Don't forget to update your personal camera inventory
CollectiBlend
Antique and classic cameras/lenses price guide.Back to the table of content
Chapter 2: The evolution of the LEICA Lens.
2.1 Part 1
The origins, 1925 to 1930. When
Barnack was busy constructing and refining his camera around 1907 to 1911 , and decided to use a format of 24x36mm on cine-film, no
suitable lens was available on the market. As any lens projects a circular
image, you need a angular coverage equal to the length of the diagonal of the
image format .
The diagonal of the 24x36mm format has a length of 43,27mm,
which gives a maximum image height of 21.6mm from the centre of the image.
Barnack therefore needed a lens with a angular view of
46o to cover the format. A lens with a focal length of anything between 40mm
and 60mm would do. Barnack, working at the Leitz factory, a well-known
microscope manufacturer, found a series of lenses, made by Leitz and named
Mikro-Summar and Milar with several focal lengths to choose from.
Figure 1: 2.1 A Mikro Summar
He chose the Mikro-Summar 1:4.5/42mm, which he fitted in at
least one of his three prototypes. The Ur-Leica, residing in the museum in
Solms, has been equipped with this lens. Some researchers have identified the
lens as a Milar or a Summar. There is no inscription on the lens, so there is
room for some interpretation. The exact focal length however can be measured
and recently the optical engineers at Solms just did this and established a
focal length of 42mm. The only lens in the Leitz microscope catalogue around
1910 is the Mikro-Summar, a six element symmetrical lens. So we may now reveal
with certainty the identity of the mysterious lens in the Ur-Leica. It is a
Mikro-Summar with 6 elements, symmetrically arranged.
2.1.1 Choice of format.
Why did Barnack settle on the 24x36 format? Previous and
contemporary designers of cameras, based on the 35mm perforated cine-film ('Kino-film')
had used 24x24, 32x44, 18x24, 30x42 formats. Barnack himself tells us that he
wanted to use the 'Kino-film' and that he decided to go for maximum area to
ensure good image quality. The dimensions of 24x36mm have a 1:1.5 relation,
just as the 6x9cm negative of the then ubiquitously available Roll film camera,
the 'point and shoot camera' of the early part of the previous century. The
Leica camera had to face a very strong competition from this format, as the
most used print format was also 6x9cm and this could be produced from a 6x9cm
negative as a direct copy, without any enlarging and subsequent losses in image
quality. And the next steps in print format (9x12cm and 13x18cm) have relations
1:1.35, so quite close to the 1:1,5 (with a slight
safety margin). It was one of Barnack's primary concerns that his compact
camera and its small negative could deliver image quality as good as, if not
better than the 6x9cm contact-print. His device would be unacceptable to most
prospective users if the image was inferior to the roll-film competitor. Here
we see already the interaction between the demands from the practical user and
the capabilities and ambitions of the optical designer that will characterize
the development of optical design.
Figure 2: 2.1.1A Barnack sketch.
Note similarity to Berek drawing The
next step, after having established the dimensions of the image area, was to
define the requirements for the recording capabilities of the lens.
2.1.2 Optical requirements
Most photographic prints in the early decades of the 20th
century had dimensions between 6x9cm and 13x18cm. A comfortable viewing
distance for these pictures is around 25cm. It so happens that we are able to
see the whole 9x12cm print at this distance without moving our eyes. The
resolution of the eye (that is the finest detail we can distinguish when
viewing an object) is of course the limiting factor for the optical
requirements. It has been established by research that the best resolution of the
eye will be found when the eye is focused at a distance of 1.2meter to 1.5meter.
But for photographic purposes we have to stick to the normal
viewing distance of 25cm. This distance has been used as the reference distance
for the definition of the circle of confusion. Assuming an enlarging factor of
the negative of 3 to 5 times for a print of 9x12cm (in fact 7.2x10.8cm) the
smallest point on the negative should be at least 0.1mm/3.
That is 1/30mm or 0.0333mm. This value is used by Leitz as the criterion for
the calculation of depth of field tables or engraving depth scales on the lens
mount. The optical calculations however are not based on this value. It is
clear that the designer will try to concentrate all light rays on as small an
area as possible.
The value of 0.03mm then constitutes the limiting case, the
minimum value so to speak, that is a point should have at least the diameter of
0.03mm or 30 micrometer (mu). When reflecting on these topics Berek and Barnack
settled for 0.03mm. They could not know at what distances the Leica pictures
would be viewed by the user, nor what enlargements they would choose. The value
of 0.03mm is the standard circle of confusion at the film, used as a reference.
It should not be considered the target value. See the chapter 1.2 for more
details on this topic.
2.1.3 The first lens for the 'Leica' format: some design aspects
The first lens, specifically designed for the Leica format
was the Leica Anastigmat , later renamed Elmax
(presumably the combination from Ernst Leitz and Max Berek). This lens had five
elements, where the Zeiss Tessar had four. Berek designed the Anastigmat with
the last group composed of three cemented elements. This lens then has the
unique distinction as being the first lens, designed with the specific requirements
for the 35mm Leica format in mind. Max Berek designed the lens around 1922 (As
the date on the original drawings at Solms tells us).
Figure 3: 2.1.3 A (Computation page: already at Hove)
Figure 4: 2.1.3 B Anastigmat
As the Elmax was a bit complicated to build (as a cemented
triple) Berek changed the design to a triplet with the last element as a
cemented doublet.
Figure 5: 2.1.3 C Elmax.
The Elmar was born. The design date on the
documents tell us that the first design of the Elmar was completed in
1925. This lens had improved imagery when compared to the Elmax. The Elmar
design is similar to a Tessar design, with one important exception, the
location of the stop. In optical design the location of the stop is part of the
tools for aberration correction. Depending on the design this location can be
more or less influential. In the case of the Elmar, the stop is located between
the first and second element. This forward location cuts off some of the rays
at the edge of the lens and while producing some additional vignetting, also improves
the central definition. As the lens is now unsymmetrical around the stop, we
see some astigmatism and coma in the outer zones of the field.
Figure 6: 2.1.3 D: Elmar diagram
The Elmar 1:3.5/50mm established the fame of the Leica
camera, by providing image quality of a high order for its day. Analysis by
modern computer based design programs of the original Elmar configuration show
that the basic design is difficult to improve upon. A tribute to the old
masters no doubt! A designer would cement two glass elements together ( or split one element into two separate lenses) to create a
glass type that is not available in the glass catalogues. Basically the Elmar
is a triplet, where the last lens element is cemented to generate a new type of
glass, that was needed for the computation.
If advances are to be made, the only route is the glass
selection. And we may be sure that the Elmar lens in its long history of 1925
to 1961 underwent changes of this kind. Inevitably older glass types were
removed from the catalogues and so the designers were forced to adapt to these
circumstances. But modern optimization analysis also shows that the gains are
relatively modest and so it may come as no surprise that the image quality of
this lens is quite stable over the years. Of course newer glasses and
specifically the coating of surfaces might give a modest gain under adverse
conditions. The design of the Elmar allowed for the correction of the basic aberrations
with minimal means. This makes the computation especially labourious and when
done with only a slide rule and logarithmic tables is very demanding on the insight
and creativity of the designer. Any four-element lens has limited possibilities
for the correction of the important aberrations. A compromise is always needed.
Broadly speaking the four-element triplet has a typical
character like good central sharpness and weaker performance in the field. A
designer can and will adopt a specific solution, that gives every version of
this type a different fingerprint. The Elmar at full aperture gives a low to
medium contrast image, with good recording of coarse detail over an image
circle of about 12mm radius. One of the advantages of the Leica, when compared
to larger format cameras is the extended depth of field when using the lens
wide open. Max Berek rightly placed emphasis on the axis (centre) performance
and also on a good colour correction over the whole spectrum, including the red
part of it. Stopped down to about 1:5.6 the contrast improves markedly,
crispening the rendition of fine detail. Berek again assumed that the lens would
be stopped down in order to increase depth of field to maximize the comparative
advantages with the roll film cameras.
Figure 7: 2.1.3 E-Late type Elmar
We may regard the Elmar as a classical example of an
excellent design that stretched to the limits the inherent possibilities of the
type. The three element Cooke triplet has comparable
performance on axis as the classical four-element Elmar type.
Computer analysis show that the
relative differences in performance are zonal astigmatism and oblique spherical
aberration, both of which are better corrected in the Elmar. The differences
are small, but quite noticeable and could make the success or failure in the
marketplace. So it was a wise choice from Berek and Leitz to opt for the
4-element Elmar to compete with the Cooke type triplets employed in most
roll-film cameras. The performance edge indeed is important and outweighs the higher
production costs, due to the more elaborate construction.
Part 2: The expansion, 1930 to 1957
Around 1925 most films were of the orthochromatic type and
had sensibilities of ISO20 to ISO30. Panchromatic films were also available,
but with reduced speeds (ISO10 to ISO15). Even with a 3.5 lens, these speeds
limited the possibilities of the Leica user to shoot pictures in low light
situations. But the Leica was the camera to record all aspects of human life in
its various surroundings. Today, when we look at early Leica pictures made in
'available light' we should marvel at the expertise of the average Leica user
to use this modest equipment. Ask any modern counterpart to capture indoor
scenes at an aperture of 3.5 with an ISO25 film and you will hear all kinds of
protests. Using current Kodachrome 25 and the Tri-Elmar 4/28-50mm or the
Vario-Elmar-R 1:4/35-70mm at 1:4 would simulate these conditions quite well.
Results would be superior, evidently, but we should reflect
for a moment on the role of the Elmar in its historical and technical context.
For many scenes, the aperture of 3.5 is still an excellent choice as it
combines depth of field, needed for most solid subjects with a slightly safety
margin for inaccuracy of focus and a good image quality. Using the f/16 rule,
we can estimate the shutter speeds to be useable with an ISO20 film as 1/20 sec
at f16 or 1/500 at 1:4 in the sun or 1/10 at 1:4 for a cloudy, dull scene with
side lighting. The latter conditions are at the limit, but within the grasp of
an able photographer.
2.2.1 The family of lenses for the Leica.
It is not clear if Barnack did envisage the use of
interchangeable lenses, when he laid out the basic principles of his design. At
first no additional lens has been offered and later they had to be matched
individually to the body. A sure sign that Leitz was still in
the beginning of the learning curve of becoming a manufacturer of high
precision camera systems. This is remarkable as Leitz were well
acquainted with the production of microscopes. These instruments were designed
to be individually calibrated, which could not be done in the photography
department. The concept of manufacturing interchangeable lenses to match with
great accuracy to any body was really a challenge to the factory. But the
availability of lenses with different focal lengths and apertures was the
raison-d'-etre for the Leica. From 1930 the Elmar design is stretched in two
directions: wide angle: 35mm and narrow angle: 90mm, with a short excursion
into the 105mm focal length, and the 135mm. In fact we have only two major
configurations in the Leica lens line-up from 1925 on. The triplet basic form uses
a configuration of three singlet elements (positive, negative, positive). If we replace the third element by a cemented
triplet, we get the Elmax. Replace the rear element by a doublet and we have
the Elmar lenses 1:3.5/35, 1:3.5/50, 1:4/90, 1:6.3/105 and 1:4.5/135mm.
Exchange all three singlets by doublets and we get the Hektor 1:2.5/50 and
1:1.9/73mm. Replace the front and rear singlets by a doublet and we get the
Hektor 1:6.3/28mm. Replace only the centre element by a doublet and we have the
Hektor 1:2.5/125 and Hektor 1:4.5/135 and Thambar 1:2.2/90mm.
Figure 8: 2.2.2 A: basic lens types
The second basic form is the symmetrical doublet or
double-Gauss type. This lens had been expanded to a six element version by Lee
and the basic configuration has a grouping of singlet, cemented double and
cemented double, singlet in a slightly asymmetrical sequence. This is also the
design of the Summar. Split the first singlet into two separate lenses and we
get the Summitar and first Summicron versions. Split the rear singlet into two
air spaced elements and we have the Xenon/ Summarit versions. Use this
configuration, add a Merté surface and we get the Summarex 1:1.5/85mm.
When adding the long focus designs of the Telyt 1:4.5/200 and the 1:5/400mm,
the complete lens line is covered by three basic designs. There are several
reasons for this economy of design types. First of all, there was not enough time
and theoretical knowledge to explore more exotic designs. Several studies done by
Leitz designers in the early period, generated some very promising and exciting
designs, but lack of suitable glass types, the required mechanical tolerances
and careful assembly made such designs not feasible in those days. So the
designer had to use his ingenuity to stretch existing lenses to cover the
requirements. Secondly it made sense in those early days to concentrate on a
few designs and glass types and try to manufacture as many lenses as possible.
