The Rare Nippon Kogaku 8.5 cm F1.5 Nikkor-S.C for Contax, Nikon, and LEICA Screw Mount Rangefinder Mounts{

/BOLD}

Historically this is one of Nippon Kogaku's most exotic and extremely interesting lenses. At the time that it was released in January of 1951 it was the fastest telephoto lens that Nikon had ever produced for theirs and any other rangefinder camera. With a lens design that is a refinement of its extremely successful 5.0cm F1.4, its performance was amazing and is respectable even by modern standards though it is not at its very best at its widest open F1.5 aperture.

When the 8.5cm F1.5 was introduced in early 1951 at almost $300, one could purchase a very nice and well equipped car for less than $2000. Inflation adjustments would argue that this $300 price would be more like a now current price of $6000 but at a time when the standard of living was much lower and each inflation adjusted dollar meant a great deal more. No wonder that over its ten year lifetime, only approximately 200 of these beauties were produced per year.

A long focus version of and very similar in optical design to the 5cm F1.4 , it employs seven elements in three groups: a front singlet, and internal triplet and a rear triplet.

From the time that it was introduced to its discontinuation in 1961, only 2000 were produced. Most were in the Nikon Rangefinder mount. About 500 were in LTM or Leica SM. A very small number, probably less than a hundred of the very earliest ones were marked with a 'C' on the side of the lens barrel and were intended for Contax Rangefinder camera bodies. These would indeed be the rarest of the rare.

The first 1500 or so had a finer knurling pattern than the last five hundred. This was the only meaningful production variation in its design.

{BOLD}
CONTAX "C" MOUNT FOCUS PROBLEMS WORSE THAN NIPPON KOGAKU EVER CONFESSED!

The problems occur profoundly even for the the ubiquitous 5cm lenses for which they have never previously been identified.

SIMILARITIES AND DISSIMILARITIES of NIKON and CONTAX Mounts

In fact the Contax and Nikon BM mount lenses can be employed interchangeably on each others mounts and will focus perfectly, contrary to some erroneous rumors; but, only if the distance scale actually engraved on the lens is used against actual physical measurements of the distance from the film plane to the subject.* (Alternatively, a variable small angular offset, counterclockwise for a Contax based lens on a Nikon body and clockwise for a Nikon based lens on a Contax body, can be employed as compared to the built-in rangefinder.) It is the effective cam pitch vs. rangefinder coupling where the respective relationships fail. Both mounts are identical in distance to the film plane at infinity focus but the effective cam pitches are sufficiently different as to make a very significant difference away from the infinity setting, especially with an 8.5cm f1.5 wide open at portrait distances. Do note that this is exactly the failure mode that has been reported from the field as having been found many years ago. One class of very early reports holds that facial photographs at f1.5 with the lens focused on a model's eye suffered from the model's eye failing to fall within the depth of field by a few centimeters, very consistent with our findings.

There have been discussions over the years as to the genesis and physical manifestation of the differences between the two mounts. About a year ago Henry Scherer authored an article on the web SCHERER Article that purported to sort out the issues with some sense of finality. Unfortunately the paper suffers from the author seeming to know the results of his data prior to the gathering of the data itself. It seems that the author, apparently committed to a particular result, ignores the extremely obvious implications of much of his own carefully taken data. Due to copyright considerations (do please see Scherer's genuinely very humorous Legalese) and to vastly improve the photographic quality of the illustrative photography; we have redone the pictures: (Contax/Zeiss bezel is the one on top) Please click on images to enlarge.


The author claims that the screw pitches being identical argues that the rangefinder coupling lever is equivalent. However, the overlay of the scales showing approximately a one to two parts in 40 distance deviation at 3 to 4 feet would appear to prove the contrapostitve. The reader should note that a one to two parts in forty is actually essentially the difference in adjustment to the total "screw" that is required for an adjustment from the hypothetical difference cited by Scherer as what he cites as the "ERRONEOUS" source of the difference in base focal lengths between early Nikkor and Zeiss normals. Scherer states at the very beginning of the article: "The first thing to do is to ruin the myth that Nikon produced 50 mm marked lenses with a 51mm focal length and Zeiss produced Sonnars and Tessars with a 53 mm focal length." but his data would appear to emphatically support just that thesis that he has disclaimed. Do note that in the SCHERER Treatment the author determines and acknowledges that the the screw pitches of the two respective camera body focusing mounts are essentially identical but that the distance markings suffer a creeping offset with the Contax body displaying a longer displacement for an equivalent focus adjustment. (The reader should note that this signifies an absolutely compelling argument for the Zeiss/Contax scale being based on a longer focal length foundation lens than that for the Nikon scale.)

