced (visual) accuracy of the finder.
The 75-135 pair could replace, or ideally complement, the classical combo of 50mm and 90mm with a wide latitude in picture opportunities.
The magnificent four (to borrow a title from that classical Western movie, the magnificent seven, deliver the best imagery money can buy in the M line. While there is a certain overlap in possibilities, the fingerprint of these four is sufficiently different to justify the existence of them all.
The Apo-Summicron-M 1:2/75mm ASPH offers imagery of a special kind with the combination of exceptional clarity of fine detail, crisp definition of subject outlines, and almost tactile plasticity of solid subjects. The just smaller angle of view than the 50mm allow for a tighter cropping of the main subject and one is inclined to spend more attention to composition, where the 50mm invites you to take pictures with a glance.