One should remember that the cost of designing a new lens was often
prohibitively high and success was not guaranteed.
The designer calculated the lens on paper and had a good
idea of the performance that it could deliver. In those days the designer had
limited means for the calculation of aberrations. Tracing rays through the
optical system is quite labourious. And sometimes, in the case of skew rays,
the mathematics are quite complex. So a designer could not trace enough rays to
get a full picture of the design. The physical prototype was needed for a
practical evaluation of the design. The design on paper was a close
approximation to the desired correction, but the designer could not predict
fully its state of aberration correction. Some mysteries remained, that could be
detected after the prototype had been built. A proven design then was often the
only or more practical way to proceed.
2.2.1.1 The Summar type.
In those days then every additional speed gain would be
welcomed. The Hektor 1:2.5/50mm was the first answer by Leitz to the request
for more speed. Its design in 3 groups of 2 cemented elements each, classifies
it as a triplet variant. It had only six air-to-glass surfaces and was intended
by Berek as the answer of Leitz to the Sonnar designs with 6 air-to-glass surfaces
too. Definition of finer detail on axis could be preserved, but at maximum
aperture overall contrast was on the low side. Films in those days had thick
emulsions and low resolution, which gave the muted sharpness impression of many
older photographs with the 35mm format. The modest wide open performance would
have been in line with what was the accepted norm then.
High speed lenses for the 35mm format were quite difficult
to compute, as the relatively wide aperture introduced a high level of coma and
astigmatism, which softens the outer zones considerably. Furthermore the
absence of anti-reflection coatings produced additional sensitivity to flare
and reduced contrast. And last but not least, the requirements of Leica users
for enhanced recording capabilities, triggered by improved emulsions,
necessitated the search for new optical designs. To reduce unwanted flare and
reflections, one had to reduce the number of air-to-glass surfaces and one
could use one of several types of the triplet (like the Hektor). This triplet
type (consisting of two positive singlet elements (crown type) at front and
rear and one inner negative singlet element (flint type) is the base
configuration. We can then play with this design and use one or more cemented
doublets to replace a singlet element or we can split a singlet into two or
more elements. The second major type is an expansion of the original Gauss
doublet (two air-spaced meniscus elements). When we place two Gauss doublets
back-to-back symmetrically around an aperture stop we get the double-Gauss type
and playing again with the design we can add a cemented element to each
negative meniscus. The classical and justly famous Double-Gauss six element
lens is born. This design type has excellent possibilities for aberration reduction.
So Berek was quite right in his decision to base the Leica high speed 50mm
lenses on the Double-Gauss design and neglect the triplet derivatives (as the
Zeiss Sonnar and Ernostar), even if these had a short-time practical advantage.
Without anti-reflex-coating, the many glass to air transitions of the Summar
lowered the overall image quality. The cross section of the Summar looks
remarkably similar to that of the later Summicron family. The Summar 1:2/50mm
exhibited a low contrast image at full aperture, with a strong presence of astigmatism
in the outer zones and a fair amount of veiling glare. The edges of the fine
detail outlines are fuzzy, which give the impression of a slight unsharpness.
Figure 9: 2.2.1 A Summar lens rigid
The Summar had not the crisp recording of coarse detail that
the contemporary splittriplet- type of lenses could offer, and its image
quality lacked punch. Leitz however, kept on studying and improving the design
and so gained valuable experience, which could be used to good effect when
Leitz introduced the Summitar 1:2/50mm in 1939. This lens has somewhat higher
contrast and noticeably less vignetting, partly due to a bigger front lens . Its general characteristics wide open closely follow
the Summar specs. Its central area of good definition is larger as is the
reduction of astigmatism. Most importantly the Summitar improves rapidly on
stopping down The improved colour correction (longitudinal) and better
definition of finer details were needed to support the new generation of
emulsions. The first Kodachrome from 1935 offered a staggering ISO value of 6,
but the monochrome emulsions, especially the Agfa Isopan Super Special already
had a speed value of ISO100, which is as sensitive as current standard B&W
films. So in a few years time effective film speed increased two- to fourfold
with finer grain than before and demands for better corrected high speed lenses
became louder and louder. The colour negative and transparency films hovered
around 20ISO for a long period and in 1950 reached ISO40. The first ISO100
colour film was the Super Anscochrome from 1957. It is also of interest to note
that the granularity and resolving power values were in a really different
category than what we expect today. The 1930 generation of films had RMS values
that were four times higher than today's films of comparable speed and offered
a resolving power of around 20 to 30 lp/mm in standard lightning conditions. This
value, low as it may seem today, still presented a challenge to the lens designers.
While the performance of the Summitar on axis exceeded the capabilities of the
film with great ease, the image quality in the field (zonal areas) showed a low
contrast at 10lp/mm already. The value of 30lp/mm (an indication of the
recording capabilities for finer textural details) was reached but with very
fuzzy edges and low contrast. The wider the aperture and/or the larger the
image field, the more difficult it will be for a designer to control the
aberrations. The low sensitivity of the filmemulsions and the wish of
photographers to record events in very dim light encouraged the designers to
compute lenses with an aperture of 1:1.5 was busy with a 1.5 design and
finished one in 1938. But in 1932 the Contax offered already a 1.5 lens of
Sonnar (triplet) design and Leitz needed a lens with similar specifications urgently.
So the company licensed the Schneider design, the Xenon 1:1.5/50mm.
From a historical perspective we may regret this decision,
as it would have very interesting to see which design of the two contemporary
giants in optical design (Bertele and Berek) would be the better one. The Leitz
Xenon allowed the user to expand his practical picture taking opportunities
into hitherto uncharted realms. The performance of the lens at full aperture is
however only acceptable for reportage type of pictures. A strong presence of
coma introduces patches of flare around specular highlights. The quite dreamy
and romantic atmosphere of pictures taken with high speed lenses at wider
apertures in those days can be witnessed in many picture examples.
2.2.1.2 The Elmar type.
The Elmar family expanded quickly from 1930 in a few years
to a range of lenses, covering 35, 90, 105 and 135mm focal lengths. Apertures
were a modest 1:3.5 to 1:4.5. The Elmar 1:4/90 at full aperture has low to
medium overall contrast and renders coarse detail with good clarity over an
extended area (image height about 12mm). The 90mm lens started its life not as
a portrait lens as they have been baptized later. In the '30's film emulsions were
coarse-grained and even at moderate enlargements, details in the subject could
get lost in the grain clumping. So the first use of the 90mm was the ability to
record details sufficiently large on film that the grain could handle the finer
image details. The so-called Berg-Elmar 1:4/105mm is a compact and very
light-weight lens of only 240grams. This lens shows the concern of the Leitz company to produce lenses that are compact and light-weight.
These characteristics match the Leica body as the camera for dynamic
photography and the camera as travel companion. The Elmar 1:4.5/135mm again has
the typical Elmar fingerprint of good central sharpness, but showed also
chromatic errors, which are unavoidable in a long focus design. Leitz replaced
this lens with the better Hektor 1:4.5/135mm. The Elmar 1:3.5/35mm at full
aperture has some vignetting and the family character of lower contrast, good
central definition of details and soft outer zones. It is a very compact lens,
as were most of the lenses for the Leica in those days. Leitz even announced an
Elmar 1:4.5/35mm in 1935. It had fix focus click stops for ease of use. This
one and the Berg-Elmar are indications that Leitz made every effort to provide
the user of the Leica with a system of lenses that covered a wide range of
applications. Every type of photography (from travel to commercial illustration)
could be covered with the Leica system. 2.2.1.3The Hektor
type. The Hektor 1:1,9/73mm was the first
attempt of .i.; to design a high speed lens with a moderate angle of field. The
triplet design however could not be stretched too much.
The Hektor 2.5/50mm and the Hektor 1.9/73mm allow the same
amount of energy flow to pass through the lens. The wider aperture of 1.9 (2/3
of a stop more than the 1:2.5) is offset by a 50% smaller angle of field. (30oversus 45o). clearly was aware
of the limits of the generic Hektor design. The 1:1.9/73mm short telephoto lens
at the wider apertures has that typical softness at the edges of outlines and
the very smooth transition of the sharpness plane into the unsharpness blur of
fore-and background, that became the defining characteristic of the portrait
lens for generations of photographers. The Thambar 1:2.2/90mm is a so-called
soft-focus lens that allowed the photographer to make portraits with the soft
effects that were very popular in those days. The extent of softness can be
regulated by the iris diaphragm and a special opaque disc,
that can be placed in front of the lens and cuts off the central rays.
As now only the outer rays will be available for image formation, we see a strong
degradation of quality as these outer rays are generally less well corrected
than he central rays, that are blocked now. Longer
focal lengths have a smaller angle of field and the correction of aberrations
is somewhat simpler as long as apertures are moderate. The wide angle have a larger angle of field and now vignetting and
zonal aberrations are becoming a problem for the designer. The first 28mm lens,
the Hektor 1:6.3/28mm, was introduced in 1935. When
the angle widens the vignetting and distortion ask for all attention. The large
front and rear lenses help reduce vignetting which indeed is gone by 1:8. The
diameter of the front lens is +10mm, twice as large as is needed from geometry
only. The overall image quality is a bit dull and fine detail is softly
rendered without the clarity and bite we do note in current lenses. 2.2.1.4 The
Telyt type. The coupling accuracy of the Leitz rangefinder/body combination is
limited to the 135mm focal length and when a longer focal length is required, a
new device, the mirror reflex housing (Visoflex or the as it has been designated
in the early period PLOOT), had to be used. This device, ingenious and a marvel
of mechanical engineering, transforms the Barnack camera into a mostly stationary
photographic instrument. For some assignments however the longer focal lengths
are indispensable. The Olympic games in 1936 were a
strong incentive to produce new telephoto-lenses and the Telyt 1:4.5/200mm and
1:5/400mm were designed with this deployment in mind. Both Telyt lenses were
true long focus lenses, and performance is good for the 200mm version, but weak
for the 400mm.
2.2.2 The war period.
During the war Leitz managed to keep the production lines
going, but production increasingly shifted to military, medical and projection
purposes. The only new lens to appear for the presumably civilian market was
the Summarex. This lens is described in the beautiful and colourful Leitz
brochure from 1943: 'auswechselbaren Leica-Objektive'. The state of the economy
would not have encouraged anyone to buy this lens, even if it were available,
which is highly unlikely. The price in 1936 for the Hektor 1:1.9/7.3 cm was RM
260.00, way beyond normal purchasing power. The Summarex 85mm had modest image
quality at full aperture but improved remarkably on stopping down to 1:5.6 and
1:8.
Figure 10: 2.3A- lens overview 1943
The last lens-series has been registered on March 15, 1944.
It was a batch of about 10 lenses of exotic specifications: 1:0.85/150mm, the
last lens having the serial number #594852. The list of specialized designs is
quite long. Examples are the Elkinor 1:1.5/300 and 1:1.5/400mm, IRSummar
1:0.85/75mm and 1:0.85/150mm, URSummar 1:1/90mm and 1:1/150mm. Coating was also
introduced on the military versions of the lenses in 1941. The civilian
production however did not receive this treatment until after the war. The
topic of the use of coating has not found its final treatment. Rogliatti notes
that coating had been applied to lenses from #587601 (Summitar) from 11
November 1945. The factory lists indicate that at 6 November 1945 the serial
numbers 600000 to 601000 have been allocated for a batch of Hektor 13.5cm
lenses. The first Summitar series begins with 603000 in early 1946. The #number
587601 is part of a batch of Summitars from #586001 to 590000, allocated in
1942! The factory archives in the optical department note that coating was
applied from October 1941 (from serial# 580000) and I quote:' from October 1941
all Leica lenses receive anti-reflexion coatings. These lenses are not
available for amateurs, but only for war photographers
(Kriegsberichterstatter)'. A report by the US Naval Research Labaratory, dated
3 October 1946, on German coating methods during the wartime does indeed note
that Zeiss, Leitz and Schott all used several methods of coating during that
period. The report remarks that Dr. Männchen from Leitz demonstrated an
(experimental) method of centrifugal coating (as compared with the Zeiss method
of thermal evaporation coating), which however produced a very soft coating.