With regard to Scherer's observation that the Nikon Body is approximately 0.31mm slimmer than that of the Contax, suffice it to say that it is a measurement that we simply cannot confirm. Despite repeated attempts with a precision depth gauge to measure the distance from the lens position ring to the front surface of film at the film plane we see no such difference between the two representative camera body families.

It is difficult to pinpoint just where Scherer's error(s) originate because he reveals precious little that is of any use about his methodology. If one carefully studies the illustration in the article just above the statement: "The measurements showed that at infinity focus the Nikon S had a distance of 26.51 mm and the Contax was 26.82. The difference between these is 0.31 mm." One can see one very obvious error in his approach. By putting the Nikon S body on the DoALL Precision Granite Block as he has, it would seem that he has introduced an error on the order of the size he has claimed that he has found, but in the opposite direction. This manner of measurement would accrue to render a measurement of the thickness of the Nikon S body as thicker than it actually is. The block rests not at the film plane but at the plane defined by the outer film guide rails. These rails are slightly raised and do not serve any purpose in the camera body of defining any optical plane. Do please click on this small image to see how these guide rails rise above the actual film plane The sole purpose of these rails is to guide the film path from the film cartridge to the take up spool, nothing more.

Discussion Of Determination Of Methodology Of Focusing Engineering Of Early Contax And Nikon Rangefinder Bodies

Abstract: A number of lenses manufactured by both Zeiss and Nippon Kogaku (Nikkor) for Zeiss and Nikon mounts were each mounted successively on standard Nikon and Contax rangefinder bodies to measure the event of focus at very close distances and near infinity (in this case at approximately 200 feet.) All of the lenses are found unequivocally to focus exactly properly on the near infinity target regardless of which body they are respectively mounted on and deviation is only found to occur at close focus distances

We found that the basic correction to the distance setting for an 8.5cm lens (or actually any lens for that matter) on the other camera's mount is approximately:
1.2 inches at 3 feet
1.7 inches at 4 feet
2 inches at 5 feet
2.5 inches at 6 feet
4 inches at 9 feet
6 inches at 12 feet (which becomes a rapidly converging series at large distances as it is the distance differential divided by the distance to the object consistent with the proposition that the two are coincident at infinity and diverge as close objects are focused) The correction misleadingly appears to grow in absolute terms but is actually shrinking very rapidly when compared to depth of field or image dimensions of the focused image. The correct way to gauge it, is to view progression of the angular offset of the two scales.

The correction between the two cases corresponds to additional angular twist offset (given identical helical screw pitches) for the Zeiss calibrated lens versus the Nikon calibrated lens which is apparently true to the historical legacies of the Zeiss rangefinders being based on a 5.3cm standard while the Nikon rangefinders have a 5.1cm standard.

In use, the offset can be carefully compensated for, for distances of under 10 feet with a very small several degree angular twist, CClockwise for Contax calibrated lenses on Nikon bodies and Clockwise for Nikon calibrated lenses on Contax bodies, after focus is indicated by the camera body's coupled rangefinder.

An interesting aside is that the simpler and apparently more elegant approach that Leica took, of putting a custom slope coupling cam on each of its lenses, eliminates the very constraining consequences of designing optics that must conform to mechanics of the body's mechanical rangefinder. Leitz/Leica could, as it did, at will bring forth a 7.3cm f1.9 Hector. If some of the embodiments worked out to be more than a very small fraction of a millimeter off, precise focus could be achieved by a rangefinder cam pitch adjustment; a much simpler tweak than the thread screw pitch adjustment that would be required by the Contax body's system.