Production was resumed already in June 1945 with serial
number 595000: a batch of 1001 Elmar 1:3.5/50mm lenses. Between October 1941
and June 1945 Leitz manufactured about 15000 lenses, and some larger batches
were reserved for the specialist lenses for military applications, some of
which were very rare, like the IRTessar 5/5000mm. It is reasonable to assume
that most of these 15000 lenses went to the government. As it is also very
likely that not all lenses, produced before October 1941 were sold out at that
date and Leitz would be able to provide amateurs with the non-coated version of
their lenses.
2.2.3 The postwar period till 1957.
The Leitz factory emerged after the war almost intact. The
buildings however were in a desolate state and the American Occupation had
copied and transferred all the valuable Leitz documentation to the USA. It is
typical of the wisdom and vision of Ernst Leitz that he was not bothered at all
by this transfer and he remarked to a close aid that as long as the Americans
allowed the workforce in the factory to continue with their work he was not
worried about it at all. He knew what is still valid today.
The real power of the factory is the creativity, brains and
motivation of the people.
Already in the spring of 1945 Leitz-Werke resumed production
with the same core product line as was available in 1939. The first new design
was the Summaron 1:3.5/35mm, introduced in 1949. Like the Hektor 28mm, the
diameter of the front and rear lenses were quite large for the same reason
(reduction of vignetting). This six element Gauss design had more potential for
aberration correction than the Elmar of the same specifications, and its
introduction so soon after resuming civilian production, showed the
determination of Leitz to improve the image quality of the Leica lens line.
During the war Leitz introduced single layer coating (around 1941) and after
the war new glasses became available, many of them specifically researched in
the Leitz glass lab, founded in 1949. In the past the optical designers had
stated too often that they could get better correction of aberrations if they
could use a glass with specific properties. These glasses however did not exist
and the idea to have to wait and see if and when Schott or other manufacturers
produced the glass so eagerly awaited for, was not in the interest of Leitz. In
1950 the Summarex 1:1.5/85mm was relaunched in a chrome version and with coated
glasses. Its inherent characteristics were not changed and, while having an
impressive appearance and a heavy weight , image
quality became quite good only after stopping down to 2.8. At full aperture the
lens gave modest performance, and its weight must have been a heavy burden for
the Leica bodies of the day. Indeed, the structural weakness of the screw mount
Leica bodies when the heavier lenses were mounted, was
one of the arguments to develop the M-series. The Summaron 1:5.6/28mm (1955)
improved upon its predecessor. Its design is symmetrical which gives better
performance when taking pictures at closer range. Also distortion and coma are
slightly reduced by this design. Overall it is medium contrast lens with a
commendable definition of fine detail over a larger image circle of about 12mm.
The outer zones are soft to very soft. The wide angle of field gave the
designers a lot of trouble and an even coverage including the edges was still a
pipe-dream. When considering the introduction of a new lens, Leitz had to weigh
two factors. The high cost of the design process and the number of units sold.
But sometimes the Leitz designers had no suitable design
available. The 21mm lens is an example. The optical solution for the design of
an ultra wide angle lens had been found by Zeiss and Schneider. Leitz offered
the Schneider Super-Angulon 1:4/21mm from 1958. This is a symmetrical design
that covers a wide angular field.
Its construction incorporates negative outer elements and a
wide spacing between the elements. This type of construction helps flatten the
field, which is very important for a wide angle. Coma and distortion are also
reduced. The Super- Angulon has indeed very low distortion and slight curvature
of field. It also exhibits the low to medium contrast typical of the period and
the rendition of fine detail is on the soft side.
Figure 11: 2.4 A-early Summicron
The collapsible version of the new Summicron 1:2/50mm (1953)
is a lens that marks the transition to the next period. While the lens itself
belongs in the first period, I will discuss it in relation with the newer rigid
version, that is specifically designed with the new
M-mount in mind.
2.2.4 Summary for first period 1925 to 1957.
Theoretical knowledge of the optical aberrations and how to
correct them is based on the work of some eminent scientists of the period
around 1900. Schröder, Moser and Miethe studied the equations needed for
the complex calculations for photographic lenses, Abbe experimented with new
glass types, and people like Rudolph, Merté and Lee explored unknown
territory to get the best imagery possible.
The designs they used, (triplets, Petzval versions,
double-gauss derivatives) were all based on constructs from the 19th century
and it is in fact remarkable to note that most of the early efforts were
concentrated on improving old designs and specifically to get rid of
astigmatism. It is also interesting to remark that in the period form 1900 to
1940 the demand for photographic lenses was so high that most designers at
first had to generate designs that could satisfy that tremendous demand. Room
for fundamental research hardly existed. Commercial arguments prevailed in this
expanding industry and the manufacturers exploited the available experience to
the utmost. The widespread use of the Leica and its 35mm format and the expansion
of focal lengths into the wide angle area, put heavy
demands on optical performance and so did the surprising introduction of colour
film around 1935. It became clear to some designers that the traditional
methods and formulae for the calculation of aberrations did not suffice and
that even the theoretical knowledge of the aberrations themselves deserved a
reworking of basic principles. The geometrical (trigonometric) approach to lens
design was insufficient to study and correct the higher level of aberrations
that needed to be controlled for better image quality. The direction to follow
was a study of the energy balances in the optical system, Berek understood this
very well and he even devoted a full chapter to this topic in his book 'Grundlagen
der praktischen Optik' (1930). This chapter 10 ('Energiebilanzen') is part of
the Leitz philosophy of lens design. The staggering amount of numerical results,
needed to analyze lenses in this new theoretical approach, could not be handled
with the classical log-tables and numerical methods. Indeed the full use of these
insights had to wait until the computer became available early in the '50s. Max
was not the only one to run against the limits of the traditional methods of computation.
Albrecht Tronnier of Nokton fame was also acutely aware of the vanishing
possibilities of the old approach. War however changed the course of humanity
for a long period. Behind closely guarded doors the research continued, and the
foundations for this new approach were intensively studied. Also new evaluation
methods for the appraisal of optical performance were introduced as the mirror
image of this research. The first studies into the importance of contrast for image
quality were laid down by a Zeiss employee, Hansen in 1943. The period from 1925
to 1957 can be characterized as the search to adapt the known optical
principles to the exacting demands of the Leica format and of colour
photography. A few proven designs were selected as the platform on which to
construct and manufacture the lenses that were commercially and conceptually
required by the quickly expanding user base for the Leica camera.
Figure 12: 2.5 A :the Leica system Study of the production
Figures over the period 1930 to 1940 indicates that a total
of 414.000 lenses left the factory, of which 169.000 were of the type Elmar
1:3.5/50mm and 123.000 of the type Summar 1:2/50mm. The Elmar 1:4/90mm and the
Elmar 1:3,5/35mm had a production of 28.000 and
26.000. More than 70% of all lenses that were produced by the 'Leitzianer' in
the Wetzlar 'Hochhaus' were of the 50mm focal length. The 35 and 90mm lenses
added another 13%. These figures are a clear indication of the commercial
viability of new designs, and when we add the worldwide recession after 1929
into the equation, it is sound business practice of Ernst Leitz to stick to the
well-proven designs.
Figure 13: 2.5B: the Leica image
In the late thirties the Leica user could choose focal
lengths from 28 to 400mm and apertures from 1:1.5, including some special
lenses like the Berg-Elmar 1:6.3/105mm and the Thambar 1:2.2/90mm and
(hypothetically) the Summarex 1:1,5/85. While it is true that some of the Leica
lenses overstretched the optical limits a bit, designing and producing the
lenses was of paramount importance for the next period. Leitz tried to expand
the system concept of the Leica as quickly as possible as he understood quite
well, that the success of the Leica depended on a comprehensive line of lenses,
that provided most users with every option they could imagine. When evaluating
the optical performance of a lens, designed and produced in those days, one
should take into account that the lenses then were not checked and tested as we
do today. The British report states that the Summitar was checked for image
quality at the factory at the aperture of 1:3.2. Presumably the lens testers
used some numerical rules to relate the quality available at 1:3.2 to the one
at full aperture. The success of the 'Leica'-camera was in part based on the
large depth of field of the lenses, coupled to the 35mm format. The depth of
field is greater when one stops down to middle
apertures, and this technique was used in most circumstances. The full aperture
setting was primarily reserved for the really difficult low light situations, where
slow shutter speeds and high speed, grainy film-emulsions would not record fine
detail at all. It was a fact of life that the image quality available at full
aperture would be less that what could be delivered at smaller apertures.
Part 3: The challenges, 1957 to 1988
The Leica company had to face
several challenges during these decades, some of which were optical in nature
and others were related to the course of the photographic industry. The German
'Wirtschaftswunder' generated unprecedented economical wealth,
photography became the prime hobby for large segments of the population in the
industrialized western part of the world. In 1955 the Leitz company,
now located in three different factories in and around Wetzlar, produced 40.000
camera bodies a year and close to 100.000 lenses.
Figure 14: 2.3 A: lens # 2.000.000
In 1964 the production of lens# 2.000.000 was proudly
announced, 12 years after #1.000.000 in 1952. In those days Leitz manufactured
a thousand different products, which were assembled from 300.000 discrete
components. Every month millions of parts had to be shop tested for tolerances,
often less than 0.01mm. Such a diversified range of products, most of which
were of high precision optical-mechanical nature, put a strain on the
production and engineering divisions and of course on the design department
too. Manufacturing technology was in essence the same as before the war,
improved and modernized of course, but manual labour constituted the bulk of the
workflow. Around 1965 the economy slowed down and the photographic industry
became dominated by the Japanese and one main camera type, the single lens
reflex. The high quality compact rangefinder camera was the second strategic product
for the market. Leitz was squeezed by both product types, for which they did
not have an immediate answer. The Leicaflex (1964) and the Leica CL (1972) were
too late and of too modest specifications to have any impact on the market.
The optical challenges were large too. Japanese designs
covered an ever growing range of lenses of improved optical performance. The
Leitz philosophy of design and manufacture of lenses made them less flexible in
the area of extended specifications. The universal adoption of computer aided
optical design gave lens designers equivalent opportunities and in the
seventies one might note a plateau in lens performance. The leading marques
produced lenses that delivered image quality on a level that satisfied most
professionals and amateurs alike. It became increasingly difficult for Leitz to
distinguish themselves and create optical systems of
cutting edge performance. The amounts of money for the required research and
development to stay ahead were simply not available, as the falling production
of the Leica cameras reduced the profits of the company and asked for
investment money too. The challenges seemed indeed insurmountable. On the
organizational and management level, Leitz worked within the constraints of a
traditional method of manufacture and decision making and the famous feeling of
Leitz for the trends in photography was fading away too. On the marketing
level, Leitz had to cope at the same time with the problem of a smaller
production volume and the need to expand the range of products. On the level of
design and innovation one has to note that the Leitz designers were quite
innovative within their chosen domain of excellence, the high speed lens of
fixed focal length, and were very advanced in the research for new solutions
(glass, aspherics, even vario lens designs and even auto focus). Leitz focused
all attention and development on one goal only and that was the development of
lenses with very high image quality, which could only be delivered when the
mechanical mounts were of very high quality too. . The true dilemma is simple:
the high optical performance necessitates an expensive production process and a
mechanically elaborate mount , which in turn defines
the physical dimensions of the lens and as these dimensions cannot grow beyond
manageable proportions, the specifications are fixed too. Leitz started the
period in great prosperity, with an enviable product line, that sold very well,
and a fine tradition of design and manufacture. The allowed themselves the
luxury of two different and sometimes competing design departments in Wetzlar
and Midland and a glass research lab. Later Leitz even added a very elaborate
and costly technical lab for testing lenses and designing new methods for image
evaluation. The big changes in the photographic world and in the photographic
industry were only reluctantly acknowledged and Leitz tried with cooperation,
and the adoption of lenses made by several manufacturers to expand the product
line, while staying faithful to the traditional values of optical quality and
superb mechanical craftsmanship.
Figure 15: 2.3B: telyt 3.4/180
Figure 16: 2.3C Macro Elmarit
Figure 17; 2.3D Noctilux 1/50
In 1972 the Leitz family stepped back from the management
and in 1974 Wild Heerbrugg was in complete control. Not much changed however
and the next period from 1974 to 1988 is characterized by a painful process of
global marginalization.
Some outstanding products were created in this period, like
the M6 and R6, the Noctilux and the Apo-Telyt 3.4/180 and the
Apo-Macro-Elmarit-2.8/100mm, but the impact of Leica on the market had been
lost.