One basic lens formula in Physical Optics is :

1/(Focal Length) = {1/(Subj. Dist.)} + {1/(Image Dist.)}

For the case of pinholes or infinitely thin lenses in a vacuum the interpretation of this formula very straight forwards. In real life applications, however, the distances are the sums of the measurable distances multiplied by the optical density (often called index of refraction) of the different optical media.

which translates to :

1/(Focal Length) = {1/∑D[SUBJECT][Index of Refraction for each D]} + {1/∑D[IMAGE][Index of Refraction for each D]}

In the degenerate limiting case, as with a pinhole or infinitely thin lens in air (which approaches the optical density of a vacuum), the sum of the two optical distances is generally exactly equal to the actual physical distance from the film plane to the subject. Of course, in the real world where the optical densities (indexes of refraction) of typical optical glasses are in excess of 1.6 Index of Refraction Table ; Index of Refraction Table and through glass light paths approach a couple of inches, as with the 8.5cm f1.5, the correction for close focus situations, as with portraits, can be on the order of an inch, which for close focusing distances is very meaningful.

One concept that survives from the early era when these lenses were current is that the difference in focus mounts impacts "tele" lenses more than the "wides" because of the inherent greater depths of field of the "wides". It is true that "wides" have enough depth of field to cover a focus shift but the "wides", having a shorter focus base, can be more severely impacted in focus center shift by a distance to the film plane differential. To this end, we compare two two Nikkor 3.5cm f3.5's. One has serial number 61244 which is not marked "C" but which was clearly manufactured approximately in late 1946 prior to the advent of he original Nikon I and, hence, is a "C" mount lens as Nippon Kogaku was not likely a producer of lenses for which there was no body at the time. Incidentally, this is but one more test of Scherer's claim of a simple lens path distance shift as being the differentiating aspect of Nikon BM lenses versus Contax BM lenses. BTW, that the wides work well, or even at all, is but one more nail in the coffin that enshrouds Scherer's conclusions.

The measurements:

Two bodies were chosen: a Contax I and a Nikon SP. To define the focal planes Nikon F Type A focusing screens were carefully aligned on the film guide rails of each camera body: The lenses: a Contax BM 3.5cm f3.5 very early Nikkor (sn 61244), a Nikon BM 3.5cm f3.5 Nikkor, a Nikon BM 3.5cm f1.8 Nikkor; a Contax BM 5cm f2 Sonnar; a Nikon BM 5cm f1.4 Nikkor; a Contax BM 8.5cm f1.5 Nikkor; a Nikon BM 8.5cm f2 Nikkor; and a Nikon BM 10.5cm f2.5 .

For very distant measurements, a 200 foot distance was paced off and verified. For the near distance, 4 feet was chosen. All physical distance measurements were determined through air and from each camera body's film plane. Do note that this is not at all identical to the dimensions that are to be plugged into the equation(s) cited above but that they are the only practicable dimensions and that they are generally the ones that are engraved on the lens distance scale as well as the ones that are marked on the camera body's rotating distance triangulation base scale. This dimension is actually the sum of (Subject Distance) + (Image Distance) through air but which solves the optical formula equation with the appropriate optical densities plugged in.

Every lens was confirmed for proper focus at the distant 200 feet measurement on both bodies. The wide angle lenses were less conclusive because focus is a much less dramatic event but there was clearly no offset that would argue for a spacing or body thickness differential between the two body standards. The telephotos (8.5cm and 10.5cm), as expected, are more dramatic in focus events. All of the telephotos demonstrate crisp focus at 200ft. It makes no difference which body is employed or which mount standard the lens has. The mounts and bodies are interchangeable for focusing at a distance.

The 8.5cm f1.5 Contax BM Nikkor is found to focus properly at 4 feet on the Contax I body. On the Nikon SP, however, the lens had to be set 1 to 2 inches closer than 4 feet and what would obtain using the SP's built-in rangefinder. This corresponds to a small angular cclockwise correction to what the camera body's built-in rangefinder would dictate. This lens, of course, confirms close range focus when mounted on the Contax I body as do the Nikon BM short telephoto lenses when mounted on the SP body.

The 8.5cm f2 Nikon BM and the 10.5cm f2.5 Nikon BM lenses reveal an opposite focus offset when mounted on the Contax I body. Both lenses need a small clockwise angular correction to the focus indication of the camera body's rangefinder to compensate for the need on the part of the lens to be focused further out than the rangefinder would indicate.

CITED BELOW for comparison are Sherer's findings and conclusions followed by ours:

"...it can be seen that if a Sonnar which is set for good infinity focus on a Contax is used on a Nikon the focus will be off by about 25 feet at infinity, 10 feet at 100 feet and then declining down to about a foot off at close distances. This will be very noticeable."