2.3.1 Evolution of the Leica camera.
During this period the photographic division of the Leitz company did indeed change dramatically. With the M3 and its
range of lenses, Leitz established itself as the premium manufacturer of precision
miniature rangefinder cameras and was without doubt the photographic company
with the highest esteem in the world. The M3 with the Noctilux 1:1.2/50mm
(1966) confirmed the position of Leitz as a world leader in advanced optics and
the role of the M3 as the available light/reportage camera.
Figure 18: 2.3.1 A M3
While Leitz concentrated a substantial part of its resources
on the rangefinder concept, the photographic world at large gradually shifted
to the SLR-concept. This change occurred in a few years and must have been a
surprise to Leitz. They were the only company in the world that expressed faith
in the rangefinder concept for a professional camera system. The factory
however could not overlook the drop in sales figures for the M-series. Leitz reluctantly
followed the trend with the Leicaflex in 1964. The birth-pangs of this camera
have been noted in the contemporary press and the announcement of a reflex
Leica were rumoured since 1958. The design and development of the reflex line
is not part of this story. Suffice it to say, that the Leicaflex and its
successor models were less successful than Leitz anticipated. The development
of a comprehensive lens-line to compete with the professional Japanese and
German models of those days was no easy task. At the same time the sales of the
rangefinder camera dropped significantly. In 1960 the sales of the M2/M3 were around
30000, and a few years earlier were above the 35 thousand level. But in 1970 the
M4 sales were around 12.000, and the Leicaflex SL had a designated production run
of 10.000. Leitz was in a difficult position as its market share shrank, while
the investment for research and development had to increase to compete with the
ever more innovative Japanese companies.
Figure 19: 2.3.1 B M System
Being alone in the rangefinder market, Leitz tried to break
out of the M3/4 mould, with bold new designs. The M5 was clearly targeted at
the professional photographer who was accustomed to features like the through
the lens exposure metering. The CL, on the other hand incorporated modern
features that tried to open up the new vast market of the dedicated amateur
photographer. These models, with their quite innovative engineering, did not
capture the mood of the market however and were withdrawn after only a few
years. In fact the fate of these models dealt a heavy blow to the confidence of
Leitz in the RF-future. Is was decided in Wetzlar to
end the RFproduction and to concentrate on the SLR as the main product line.
The dedication and enthusiasm of a few persons, saved
the M-system: the production of the camera was transferred to Midland (the
M4-2).
Figure 20: 2.3.1 C: Leicaflex
The Leicaflex SL2, the last of the 'true' Wetzlar reflex
bodies, had run out of steam already at its introduction and Leitz, following
the maxim 'if you cannot beat them, join them', teamed up with Japanese
suppliers to produce the next generation SLR, the Leica R3. The history of the
R-development is not part of our story. Suffice it to note that the Leica R3
and the successor model R4 sold well, but did not inspire that confidence in
the solid engineering Leitz were famous for. The R5 and especially the R6 and
R7 restored the reputation, but now we already entering the nineties and that is the Solms-era. The cooperation with Minolta did extend
into the optical area, and Leitz used several Minolta designs to fill in the
gaps of the lens line for the R-system.
In the mid-seventies the photographic industry reached a
performance plateau: the SLR development had virtually stopped after the
incorporation of Auto-Exposure techniques. The lenses too, helped by computer
assisted design programs, had evolved to a state where it was becoming
difficult for the casual and critical observer alike to see big differences in
image quality between the leading marques. The independent suppliers of lenses, improved their optical quality to a level that
started to challenge the leaders, at least optically. And with a very
favourable priceperformance ratio, coupled to innovative designs, these makers
made substantial inroads in the traditional markets. On the optical front, the
Leitz designs, while still first class, ran into stiff competition from several
Japanese and German companies, whose design teams, aided by computer programs
and a large selection of commercially available glass types, produced
lenses that were sometimes uncomfortably close to the Leitz designs in image
quality. The domains where the Leitz performance could show its advantages
(slide projection and large format B&W prints) were shrinking. The
increasing use of the colour-negative print also diminished perceivable
differences in mage quality. Leitz had one very important advantage,
that is often overlooked: the mechanical quality and engineering
precision of the lens mounts is second to none. The longevity and durability of
Leitz lenses can be seen in any older lens that is in use today. Leica lenses
also stay within tolerances, even after prolonged and heavy-duty use.
Figure 21: 2.3.1D: Lens mount
The Leitz company started the 35mm
boom in photography in 1925. The company was sold to Wild in December 1986 and
the production facilities transferred to Solms in June 1988. It is a strange
quirk of history that Leitz always went against the current mood: in the
twenties when against all odds and advice the Leica Standard was produced, in
the fifties when the decision was made to continue the development of the RF
system and in the eighties when they decided to stick to the manual focus SLR.
The Leitz management then had to accomplish a true mission impossible. They
felt pressed to offer a comprehensive range of lenses for two different
systems, they had to improve the optical quality and at the same time contain
the cost of production to adjust to a lower sales volume of the R- and Msystems.
These decisions had important consequences for the
development and characteristics of the Leitz lens families. For the rangefinder
and reflex systems the company wanted state of the art performance for their
lenses. Where 'state-of-theart' was to be defined within the
limits of the Leitz philosophy of lens design. The designers had set
themselves a certain minimum level of image quality. When you have designed a
lens (focal length, aperture, volume, number of lens elements etc.) you can
calculate the (theoretical) performance of a lens with a technique called 'ray intercept
curves'. These curves can be graphically displayed and their shape is very important
as they give the designer a full insight into the quality of the lens. Now you can
define that the bundle of curves should be within certain upper and lower
limits.
Any lens that has curves that extend
beyond these limits, delivers less image quality than a lens that has graphs
within these boundaries. Leitz did design several prototypes for the
R-system, but performance at full aperture could not be fitted within these
optical limits (and the production constraints). They might have been very good
lenses by most standards and would certainly have drawn attention to the Leica
R, but Leitz, bound by their own strict performance levels could not be persuaded
to manufacture this lens.
2.3.2 An Herculean task From 1957 to 1988
Leitz introduced more than 40 new or redesigned lens types
for the M-system and more than 50 lens types for the R-system,
on average three lens types per year. The Leicaflex has been introduced on the
market in 1964, but the design of lenses for this system started around 1960.
At first the designs for the R-system were optically completely different from
the M-equivalents. In 1964 the Leicaflex was accompanied by four lenses:
1:2.8/35mm, 1:2/50, 1:2.8/90 and 1:2.8/135, the classical quartet. All four
were new optical designs, and of Canadian origin with the exception of the 35mm
lens, which was created in Wetzlar. The larger throat diameter of the R-body and
the bigger distance from lens flange to film plane (27.8mm for M and 47mm for R)
force the designer to create retro-focus designs for lenses with focal length
below 50mm, which are more difficult to design. The retro focus principle also
requires larger front elements. There is a 'hard' rule in optical design that
states that a physically larger lens is easier to correct for aberrations than
a smaller lens. And the type of correction could be different too. Leitz then
needed two distinctive design philosophies so to speak, one to optimize the
smaller M-lenses and one to optimize the larger R-lenses. How did Leitz cope
with these incompatible challenges? In the beginning the factory followed two
different design strategies, one for M and one for R. In the M-domain, they
pursued to improve the image quality of existing designs and to develop lenses
with daring specifications, like the Noctilux. These new lenses were made
possible by the advances in optical knowledge, partly by theoretical study, creative
inspiration and the assistance of the computer and partly by the research into new
types of glass. In the R-domain, Leitz started quite conservatively with lenses
of moderate specifications and forged ahead later with longer focus lenses with
apochromatic correction. Of course there was a sharing of insights between the developers
of the M and R lenses and this can be seen quite clearly in the development of
the 50mm lens.
Figure 22: 2.3.2 A reliance on manual labour.
One of the first strategies to be used by Leitz is the
pooling of resources by designing lenses that could be used in the M- and
R-systems. One example was to use the long lenses, developed for the Visoflex,
in a different mount for the Rsystem.
The later Summicron 1:2/50mm lenses, the Summilux 75/80
lenses and the Elmarit 1:2.8/135mm lenses are examples of this trend too. We
should note however that the two design-departments (ELW and ELC) were
thousands of kilometres apart, and pursued their own design philosophy under
two independent management teams and so the potential for competition was as
much part of life as the potential for synergy. Another strategy is to
cooperate with other lens manufacturers. The range of lenses that Leitz could
develop themselves was limited, so for a number of lens designs
, notably zoom lenses and very wide angle lenses, the factory used
lenses from third parties, like Schneider, Zeiss and Minolta. Behind these
strategies, the true battle of that period is between the drive to improve the
image quality and the possibility to translate this improvement into the
production of the lens, while holding down the cost of manufacture. The use of
the computer, new research into optical glass, better grasp of the optical
theory and more insights into the design process all helped to improve designs
significantly. There is a hint of a true Greek drama here, that Leitz succeeded
in advancing the theory of optical design, but could not find the right way to
produce these exciting designs. First of all: the method of producing,
assembling and mounting lenses did not change over the years and had much in
common with the way lenses were manufactured in Berek's days. But the improved
designs needed tighter production tolerances, that
were not possible without a higher cost. And that was the economic constraint
as the lower sales volume could not bear this.
2.3.3 Evolution of the lens systems for M and R.
The development of the lens systems for R and M has to be
described and interpreted against this background. We can illustrate these
developments with a few significant acts. From 1954 to 1971 (the period of the
M2/3/4) Leitz expanded the M-line into a full blown system concept, with lenses
from 15 to 800mm. At the same time the factory redesigned many of the lenses,
for improved performance, and/or easier handling. In several instances the
drive for enhanced optical quality collided with the limits of production
technology and/or production costs. After 1971 (introduction of the M5 and CL)
the rangefinder concept was developed more selectively with lenses in the 21 to
135mm bandwidth. But the reduced sales forced the design-department to
economize on the introduction of new designs and find ways to control the
production costs of existing designs. Between 1975 and 1980 Leitz introduced a
number of very high speed lenses for the resurrected M4-2 (and M4P), that could
be seen as the state-of-the-art profile for the rangefinder concept.
From 19080 to 1990 no new lenses for the M-system have been
constructed.
Figure 23: 2.3.3 A CL
Figure 24: 2.3.3 B M Lenses
For a long time Leitz maintained and
expanded the M-system as the dedicated tool in the great tradition of the
documentary photography. Many lenses were redesigned for improved
optical quality and/or better handling. (smaller
volume, less weight).
Some exciting designs like the Noctilux 1:1.2/50mm indicated
the dedication of Leitz to deliver the best possible optical quality to its
demanding customers. But the system-concept of the M was not abandoned after
the introduction in 1964 of the Leicaflex and for a longer period Leitz had to
develop two lens lines which at the long focus end were competing with each
other. This also blurred the distinction between the two systems. With the
Telyt lenses (400mm to 800mm) Leitz used the same optical cells in different
mounts, which reduced cost of development, but hardly of production, one may assume.
Figure 25: 2.3.3 C Noct 1.2/50
The evolution of the R-system can also be sketched in some
clearly defined stages.
After 1964, the year of the introduction of the Leicaflex,
the development of the lenses for the R-system proceeded slowly. For the
R-system Leitz concentrated on these focal lengths they had most experience with, that is the lenses that were also the backbone for the
M-system: 28mm to 135mm. The maximum apertures however were quite modest (even
for those days). The 35mm, 90mm, 100mm, 135mm and 180mm all had apertures of
1:2.8. The first 1:1.4 design was the 50mm Summilux in 1969, ten years after
the M-version.
Figure 26: 2.3.3 D R-1.4/35
The aperture of 1.4 for the 35mm had to wait till 1984,
before Leitz felt confident enough to produce he
Summilux 1:1.4/35mm. The 1:2/35mm lens arrived in 1972, more than a decade
later than the M-version. One may only guess why Leitz has been so conservative
in the specifications for the R lenses. The obvious commercial argument might
be that they did not want to have competition between the Leitz camera-systems.
Leitz obviously had defined the SLR-system as a complement to the M-system, not
as a competitor. More philosophically one might argue that the focusing screen
and the mirror box with its time lag, typical of the SLR-system, dedicated the
Leicaflex to a different type of photography and Leitz designed lenses exactly
for that goal. Leitz, with its rangefinder expertise, may have had the opinion that
the inaccuracy of focus with the split-image rangefinder of the SLR precluded the
use of very high speed lenses. And on a more mundane level, one should be aware
that Leitz had hardly any experience with the design of lenses for a SLR system.
Figure 27: 2.3.3 E R19
Around 1975, more than 10 years after the Leicaflex appeared, the first wavelet of innovations for the R-system
was announced: the 19mm and the Apo-Telyt 3.4/180mm, the first apochromatically
corrected lens, developed by Leitz Canada. It is interesting to note that the
Wetzlar designers were more concerned with the technique of aspherics and
Midland with the apo-correction, at least in those days.
The M5 and the R-body however required many lenses in the
wide angle and standard focal lengths to be redesigned as retro-focus lenses.
Leitz however had at first hardly any experience with retro-focus lenses for
Slr-systems. That learning curve is visible in the many redesigns of the 28mm lenses
for the M-system and the 35mm lenses for R-system. The first original Leitz
designs in the very wide angle domain are the Elmarit-R 1:2.8/19mm and the
Elmarit-M 1:2.8/21mm in 1975 and 1980, an indication of the long gestation
period that Leitz needed to feel confident to design these types of lenses.
After 1980, when the development for the M-system stopped, the R-system
expanded rapidly with landmark designs like the Summilux-R 1.4/35mm design
(1984) with floating elements and the Apo-Elmarit-R 2.8/100mm (1987). At the
same time, Leitz adopted many designs from Minolta and others. It is evident
that Leitz had not the capacity nor the money to
undertake a massive development program for two different systems and we see a
relocation of forces over time, and a switch of main focus from Rangefinder to
SLR after around 1978.
Figure 28: 2.3.3 F: expanding system
2.3.4 The quest for image quality
As suitable glass types seemed to be the biggest obstacle
for the advancement of new designs, Leitz set up its own glass laboratory and
around 1953 the first fruit of that effort was the introduction of the
Summicron 1:2/50mm (collapsible). Its design was completed in 1949 and a
slightly different computation became necessary in 1952 when Schott glass had
to be incorporated. After the introduction of the Summitar, the 'Leitz
Rechenbüro' continued to explore this design as it was clear that the competition
would not sleep and a high quality 2/50 lens was of utmost importance for the
critical Leica user. In those days the limits of the diameter of the lens mount
of the current models became painfully clear. When wider apertures were needed
or a higher level of correction, the diameter simply was too small. The patent
literature of the four-tap bayonet mount for the later M-models, specifically
mentions the reduced vignetting as one of the prime characteristics. The first
Summicron version closely followed the design parameters of the Summitar and
the increase in image quality was slight. It too is a low contrast lens, and shares
the optical 'fingerprint' with its predecessor. The
rigid version of the Summicron 1:2/50mm arrived on the market in 1957, but was
on the drawing board several years earlier. It utilized the optical
improvements that were made available with the wider throat for the M-body.
The seven-element Summicron split the front element in two
separate lens elements and used the air space between the two elements as an
additional lens (an air lens or Luftlinse). Leitz also separated the front
doublet and used the same technique for enhanced correction, specifically for a
reduction of spherical aberration and of the secondary spectrum (the chromatic
aberrations). The Leitz designers also incorporated in the new design the new
Lanthanum glass types, developed by the research lab. The air-lenses are a
mixed blessing though as they allow for a better control of aberrations, but
also introduce more flare and they are very sensitive to small deviations in
tolerances. After 1957 Leitz introduced in quick succession several lenses for
the M-system, that can be regarded as the first series
of lenses with markedly improved imagery, when compared to the predecessors.
The image quality of these lenses was a true testimony to the progress in the
optical and mechanical fields, that the Leitz
engineers had been able to make since the war ended. In the late thirties Leitz
had noted that the traditional methods of design, the knowledge about aberrations,
the commercially available glass types and the methods of production would
seriously handicap the search for new optical solutions. In 1949, the year that
Max Berek died, the company had inherited a large body of knowledge and had accumulated
a vast pool of experience. One should remember that much of optical theory
originated in the microscope division, which was very valuable but needed some
fundamental transformation before it could be used with good result in the photographic
department. The microscope lens has a much smaller field than the standard
photographic lens and aberrations which are very troublesome in photographic
applications are less so than in the microscope. The converse is true too, of
course. As example I may remark that apochromatic correction of microscope
lenses was standard practice, long before this approach was applied in photographic
lenses. A major part of the research and design efforts of Leitz in the period
from 1950 to 1980 was devoted to the exploration of the double-Gauss highspeed lens.
As I mentioned before, this lens type is very flexible and can be adapted to
several classes of demands. A lower speed wide angle lens as the Summaron 1:2.8/35mm
and the high speed Summilux 1:1.4/50mm or the Summicron 1:2/90mm ask for a different type of aberration correction. The
difficulties in finetuning the several sets of requirements can be seen quite
clearly in the long list of lens versions for the M- and R-systems. Some
designs were replaced very soon by improved versions: witness the Summicron
1:2/90 (1959) and the Summilux 1:1.4/50mm (1961). Here we see a trait that is
most characteristic of Leitz lens development. The optical department in the
Wetzlar 'Hochhaus' tried to refine and improve the existing lenses
continuously, and they committed a fair share of their resources to that end.
Leitz would never be content with the idea that a certain lens is good enough.
His drive was to provide the photographer with the optimum optical performance.
The many redesigns of classical focal lengths in the 28mm-135mm range do show
this intention. Sometimes a lens is not changed for decades. The Summilux
1:1.4/35mm is an example. At full aperture it has low contrast and it suffers
from coma. Its overall characteristics would indicate it to be a Summicron 1:2/35mm, opened up one stop. Given the constraints (optical and
mechanical) of a high speed wide angle lens, the definition of this lens could
not be improved upon. It is perhaps more accurate to say that the Leitz
designers could, within certain limits, create a (theoretical) lens that could
deliver better performance, but it would be very difficult to produce such a
lens. We do not often realize that suitable glass must be available, that the
production tolerances must be realistic and that the amount of manual
adjustments during assembly should be low. The designer might calculate a very
fine lens, but if the glass needed has a thermal expansion that is not in line
with the other glasses, or has properties that make it difficult to manipulate
(grind, polish), he cannot use it. It took almost 30 years before the technique
of aspherical lenses could break this Gordian knot for this type of lens. The
first Noctilux 1:1.2/50mm with two aspherical lens-surfaces (1966) does
indicate the relationship between the several variables. This lens was
optically a breakthrough, but the amount of manual labour, the exactitude of
the tolerances and the overall optical/mechanical sensitivity denied the lens
the commercial success, Leitz had hoped for. In the category of very high speed
lenses (1.4 and wider) Leitz redesigned the 1:1.4/50 in 1961 and introduced the
R-version in 1970. The R- lens stayed in production until 1998 when a much
improved version with 8 elements appeared. The M-lens is still in production,
but begins to show its age. It is however not that easy to design a version that
delivers improved image quality while staying competitive in price and physical
dimensions. The long production period of the Summilux 35 mm (M and R-version) is
a clear indication how difficult it is to improve on a well designed lens when
the parameters are really difficult (1.4 aperture and an angle of 64o are heavy
obstacles for a designer.). The Summilux -R for the 35mm focal length was a
Wetzlar design and they needed a vast array of optical means to improve upon
the M-version from 1961. The M-version had to be designed within the limits of
a compact lens and it may represent what was possible with conventional means.
The R-version could be designed with more liberty concerning the size and so a
ten element optical system with floating elements emerged from the drawing
board.
2.3.5 Containment of costs.
During and after the mid-seventies, Leitz tried to
rationalize their lens systems, sharing mounts and lens elements wherever
possible. As example we may look at the Summicron 1:2/50mm standard lens. This
lens family is an example of the double goal for more image quality (and cost
containment at the same time). The Summicron (type I) was the collapsible one
from 1953 or 1954. The Summicron (II) for M was introduced in 1957. For the new
R-system (1964) Leitz Canada designed a different lens, now with six elements
and higher level of symmetry and different glass types. A few years (1969)
later ELC introduced the Summicron (III) for the M-line, a design that differed
from the predecessors and generally offered higher image quality.
In 1969 the Summicron for M had the same overall
construction as the Summicron for R, introduced in 1964 and the reduction from
7 to 6 lens elements improved the imagery and at the same time reduced the
assembly and production costs. There was evidently one element less to grind,
check and mount and the performance was improved at the same time. The next
stage of the design for the R arrived in 1976 with the Summicron-R (II). This
lens is a highly evolved design that delivers high image quality, better than
the predecessor, especially in its behaviour at full aperture.
It gave a high contrast image with quite even performance
over most of the image area. An almost identical version for the M-system was
introduced in 1979 with the Summicron-M (IV). Both versions are remarkable as
they have 5 plane surfaces, instead of the curved surfaces of the other types
of Summicron lens. A plane surface is easier to manufacture and handle, so its
use will lower the production cost. In addition a plane surface does not
generate unwanted reflections, and it is sufficient to use a single-layer
coating.
Figure 29: 2.3.5.A M50
Figure 30: 2.3.5B R50
On the other hand a plane surface has less potential for
aberration control. The key for explaining this apparent contradiction (simpler
design and higher image quality) is the improved optimization capability of
modern optical design programs. The Tele- Elmarit 1:2.8/90mm is
a very interesting example. This lens has been introduced in 1973, as a
parallel development of the Elmar-C 1:4/90mm for the Leica CL. It shared no
components with the 1964-version for the Leicaflex. In 1980 a completely different
and improved 90mm for the R-system appeared, but this lens was not adapted for
the M-system until 1990. There is an interesting story behind the Tele- Elmarit
for M. The 135mm lens was not selling well and disregarding its outstanding imagery,
was to be axed from the catalogues. But the focal length of 135mm had to be
taken care of. In the Leitz archives we find the key for the strange role of
this lens. The Leitz designers had in mind a teleconverter 1.4x
for the M-system and the compact Tele-Elmarit. This combination would deliver a
very small and lightweight 135mm lens of acceptable aperture (1:4.0). This idea
however has been shelved, because of lacklustre performance. It does indicate
that creative solutions were actively searched for.
Figure 31: 2.3.5.C Novoflex
A third example is the Novoflex system, which used the
optical cell of the Telyt lenses, made for Visoflex and R-bodies too.
2.3.6 Cooperation
Here we can identify one of several strategies that Leitz
choose to use: cooperation with several lens manufacturers. In hindsight it is
easy to comment upon the decision of Leitz to neglect the zoom lenses for the
R-system. The official position of Leitz regarding this type of lenses was the
statement that the optical quality of Leitz lenses with fixed focal length
could not be attained by the zoom-lens. I would regard this as a marketing act.
Technically Leitz were not equipped to design zoom-lenses that lived up to
their standard of optical quality at those days. On a more philosophical level,
I would venture to remark that the zoom lens with its physical bulk, its many lens
elements and complicated engineering and indeed its lower quality, did not fit
in with the Leitz philosophy of lens design. Leitz tried to use the least
possible lens elements to get optimum results by studying the character of a
lens system in depth first. The basics of a complex zoom lens are very
difficult to study as there are so many variables involved. A zoom lens can be
designed with optical compensation or a mechanical compensation as a tool for
correcting aberrations. When the first zoom lenses arrived on the market, both
methods were used, but none of them gave the results Leitz was looking for. The
optical compensation did not give exciting results, but the mechanical one was
an engineering nightmare, at least in the Leitz tradition of mechanical
precision.
Figure 32: 2.3.6
A Angenieux Whatever Leitz' view on
the zoom lens, the market demanded a range of zoom lenses and so Leitz had to
offer them in order to broaden the appeal of the R-system. The first supplier
of zoom lenses was the French company Angenieux.
This company offered three versions:1:2.8/45-90mm,
1:2.5/35-70mm and 1:3.5/70- 210mm. The cooperation was quite limited. Leitz
allowed Angenieux to use the bayonet-mount of the Leicaflex, and promoted the
sales of these lenses. The relationship with Minolta was much tighter. There
was in fact a technology transfer deal, and we may remark that in optical
matters Minolta learned more from Leitz than the other way around. The Japanese
contacts however established a base on which Leitz and later Leica could expand
their contacts with Japanese suppliers. The Minolta zoom lenses were delivered
as Leitz lenses and they were 'adopted' as Leitz products. The optical
performance of these lenses was for many intents and purposes good enough for a
broad range of users. Branded as Leica lenses, they did not perform with the
characteristic Leica fingerprint. And critical users might question the choice
of the Leica system if the backbone of that system consisted of third party
lenses.