The conclusion of this work is simply that a Zeiss Contax lens to be used on a Nikon camera must have the lens cartridge moved 0.31 mm further out in the focusing mount and a Nikor [Nikkor] lens to be used on a Contax must be moved about 0.31 mm further in. When this is done there will be a slight error in the distance indicated by the engravings on the lens bezel, but the rangefinder will be accurate and the lens will produce sharp images from infinity to close distances.

I am still trying to finger out just what "25 feet at infinity" is.

For a fact, it approaches infinity near infinity since a small part of infinity is still infinite as any serious student of math should probably know. Scherer apparently did not attend that course.

Our very carefully gathered measurements yield a pattern of results that is in no way similar.

IN VERY SHARP CONTRAST TO SHERER'S FINDINGS WE FIND THAT:

Every single lens, whether Contax BM or Nikon BM, without exception, shows confirmed focus at the film plane for the 200ft distance whether mounted on the Contax body or on the Nikon body.

For a Four foot distance: Each lens that is Contax BM focuses One to Two inches closer than is indicated by the Nikon built-in rangefinder. A variable small angle CClockwise correctional twist versus the built-in rangefinder is shown to be needed for proper focus. These lenses, of course, work perfectly properly on the Contax body. Each lens that is Nikon BM focuses One to Two inches further out than is indicated by the Contax built-in rangefinder. Of course, these lenses operate perfectly properly on the Nikon body.

It is also obvious from careful studies of the depth of field legend on the bodies compared to the offset that, though it has been relatively unreported, the disparity between the two triangulation focusing schemes is relevant to other lenses in each of the stables. Interestingly, both the 5cm f1.4 Nikkor and the 5cm f1.5 Sonnar experience focus offsets that well exceed their respective depths of field when focused "wide open" on the other's focusing system at close film plane to subject distances. Do please carefully study the illustrations below and the last (9th) illustration of the focus offset above. It is another unexpected consequence that both the 5cm f2 Nikkor and Sonnar will actually fall outside the edge of acceptable depth of field when mounted on the wrong body wide open at close range.
The reader should note that it is only the half depth of field that is the real buffer for focus certainty since focus is generally achieved at a center point and not an end point. Hence when the depth of field lines connect the spread of focus the needed proper criterion for depth of field covering focus has only been half met. Hence both the 5cm f1.4 and the 5cm f2 are at the marginal end of depth of field performance, not just at 3 feet but at 4 feet and 5 feet as well.

Obviously, other lenses such as the 8.5cm f2 Nikkors and Sonnars yield very noticeably unacceptable results under similar circumstances. Further, such photographic tools as hyperfocus do not work properly unless focus is compensated. Lenses, such as the Micro-Nikkor, wherein depth of field is actively employed, are rendered limited in use unless its focus offset is also assiduously compensated.

* The reader should note that of all of the 5cm lenses, only the external bayonet mount variety of the 5cm f1.1 has its own focus helix and engraved distance scale which can be used against physical distance measurements. All of the other Nikkors and Sonnars employ the camera body's built-in scale and associated focus helix and that is exactly where the error arises when a 5cm "normal" is employed on the other manufacturer's body.

It turns out that this is a source of a very significant family of problems for vintage rangefinder cameras and their lenses. Nippon Kogaku apparently screwed up very significantly in this area of copying the Zeiss/Contax rangefinder technology scheme. How profoundly and embarrassingly it blundered may go a long ways towards explaining its reticence to illuminate the issues over the decades since the problems first surfaced. Part of the problem is the inherently flawed nature of trying to couple a mechanical rangefinder device that has a well defined "LEVER" with a lens which focusing characteristics may differ from sample to sample. So long as the subject is far away (generally 8 feet or so) and/or the aperture is small (generally f5.6 or smaller) the problems are subtle enough as to not be easily measurable or even detectable. However, close up (like at 3 feet), and wide open (like at f2, f1.4 or f1.1) small deviations in actual lens focal lengths on the order if a fraction of a millimeter will have a measurable impact on the focus ability of the lens/rangefinder system. With the focal length standards differing by a couple of millimeters between the two rangefinder systems, the impact on proper focus of all but wide angle, small aperture lenses is very significant. Probably the reason why it has not been widely previously reported is that it is most significant in the much lesser employed regime of close up wide open aperture exposures. Poor pictures have probably been attributed in the past to photographer error(s). For Nippon Kogaku to not have reported this issue with its "NORMALS" when mounted on Contax bodies, where it clearly is a very significant issue, can well be viewed as acts of deliberate self serving negligence. Since this hypothesis is easily photographically tested and applies to lenses that are widely held, we welcome efforts putting it to careful rigorous tests. We suggest possible simple tests as follow:

The simplest set of tests is to mount either an f2, f1.4 or f1.1(internal mount only) 5cm Nikkor "NORMAL" lens on a Contax body; set the camera body's bezel distance scale to 3 feet; then take a sequence of maximum aperture pictures at distances (physical distances from film plane to subject) of 3'2", 3'1", 3'0", 2'11", 2'10", 2'9" and 2'8". Since the Contax bezel/focus-helix is based on an approximate 2mm longer base focal length standard, the Nikkor lens on the Contax body will find itself racked out to a longer distance than will focus optimally at 3 feet. In fact, when the Contax bezel is set to three feet the Nikkor lens finds itself focussed at just under 2' 11", a disparity not at all covered by the depth of field at maximum aperture.

With the external mount 5cm f1.1 Nikkor a test that need not involve pictures suggests itself. First verify that the Contax aligns properly at 3' 0" against an actual measured distance standard. Then mount the f1.1 Nikkor on the Contax body. Focus the lens employing the Contax rangefinder system on an exact 3 foot distant subject. The angular displacement, based the longer base focal length of the Zeiss/Contax standard will cause the 5cm f1.1 Nikkor to be at under 2' 11" and to be not optimally focused. A sequence of photographs, as suggested in the paragraph above, can be done to confirm focus center but, since the external mount 5cm f1.1 has its own camera body independent focus helix, there would seem to be no point in performing this confirming exercise.

Leica, which has in the past specified its lens executions to the nearest 0.1mm, can adjust for this issue with a custom adjusted cam pitch. SLRs finesse this entire issue with a user employed surrogate focusing reflex film plane.




Summarizing: With this new view of the of the causation of Contax/Nikon rangefinder issues, some separate issues would seem to get tied together into a tidy unified package.

Even Zeiss's long ago accusation that Nippon Kogaku's initial formulation of the 5cm f1.4 is really an f1.5, may tie in. What may have happened is that the people at Zeiss, believing that NK's 5cm ?normals? are 5.3cm in focal length and not actually the 5.1cm basis of NK's erroneous copy of the Contax coupled rangefinder, probably simply measured the diameter of the front element of NK's ?normal? and noticed that it is too close to that of its own f1.5 to be legitimately be designated a real f1.4. A later slight reformulation of that lens actually increased that front element diameter somewhat in order to reduce vignetting but probably not to the end of legitimizing the f1.4 assertion.

All Contax and Nikon calibrated lenses work very satisfactorily on the other's bodies at or near infinity employing the bodies' built-in rangefinder. All lenses that have their own distance scale (among the ?normals? only the external bayonet mount f1.1 can be included in this list.) will focus perfectly, away from infinity, if the body's rangefinder is ignored and actual physical distance measurements are employed. Alternatively, the built-in coupled rangefinder can be directly employed in conjunction with the set of offsets enumerated in the table above.

These correction offsets apply equally in angular displacement/distance offsets to all lenses used on the other's bodies; this is true even for the wide angle lenses though impact is unlikely to be easily measurable on any lens except, perhaps, for the 3.5cm f1.8 . Though Nippon Kogaku only admitted that there is a problem for longer focal lengths, it is pervasive as a problem with many other lenses as well, for close up wide aperture exposures. Perhaps a good rule of thumb is that for non retro focus optical designs (this is actually virtually all of them), any front element dimension greater than ½ inch signifies close focus open aperture problems. The most astonishing aspect of these findings is the relevance to 5cm ?normals?. All of them are impacted to varying unacceptable degrees. Particularly any with a maximum aperture of f2 or greater cannot be used near and under 5 feet without corrections. And, when a Micro Nikkor is employed on a Contax body, there must be aggressively executed adjustments to how the depth of field scale is employed.