Figure 33: 2.3.6 B Minolta
Figure 34: 2.3.6.C Schneider
Figure 35: 2.3.6.D Zeiss
Leitz clearly lacked knowledge and possibly resources to
expand the Leicaflex and Leica R system with only Leitz designed lenses. The
Mirror lens, the Fisheye lens and the 15 to 24 wide
angle lenses came from various sources (Zeiss, Schneider and Minolta). The same
strategy was followed with the zoom lens. Angenieux, Minolta and were the
suppliers. There is some discussion if the third-party lenses were just rebranded
designs or if Leitz added some value to these lenses. The Angenieux 1:2.8/45-90mm
was the first in 1969, in 1974 followed by the 1:4.5/80-200 from Minolta. The
next was the 1:4.5/75-200, again by Minolta and the redesign in 1984 while the
4/70-210 had additional Leitz input? The 3.5/35-70 had two versions, one in 1982
and the second in 1988. (both Minolta design). (The 3.5-4.5/28-70mm in 1990 was a Sigma product.) We have
to wait till the next period to see what Leica designers can do with a zoom
lens after they had acquainted themselves with the technicalities and optical
properties. The very wide angle lenses for both M and R were supplied by
Schneider (21mm) and Leitz also used Schneider designs for the wide angle
lenses with perspective control. Leitz did not design their own zoom lenses,
and at first used Angenieux designs. This strategy of cooperation and adoption
of lenses from third party suppliers continued and expanded during the seventies
and early eighties with several Minolta designs, (fisheye lens, wide angle lenses,
mirror lens and zoom lenses) and a Zeiss super wide angle (15mm).
2.3.7 ELC and ELW
In 1949 Max Berek died and Dr. Zimmermann took over. Some
years later a second optical design department was established in Canada under
the direction of Dr.
Walter Mandler. The fifties presented several challenges to
the Wetzlar company.
The introduction of faster means of computation (desk
calculator and electronic computer) allowed the exploration of better designs
and as a result deepened the theoretical understanding of aberration theory.
This research could be transformed into commercially viable designs if new
glass-types were available and if the small engineering tolerances that were
required could be held in manufacture and assembly. And last but not least, the
demands of the Leica users for ever better imagery
intensified the exploration into the realm of image assessment. The arguments
for the establishment of the Midland branch have been the bleak cold war prospects
in Europe and the prominence of the American market for Leitz. The choice of
Midland as a town has been a decision of Leitz himself as he had some private
relations with the town. Midland was set up as a design and production facility
and many lenses have left the factory since the first days. The first lens to
be manufactured was the Summarit in early 1952 with serial number 987101. After
about 1955 some fundamental changes occurred. One of the most far-reaching is
the gradual shift of balance of camera lens design to the Canadian factory.
Many of the lenses for the M-system and a fair proportion of the lenses for the
R-system originated from ELC (Ernst Leitz Canada). Mr. Mandler, who already
worked in the Wetzlar factory when Max Berek headed the department, went to
Midland, Canada in the mid fifties to organize and manage the optical department
there. Both Wetzlar and Midland used electronic calculators and later computers
for faster and more elaborate computations. At first the speed of the computer
was employed as a convenient replacement of the older logarithmic tables and to
speed up ray tracing.
Later algorithms were programmed to compute and analyze the
aberrations themselves and in a third stage additional computer programs were
written to optimize (automatically correct) optical designs. In Wetzlar the
foundation of aberration theory and its corrections originated from the design
and construction of microscopes and its lenses. The specific concerns and
characteristics of 35mm photography were added to this body of knowledge and
experience in proprietary programs. In the mid seventies prof. H. Marx almost
single-handedly designed the computer programs that are still used by Solms
designers, all be it in much improved format. The COMO program is the heart of
the computer programs in use now in Solms. Correction, Optimization, Minimization,
Orthogonalisation are the keywords in the name.
Figure 36: 2.3.7: Marx pictures on CD!!!
Research into exotic designs continued in both departments.
Looking at the list of prototypes, it seems as if the Wetzlar people were
deeper into the fundamental research. In Wetzlar we find lenses with three
aspherical surfaces, and specially designed glass composites. The prototype
1.2/50mm lens with aspherics and segmented glass is a fine example of this
drive to the limits of design and manufacturing technology.
Figure 37: 2.3.7.A lens picture
Figure 38: 2.3.7.B drawing
Meanwhile in Canada, the designers adopted the optimization
programs at an earlier stage and so were often able to design lenses faster
than their Wetzlar counterparts.
ELC designed lenses were also optimized for economical
manufacture. More so than the Wetzlar designs. In a
period that Leitz became more and more vulnerable to the prevailing market
forces, this was an important consideration. Recall that around 1970 the
M-system was far beyond its zenith and the R-system faced stiff competition.
The M5 and CL bodies needed new designs (resp. more retro focus lenses and more
compact ones) and the Leitz philosophy to stick to high quality fixed focal
lengths and thus neglecting the trend of the zoom lenses, put a heavy load on
the design teams at both sides of the Atlantic. Commercial considerations however
forced Leitz to include zoom-lenses in the catalogue and just as a decade earlier
with the Schneider cooperation, Leitz added some Minolta zoom-lenses to fill the
gaps.
2.3.8 Glass research.
The glass lab existed from 1948 to 1990. The Leitz glass lab
is the focus of many stories. The glass lab, by its very dimensions was not
equipped to handle or melt large blocks of glass in any quantity. The true role
of the glass lab was the research into the composition of new glass types, with
properties that could be used by the optical designers to help them with the
computation of new lenses. This role was an important one and many of the glass
types now in the catalogues of large manufacturers like Schott or Corning have
been first explored in the Leitz glass lab.
One may say that without the glass lab the Leitz designers
would not have been able to acquire the tools and the knowledge to improve the
Leitz lenses. The original building is still to be seen on the old Leitz
Wetzlar Industry Park (Werksgelände), tucked away between the imposing
main ten-storey buildings. It has an area of about 10x20 meters and these
dimensions indicate its role. There were four small melting pots,
that could handle 5 to 10 kilograms of glass at a time. The dilemma for
the optical designer is that he needs special glass to optimize his designs,
but only in small quantities. The direction of research for Leitz was in the
exploration of high refractive glass with very good colour transmission. That
proved to be a problem and only when one started to use lanthanum oxide and
(radio-active) thorium oxide, both characteristics could be combined. The use
of rare earth additives (lanthanum) is an invention of Kodak in 1938 and later
widely used by other manufacturers.
Figure 39: 2.3.8 A glass map of Schott two versions!
Figure 40: 2.3.8.B Leitz additions
The glass lab did introduce many new formulations of special
glasses that could be more easily machined. Often glass with exotic
specifications can not be handled in the manufacturing process, due to specific
problems: glass too brittle, glass starts to discolour, etc. As the lab had no
production capacity, they had to make an arrangement with a glass manufacturer,
like Schott or Sofirel or Corning who would then produce the glass under
several types of agreement. The true significance of the glass lab is for the
optical development that it helped to support and even foster and for the microscopy
department where one needs glass of more demanding specifications and
tolerances than is needed for photography. Only very few
glass has been melted there for production purposes. The well-known
Noctilux glass (900403) has been produced for some time in the glass lab, as
Schott could not comply with the specifications. Later the glass has been
replaced by one that could be obtained from regular sources.
2.3.9 Design and manufacture of Leica lenses
The design department needed about two years and sometimes
longer to complete a new design. Often the designers had to address other
topics and could not work continuously at the project. Sometimes optical
problems proved to be difficult to tackle. And sometimes one had to make a
fresh start At least a year was needed for testing and adjusting the production
cycle to this design and then another period was required to stabilize the
production to the manufacturing quality that was expected.
Figure 41: 2.3.9.A Centering of lenses.
The manufacture of the lenses was separated, physically and
as organizations, from the design department. The optical designers were
primarily focused on the goal of optimum image quality and left it to the
production engineers to find a way to hold the necessary tolerances for machining
and assembling the mechanical and optical parts to the required specification.
It is clear that some professional tension existed between both departments.
The optical people might specify a certain glass, because it had the required
optical properties. During the processing stage, one could discover that the
glass would discolour a bit, and the transmission properties were altered. In
such a case, an alternative glass had to be searched for. Or the rim of the glass
was too thin and would give problems when fitting the lens in its mount. For a classical
Double-Gauss design, the optical department would specify the thickness tolerance
of the cemented doublet (the middle part) as 0.1mm and the sum of both doublets
as 0.02mm. Such a matching scheme is only possible as lenses are manufactured
with tolerances on both sides and this matching may be labourious and error
prone. The mechanical department then would ask for a different set of tolerances.
Figure 42: 2.3.9.B Checking mounts
Thicknesses of lens elements and spacing are very critical
and the natural tolerances within a production process need to be taken into
consideration. A thickness tolerance of 0.1mm is acceptable for larger scale
grinding of lenses, but to keep a tight tolerance of 0.01mm in even a small
production run, is a hell of a job and a very expensive
one too. The optical designer has the goal to find the best optical performance,
but also to create an optical system that can be manufactured economically. The
division of labour asked for an elaborate interaction between the departments.
After the lens is computed on paper a few prototypes are manually assembled,
which includes the grinding of lenses. When prototypes are
assembled, elaborate testing in the laboratories (to check the residual
aberration, but also to test the quality of manufacture) and in practical photo
shootings (to check the visible image quality) is required. If the
visible image quality is not as expected (compared to predecessors and other
lenses) the defects are noted and the optical department has to change some of
its parameters. Or the lens may be optically on target, but it is too difficult
to produce. There are no clear-cut solutions in this interplay of mathematical and
engineering parameters. Sometimes an optical redesign would take a long time to
accomplish and might not deliver the required improvement. In such a case it
would be cheaper to make educated guesses and make new prototypes on the basis
of these proposed changes.
Figure 43: 2.3.9.C Interferometric testing of surfaces. Interferometric
checks
To understand the magnitude of this task you may reflect on
the fact that the optical designers are looking at the sub-micron level when
they discuss tolerances and changes (much less than 1/1000 of a mm). The mechanical engineer however will consider
tolerances of 1/100 of a mm as quite exacting. This
'engineering' gap of a factor of 10 to 20 in requirements, presented a
challenge to the Leica engineers working in the optical and mechanical
departments. A challenge by the way that still exists today.
As Leitz took very great care that the lens at the end of the assembly line was
indeed within the specified tolerances, the process of cost cutting had its 'natural'
limits. Lens assembly at Leitz was and is a labour intensive process,
that only a qualified and experienced workforce can handle. Any new lens
design requires that the workforce has to be retrained to produce and assemble
the lens. A designer can easily create an exciting design that is impossible or
very difficult to construct. Lens elements are polished on a single block and
the more elements can be polished in a single run, the cheaper it will be. But
then the curvature of the elements should be as flat as possible. That limits
the designer's freedom. Any lens element or steel or brass or aluminium
component has some deviation from the zero-tolerance norm, however careful the
production process. A clever designer has to allow for some tolerance and make
sure that image quality does not drop beyond a certain specified level, as some
manufacturing deviations are to be expected. This is part of the optical design
itself. But many components will have different values within the tolerance bandwidth.
Throwing away is no option, so the technique of matching and compensation is
required. This matching and checking and correcting takes time and the design
should allow for this technique. A common method at Leitz was to assemble lens
elements in sub-units that could be individually and manually adjusted to the
required specifications. Leica lenses are solidly build
with finely machined components. These characteristics are the result of the
design and production methods used then and exemplify the level of quality
Leitz did build into any lens.
There is a persistent view that the Leica lenses till the
sixties are better built than later series, with a 'cost is inconsequential ' consideration. That would be economical suicide of course.
But is there a grain of truth is this view. As with many Leica topics, a
clear-cut answer is not easy. Brass and aluminium and a layer of chrome give an
impressive appearance and feeling. Part of the truth is this: older designs
were assembled form subassemblies, that housed some of
the lens groups. This was done to be able to adjust the lens elements and lens
groups to the required tolerances as specified by the optical department.
Manufacturing technology of mechanical parts, in those days, could not hold the
very fine tolerances needed. The same applies to the individual lens elements.
The assembly workers used parts that were classified into tolerance groups,
finely graded in plus and minus tolerances so as to combine a suitable pair to
match the numerical demands. As so many lens elements have to on stock to help
the matching process, costs will be higher too. It is true that Leitz built their
lenses with high quality materials and with great precision, but the manufacturing
cost of the lens governed the production process.
2.3.10 Summary for the second period: 1957 to 1988.
The level of performance of the lenses for the R and M
systems had improved over the years to an enviable level and it is quite easy
to understand that any further improvement in image quality required more and
more study in the optical department. It is one task to design a new or
improved version with a lower level of aberrations as calculated. It is quite
another task to transfer this improvement into production, given the much
tighter tolerances and it is a third task to make these optical improvements
clearly visible for the user in practical shooting assignments. It is the same
story as with an athlete. To cut the time for the 100 meters from 13 to 11 seconds
takes a short training period.
To go from there to the sound-barrier of 10 seconds takes
years of intense training.
So it is with a lens. To control the classical third order
aberrations is relatively easy, once you have understood the principles. To get
a grip on the next level of aberrations is much more complicated. When we take
a closer look at the Leica lens history we note a slightly different pattern of
evolution for the M-system. Here we see four trends: the change to retro focus
designs in the wide angle group of lenses, the simplification of optical
systems for many lenses in the 35 to 90mm category with some improvements, a
markedly higher performance in the 135mm group and a special effort in the 50mm
group of lenses to advance the higher speed lenses to a class of its own. For
the R-system we see some similar and some specific trends.
Again we see a simplification of designs with the additional
aim to provide enhanced performance. Then we note the adoption of third party
lenses to close gaps in the proprietary lens-line to provide a comprehensive
system of lenses and a specific strategy to enhance performance of the longer
focus lenses markedly by apochromatic correction. The complete strategic
neglect of the zoom-lenses may seem strange at first, but can be explained if
not excused by the concentration on the core strength of the company: fixed focal
length systems. Still Leitz had a special team of people that studied the
theory and possibilities of the zoom lens. It seems as if Leitz were reluctant
to explore new worlds. The invention of the Correfot is another example of a
Leitz innovation, that was not developed further. When
the M5/CL pair failed in the marketplace and the R3 succeeded the Leicaflex
line, the spirit of the Leitz company faltered. With
the M5 and the Leicaflex SL, Leitz tried to redefine the classic rangefinder
concept and define the SLR-camera within Leitz parameters and in both cases had
to admit that the Leitz approach failed to capture the market. Designs like the
Noctilux 1:1/50mm and the Apo-Telyt -R 1:3.4/180mm were
true to the optical mission of the company. The exploration of the ultra high speed
lenses for M and R too (a 1.2/50 design for R has been contemplated) and the design
of the best possible image quality by using special glasses and apochromatic corrections
continued. Quietly however, a new level of possibilities for optical correction
gave rise to designs of less spectacular specifications. The Elmarit-R 1:2.8/90mm
(1983) and the Apo-Macro-Elmarit-R 1:2.8/100mm (1987) were of much humbler
specifications, but they were signposts of the direction to follow. The Apo-Macro-Elmarit-R
1:2.8/100mm is the last lens that, supervised by Mr. Vollrath, left the Wetzlar
design department. It became the standard by which all other Leica lenses were
measured. At full aperture it delivers outstanding image quality over the whole
image field with very high overall contrast. The outlines of larger subject details
are recorded with almost 100% contrast transfer. And finer detail still is delineated
with a rarely seen 95% contrast transfer. This high edge contrast is supplemented
by a very crisp recording of the finest possible textural details that are rendered
with very good clarity The last lens to leave the
Wetzlar factory was a Summicron-M 1:2/50mm with serial number 3451920 at the
end of 1987. The new owners separated the camera manufacture and optical design
department from the Wetzlar microscope and instruments division and set up a
new and independent factory in an existing building at the outskirts of Solms,
about 7 kilometres from Wetzlar.
Part 4: the new generation from 1990
From 1980 to 1990, no new lenses for the M-system have been
produced. If we look closely at the M-lenses, we note that the last lenses
specifically designed for this system have been the Noctilux 1:1/50mm from 1975
and the Elmarit-M 1:2.8/21mm from 1980. The 2/50mm, the 2.8/90mm, the 2/90mm,
the 2.8/135mm and the 1.4/75 were all developed in parallel to the R-system.
The Summilux 1.4/75mm is the last lens that left the drawing boards of ELC. ELC
became a manufacturing division and a design department for non-photographic
lenses.
During this decade the introductions of new lenses for the
R-system were quite few too. The 1.4/35mm (1984), the 2.8/280mm, the 2.8/100mm,
(1987) and the 2.8/19mm were designed during this decade. The centre of optical
excellence shifted back to Leitz Wetzlar, but the much reduced resources for
R&D and the precarious state of the photographic department discouraged the
pursuit of new optical designs and techniques of manufacture. In hindsight one
has to make some critical remarks about the decision to split the optical
design department over two continents. The official argument by Leitz for this
division of labour is the threat of the Cold War and the possibility of a
nuclear clash between the super powers. Leitz wanted to preserve the optical
expertise of the factory and thought it wise to duplicate the design
department. A more practical argument might have been the growing demand for
military optical equipment by the American army. The Canadian optical department
developed, only partially influenced by the Wetzlar department their own design
philosophy, and used different tools to compute the lenses. The fact that many
lenses were designed in Midland and manufactured in Wetzlar,
certainly induced some tension into the production cycle. It was however
accepted practice to split the design and production departments into two
groups with different and separate responsibilities. The dominant position of
ELC during the period 1955 to 1975 gave the Leica lenses a specific flavour and
character. It also helped the Wetzlar team to pursue fundamental research into
optical design and aberration theory. This work laid the foundations for the
next generation. No doubt the practical dominance of the ELC designs,
frustrated the Wetzlar engineers and designers. In practical testing the ELW
designs were often better than the ELC counterparts. Still the Midland
proposals were selected as the production versions as they could be produced
more quickly or efficiently. And that was an important consideration in those
days. The theoretical studies in the Wetzlar department became, in a certain way,
a goal in itself and the famous optical lab was the result. It did not live up
to its promises however and it was finally closed. The head of the lab, Mr
Thomas, noted in LFI (7/1982) that the role of the lab had been reduced to the
testing of lenses of the competition.
2.4.1 New visions
Lothar Kölsch became the new Head of the Optical
department at Leica, Solms. The first lens to be produced by the new team, working
in the Solms factory, was the Summilux-M 1:1.4/35mm aspherical. The patent is
filed in 1991, and its inventor is identified as Walter Watz, one of several
unknown masters of optical design. As an important aside I would like to draw
attention to the many anonymous individuals, who worked in Wetzlar and who now
work in Solms, outside of publicity. Their tireless research into the
foundations of aberrations and their creativity in transforming that knowledge
into an optical system, consisting of glass and metal, that will satisfy the
demands of exceedingly critical users, has to be mentioned and acknowledged.
Figure 44: 2.4.1.A Summilux aspherical
This first Summilux with two aspherical lens surfaces is
evidently the direct descendent of the original Noctilux 1:1.2/50mm from 1966.
During the design stage of the 1.4/35mm lens, it became clear that the use of
aspherics alone would not bring the desired improvements. A high speed lens
with a focal length of 35mm is more difficult to correct than a 50mm lens,
because of the increased influence of zonal aberrations. One can identify these
zonal errors when one takes a look at the performance of the original 1.4/35mm
lens for the M-system. The new design performs very well already at the wider
apertures, with a high contrast image at full aperture and a clear, crisp
definition of finer detail in the zonal areas. The revolution of the aspherical
design is not so much the use of the aspherical surfaces, but the radical
departure from the classical and time honored Double-Gauss principle. The system
comprises nine lens elements in five groups, with the first surface of the
first element and the last surface of the last element of a concave shape. The
design is more symmetrical too, which helps reduce some aberrations. The path
of the light rays, passing through this system, is more 'relaxed' than in a
traditional Double- Gauss system, a characteristic, that is not obvious and of
a more philosophical nature. It does signify a new aspect of lens design at
Leica. That is the study of and reflection on the fundamentals of a design. The
current approach to lens design in Solms is characterized by several
interlocking aspects. The most important trait is the search for the most simple solution for the required lens parameters, as
maximum aperture, focal length and physical size. To find such a solution, the
designer needs to study the intrinsic behavior and possibilities of the optical
system. Supported by quite sophisticated and indigenous computer programs, the
designer will then search for an optimum solution, that
can be manufactured within the required tolerances.
New insights into the character and potential of the glass
types available on the market bring additional advantages for aberration
correction. If I had to single out the most important characteristic in the
current design approach, it would be the tight cooperation between the optical
designers and mechanical engineers. The second factor which helps explain the
performance and the cost of a new Leica design is the principle of manufacture,
where the fabrication process is adapted to the requirements of the selected
glass and of the tolerances needed. This is a large step forward when compared
to the previous method of manufacture that was the same, irrespective of the
specifics of lens design and glass manipulations.
Figure 45: 2.4.1.B grinding the aspherical
The two aspheric surfaces were grinded and polished
mechanically on a spherical glass surface. This method was an improvement on
the original manually grinded aspherics, but still quite labourious, with a
high rejection rate. The second version, recomputed in 1994, employed a new
press-molding technique, in combination with Hoya glass. Leica has been heavily
involved in the invention and development of this technique.
Figure 46: 2.4.1.C 2/180
From 1993 to 1996 the designers gave all attention to the
renewal of a number of lenses for the R-system. A new series of high speed
telephoto lenses with exceptional apochromatic correction and high speed were
developed. The Apo-Telyt-R 1:4/280mm is probably the best corrected lens in the
whole R-stable. This one is indeed diffraction limited. The Apo-Summicron-R
1:2/180mm (1994) and the Apo- Elmarit-R 1:2.8/180mm (1998) offer performance
that is truly a quantum leap ahead of the predecessors and do indicate the
mastery of the optical designers. These lenses also embody new techniques, like
internal focusing that is in addition employed for further aberration
correction and macro-functions. New mounting techniques deliver a very smooth,
fast and accurate focusing operation.
Figure 47: 2.4.1.D 2.8/180 apo
In the field of zoom lenses, Leica has made impressive
progress and reduced their dependence on third party suppliers. And one might
add, with a vengeance as the new variableness, including the first one, the
Vario-Apo-Elmarit-R 1:2.8/70-180mm, equal and often surpass the image quality
of the equivalent fixed focal lengths. In quick succession the family is
expanded with a 4/80-200 and a 4.2/105-280mm version. It does show that as soon
as the designer has studied and understood the basics of a design, it is
relatively easy to develop several versions. This same technique of spin-off we
see in the Summicron-M 2/35mm ASPH, which is a variant of the 1.4/35mm ASPH.
The Elmarit-M 1:2.8/21mm ASPH and 1:2.8/24mm ASPH share family characteristics
and especially the 24mm lens delivers outstanding imagery and is a landmark
design.
Figure 48: 2.4.1.E Vario 70-180
Figure 49: 2.4.1.F vario 80-200
The first combination of apochromatic correction, which is
primarily a glass selection technique and the use of aspherical surfaces which
can be used for several purposes (see chapter 1.2) we find in the
Apo-Summicron-M 1:2/90mm ASPH. This lens is arguably one of the very best
lenses Leica has ever designed, as of this moment of writing. Already at full
aperture it delivers outstanding quality that even surpasses the performance of
the Summicron-M 1:2/50mm at 1:2. In the same league we can place the
Apo-Telyt-M 1:3.4/135mm, a compact and light weight telephoto lens, that approaches the old dream of a lens that does not
improve on stopping down. At apertures from 1:5.6, the Apo-Telyt is only
slightly better than the previous Tele-Elmar-M 1:4/135mm (thanks to the
apochromatic correction, which brings a small but visible improvement, even at
smaller apertures). For most purposes these differences are not that important.
The exceptional performance of Leica lenses is best appreciated at the wider
apertures. To deliver this quality, the manufacturing tolerances, the
production technique and the quality checks must all match. The lens that in my
view exemplifies this marriage of mechanical and optical expertise and
production technology is the new Tri-Elmar-M 1:4/28-50mm. Its image quality is
better than most Leica-M lenses of comparable fixed focal length.
But its mechanical layout is the true measure of excellence
in this case. This lens can be used on all Leica M bodies, even the earliest
ones, like the M3. The rotating ring that selects the different focal lengths
has to accommodate different spring tensions and cam curvatures. The solution
is mechanically very complex, but it brings zoom convenience in a rangefinder
model and so defies obsolescence. The engineers however have designed an even
better solution, which will be introduced at Photokina 2000 and sports improved
ergonomics and redesigned internal mechanics.
Figure 50: 2.4.1.G Tri-Elmar
Many observers of the Leica world would select the two
Noctilux versions as the defining icons of the optical excellence in the
rangefinder domain. I would beg to differ. The Noctilux lenses are optically
quite good and do radiate exquisite mechanical engineering and optical
expertise. The Tri-Elmar, Elmarit 24mm and Apo-Summicron 90 and Telyt 135mm
point to an exciting future direction and are more versatile and effective in
everyday use. The position of the R-system is not so easy to sketch. The
current Solms-designed vario-lenses are outstanding products, but physically at
the limit. Now Leica will tell you that this volume is needed to deliver the
optical quality and I do not question this statement. Still the challenge is twofold:
to design fixed focal lengths with even better performance than the current vario-lenses
is not easy and the efforts and cost must be offset by a sizeable sales volume.
The demands for designs of variableness are threefold: a wider zoom range; ergonomically
convenient to use in handheld situations; small physical dimensions and
comparable, if not improved performance versus the fixed focal length lenses.
These requirements are very difficult to combine in one set.
The current range of Rlenses is composed of three groups. In the first group,
the vario lenses, we find new lenses, designed by Leica itself and some lenses
supplied by third party manufacturers. In the second group, the fixed focal
length, we have a number of recent developments, like the 2.8/28mm and the
1.4/50mm, both of outstanding quality and a family of telephoto-lenses in the
180mm to 280mm class of really superior performance. A number of the older
fixed focal lengths lenses, however, begins to show
their age, like the 35mm lenses, the 60mm and the 80mm. As with the vario
group, , there are a number of lenses from third parties, like the 15, 16 and 24mm
and the PC 28mm. The third group consists of the module system from 280mm to
800mm, which is specifically designed for economy of cost and flexibility of
use. This system has added value if the user needs more than one focal length
and so the target group of users is limited.
2.4.2 Current status of optical design and lens manufacture.
The improvements in image quality in current Leica lenses
are visible for anyone with an experienced eye. Some Leica users might defend
the idea that the older lenses are as good as, or even better than the current
ones. Image characteristics, like 3-dimensional representation, plasticity and
smooth unsharpness gradients, cited as being instrumental for the unique
fingerprint of older Leica lenses, are supposed to have disappeared. Some observers
have tried to discern two different schools of design theory, the German and
the Japanese philosophy, the latter putting all emphasis on image contrast and
the former school going for a smoother image quality. In all fairness, these
two schools do not and have never existed in this extreme juxtaposition. Of
course designers differ in their balancing of aberrations and the rigor of
aberration corrections. It may be the case that Japanese designers follow a
different route to design a lens. But generally any designer everywhere has the
same type of tools and uses the same aberration theory. Why and how are the current Leica lenses improvements on the older
designs and how do they differ from others? Modern optical design is a creative
activity, based on experience, insight, a very
thorough knowledge of optical theory and even a sprinkling of luck. We know that
optical systems for photographic purposes are not completely aberration free.
(There are exceptions. Some R-lenses, and many microscope
and telescope lenses, are indeed diffracted limited (which amounts to stating
they are aberration free).
Most photographic lenses will exhibit, after correction and
optimization, some residual aberrations. These residuals have to be carefully
controlled and balanced.
The optimization techniques will help of course, but no
computer program can improve on a design that is not promising from the start.
Leica lens designs will start a design with the absolute minimum of lens
elements. Having less elements one can study the aberration
content and the contribution of every lens element to the aberrations more
carefully and more effectively. Current computer programs are very powerful
and can make the best of any design by a kind of brute force method. By using
more and more lens elements to counter the aberration content, a lens can be made
to perform quite satisfactorily. It will not show the finesse and elegance of design
of the Leica lenses and it most certainly will not perform on the same level.
2.4.2.1 Progress in design methodology.
The art of designing optical systems is a most fascinating
subject. The lens designer today still uses methods, concepts and terminology
that have been developed in the previous century. The employment of high speed
computers and very sophisticated software has freed the designer from the
drudgery of the past, when slogging through labourious computations to trace
the light rays, using tables of logarithms or mathematical approximations was
the only way to design a lens. The current practice of optical design then is a
fine example of a successful amalgamation of a century-old (time honored)
science and a very potent modern technology (the high-speed computer). The
genius of Gauss, Clark, Rudolph, Lee, Berek and others should be acknowledged
with great admiration, as they combined quite crude computational methods with
a most profound insight into the fundamentals of aberration theory.
The computer, as we all know, is a very sophisticated device
for number crunching with phenomenal computational speed. It took Rudolph and
Lee and later Berek at least five years of dedicated attention to trace enough
rays with the required precision to get a good grasp on the aberrations in the
optical system. Nowadays a potent personal computer can trace 600.000 skew ray
surfaces per second. In 's time this amount of skew
rays would take a big part of the lifetime of one designer. Before the advent
of the computer a lens would be designed and evaluated by tracing real (trigonometrical)
geometrical rays from lens surface to lens surface. In the past, every calculation
of a surface took a few minutes. You need to trace a few hundred rays through
all surfaces, often more than 10 or 12 surfaces and do that again and again after
you find that the image points you calculated so diligently are still distorted
by aberrations. All the points calculated with this method are meridional rays, that is rays that lay in the same plane
(are coplanar with) as the optical axis. When we trace a ray on a flat piece of
paper, as we all do, the paper plane is the meridional plane. This a two-dimensional plane. But we also have skew rays, that is rays that go into the lens from an
oblique angle. Now we need three-dimensional ray tracing, which is almost impossible
to do by hand or even pocket calculator. So in the past these rays were not
traced at all, or all kinds of 'tricks' or approximations were used to study
the skew rays. Many of the troublesome aberrations (coma, astigmatism as
examples) are caused by these skew rays. So in the not so distant past ,designers had to accept they could not all the
information they needed to fully correct a lens. So working with imperfect
knowledge, the designs could not be as fully corrected as was theoretically feasible.
If we look at the older designs with modern expectations, their image quality may
be a bit disappointing. The designers of those days (1930 till 1960)
compensated their approximation strategy with admirable creativity and let us
call it gut feeling to select basic designs that work and serve the
photographer well. Leitz was one of the first to implement a computer in order
to support the designers, The Zuse Z5 was introduced
in 1953 and helped the designers to compute the oblique pencils of rays.
Before the introduction of the computer Leitz designers used
the Seidel-Theory to circumvent part of the complex computations of oblique
rays. This knowledge and experience proved to be of great value after the Z5
became the preferred tool for ray-tracing. It is fascinating to know how this work
was organized at Leitz. The romantic idea of one designer working for years on
his design is of course a myth. In reality, the chief designer supervised a
group of workers, mostly women, who carried out a good part of the
calculations. The chief designer gave instructions and got the partial results
of every ray tracing equation from them at the end of the day and then decided
to proceed or adjust his design. The resulting design is invariably a compromise,
composed of both known and unknown factors. The development of lens design
methodology interacted with the development of aberration theory and both
allowed lenses to be designed with a much higher image quality. Now we have two
new areas to give close attention to. The first area is hardly ever mentioned,
but in fact it might be designated as the true revolution in lens design. The manufacturing
processes (grinding, polishing, coating, centering) and the mounting procedures
are of paramount importance for the image quality .
Studying fabrication tolerancing to ensure that the image quality as defined by
the designer, is a major part of the design process.
Here the art of optical design and a superior grasp of the manufacturing
process and the limits of tolerancing join to create the results we see and
expect (if not demand) from current Leica lenses. The Leica aspherical lenses may
only be serviced in the factory itself. An indication that
the mounting procedure is quite demanding. In this case we see clearly
the value and necessity of the manufacturing processes in boosting or
supporting the higher level of image quality possible.
2.4.2.2 Progress in manufacturing processing and mounting techniques.
I have remarked several times and documented in the lens
reports, that Leica lenses are designed for higher contrast at much higher
frequencies than in the past. A design that is can record higher frequencies
with high contrast needs very small tolerances. Contrast of very fine detail is
very sensitive to focus and alignment errors.
Figure 51: 2.4.2.2.X precision grinding
Decentring is very bad for the ultimate image quality. In
the past, when manufacturing tolerances were not so tight, groups of lens
elements were mounted in their own subassemblies. At the final mounting stage
these subassemblies could be adjusted a bit to compensate for decentring and
other errors. An optical surface has two related qualities: accuracy and
quality. Accuracy refers to the dimensional characteristics: is the surface
uniform in its radius value?. Quality refers to the
finish of the surface: is the polish incomplete (rough spots), does it have
pits?
Figure 52: 2.4.2.2.A Interference patterns check
Figure 53: 2.4.2.2.B Checks for centring etc
If the surface quality is not immaculate the following
process of thin film coating with its thinness of several microns will also be
irregular. Relatively speaking of course: we are dealing with
wavelength dimensions where a micron is a big distance.
Some requirements at Leica demand that a polished lens must
be coated within hours after the polishing as the surface can deteriorate when
not protected immediately. A Leica lens is now designed by a team of optical
and mechanical engineers who work together from the start. The production
engineer has the last word: if the tolerances are unrealistic, the optical
designer has to start all over. In the past the departments were more divided.
Then the optical designer designed the lens as he deemed necessary according to
its state of aberration correction. Then he almost literally would hand over the
design prescriptions to the production department. Sensitive glasses with
strong radius and thin ends that were very difficult to mount without stress
gave the designers of the mounts headaches. And sometimes the designed lens
could not be built within required tolerances. One had to accept a slightly
lower image quality or a production with a much higher rejection rate. At the beginning
of this section I mentioned that physical ray tracing is still the approach used
by designers to study and correct aberrations. Physically we should not look at
individual rays but at the total energy that flows through the lens at one
single instant in time. When we take a picture all energy reflected from the
subject enters the lens in one instance (1/250 sec as example) and the total
light energy flows through the optical system in one flash. Leica designers
take care that this flow is eased from lens element to lens element. Abrupt
changes in the path of a ray that occur when glasses of highly different
refractive indices or greatly varying curvatures are employed are avoided. One
might see a Zen approach here.
Figure 54: 2.4.2.2.C complex mount
Can modern Leica lenses be improved? Yes they can and Leica
is constantly searching for economical means to produce improved versions of
existing lenses and to expand research into exciting new designs. Just as with
the Olympics the levels of achievement are set higher every time, so these new
generation Leica lenses also set the quality level a few notches higher.
2.4.2.3 Procedures for testing a lens.
When a lens is designed, it is obvious that you calculate
with the exact figures, that is you assume a
zero-tolerance. After the ideal design, you have to allow for fabrication errors
and tolerances. It is also obvious that if you optimize the design to the
utmost level, any manufactured lens will be below tolerance and these errors
will degrade performance. So the assumption of fabrication errors must be part
of the design process. A small list of fabrication errors would comprise:
surface curvature errors, index and dispersion errors, thickness errors of the
lens elements or air spaces, axial displacement errors, surface tilt errors
(decentring). All of these errors occur and have to accounted
for in the design and manufacture. The nature of fabrication errors implies
that errors occur at both sides of the target figure. If a tolerance of 5% is
specified, then in reality the values will be + 5% around the target. Some of
these errors cancel out, and sometimes you have a lens that is above the target
value and sometimes below this norm. You may however not assume that the spread
around the target is normally distributed, with the same number of lenses above
and below he target. Leica specifies the target line
very high and so most of the lenses will be a bit below the norm, but well
within the tolerance band.
Figure 55: 2.4.2.3.A Testing MTF
There is however a second method to handle production
errors. That is the use of compensators in the lens. One could designate one
lens-element or group of lenses as a mechanical compensator and use the
possibility to shift this element a little or one can shift the whole lens s
bit, which is identical to a focus shift, which can improve the optical
performance significantly. (see the chapter 1.2).
Leica uses all these methods of testing a lens and adjusting the parameters.
Every lens that leaves the assembly department is tested. For lenses that have
a stable performance pattern and are relatively easy to assemble, the
projection test is used. A complicated test pattern (of Leica design) is
projected through the lens on a wall and studied at close distance to check for
errors. If the reproduction of the pattern is within specifications the lens is
passed. Some types of lenses are individually tested and adjusted with an MTF
device. Here the method of focus shift adjustment is employed. The optical cell
(that part of the lens that comprises the glass) has its own effective focal
length and the mount of the lens has a mechanical flange focal distance. The
lens is adjusted for best focal position (optimum focus plane) with the help of
very thin shims (0.01mm) that can be inserted between the bayonet flange and
the lens mount. The target value is the contrast value. These measurements are done
for 8 circular positions on the lens in steps of 45o. After the lens is
adjusted for optimal focus, a new series of contrast measurements is conducted
to check for decentring and other errors. This procedure ensures that every
lens meets the specifications, stipulated at the design stage.
Back to the table of content