Nikon SP

Nippon Kogaku had already decided not to build it. Management had judged that piling further complexity onto a coupled-rangefinder body was irrational, and the view inside the company was settled: 「Ｓ２の後は一眼レフだ」 — after the S2, the SLR. What produced the Nikon SP instead was the American distributor. Ehrenreich, importing into the United States, wanted a camera that could go against the M3, and the consequence was that the SP, the S3 and the Nikon F were all developed at the same time. Fuketa Masahiko (更田正彦), who designed the F and later became a Nikon vice-president, said afterwards that without that pressure it was very likely the company would have gone to reflexes sooner. The usual telling of this story — that the M3 shocked Nikon out of rangefinders — has the causation backwards: the M3 is the reason the most elaborate Nikon rangefinder exists at all. The same testimony records that a new lens mount was considered for the SP, with an adapter to keep the existing lenses usable, and dropped only because the adapter could not meet the shock requirements. The name was settled in May 1957, the P standing for professional.

Everything else about the camera is downstream of the finder. Nikon began work on it the moment the S2 was finished, with the declared objective of beating the variable finder on the M3; specifications were fixed in September 1954, sectional prototyping began early in 1955, and as many as ten different finder prototypes were built. Two approaches lost. A mask-switching real-image finder gave the clearest view of the field but was simply too large for the body Nikon intended to build; a virtual-image zoom finder was beaten by the switchable brightframe design for the same reason, zoom finders of the mid-fifties being bulky things. What shipped was two finders behind one long window. The main finder is life-size, carrying brightframes for 5cm, 8.5cm, 10.5cm and 13.5cm in white, white, yellow and red, with parallax corrected automatically by a linkage to the rangefinder; beside it sits a separate 0.34× finder whose whole field is the 2.8cm view, with a 3.5cm frame marked inside it and no rangefinder image at all. Six focal lengths without an accessory finder, as Nikon Inc.’s dealer price list of 1 September 1957 advertised in as many words, and 写真工業 magazine noted that November that nothing since the Leica M3 had let a photographer work across so many lenses without special finders. It was expensive to make: when Nikon built the finder again for the 2005 reissue it counted 28 lens and prism parts against 13 in the S3, and roughly ten times the assembly time of the FM3A then coming off the line.

For all that, the finder drew criticism immediately, and Nikon’s own record preserves it — after the September 1957 release, critics called the viewfinder slightly dark. The compensations are real. The brightframes are front-illuminated, uniquely among the S-series Nikons, so they do not glare, take on the photographer’s skin tone, or reverse black and white when the sun is on the face, all of which the S2’s Albada frames can do; and the semi-reflective rangefinder patch is a generous trapezoid where the S3’s is a plain rectangle, small enough to disappear inside a dilated pupil indoors. But the number collectors most want is the one nobody published. Nikon appears never to have stated a rangefinder base length for the SP anywhere: not in the corporate history, not in either Camera Chronicle volume covering the camera, not on the Limited Edition page, not in the manual it hosts. Sixty millimetres is what the 1953 finder study settled on for the life-size layout the S2 shipped with and the SP inherited, and it is the figure carried here for want of a better; 58mm also circulates for the SP specifically. Either figure loses the comparison that gets made: 60mm at life size gives an effective baselength of 60mm, and 58mm gives 58mm, against 62.3mm for the M3, whose 68.5mm base is read at 0.91×. That is the whole arithmetic behind the long-running argument that the Leica focuses a fast long lens more precisely — an argument conducted, it is worth saying, entirely on numbers Nikon never printed.

Underneath the finder, the SP is the S3. Ed Romney’s factory-derived repair manual opens by saying so outright, and Nikon’s internal order numbers say it from the other end: the S3 was 26F1B and the SP 26F2B, the simpler camera carrying the earlier number because the two were laid out as one series and the complicated one was simply built first. Fitting the wide finder inside S2 dimensions — an explicit design goal — cost something real: the rewind shaft has to be geared around the wide finder’s optics, and hard use is said to develop play there. Elsewhere the changes went the other way, with thicker-walled internal parts, and with almost every external edge rounded and every corner radiused after complaints that earlier Nikons hurt to hold all day. The shutter was new work. Curtain travel came down to 14.5 milliseconds across the 36mm frame, against 16 for the S2 and 18 for the Nikon I, helped by a rear-curtain accelerating device and a brake built from a spring coiled around a rotating cylinder to kill the flick the older shutters made; flash sync moved to 1/60, and the speeds were coloured green, red and white by synchronisation range, the green in fluorescent paint so it could be found in the dark. From 1959 the rubberized silk gave way to titanium foil, and which camera received titanium first is a question Nikon answers two ways in its own publications: the corporate history dates the SP’s titanium curtains to production after May 1959 and calls it a world first later carried over to the Nikon F, while Camera Chronicle Vol.10 says the curtains were completed after the F was announced and that the SP was modified afterwards to accept them. Japanese Wikipedia and nikomat both follow Vol.10, and the F was itself announced in May 1959, so the priority claim is best left open — the more so because service departments re-curtained these cameras for decades, and a titanium shutter tells you nothing certain about a given body’s date.

The SP was also built around its motor. The S36 was released alongside the camera in September 1957, arriving as a complete replacement back, adding a little over an inch of height and eight ounces, and running single frames, short bursts or a whole cassette at about three frames a second, with a subtractive counter settable anywhere from 1 to 36 and a clutch that slips rather than tear the film off the spool; a body had to go to a service centre for a small modification before it would take one. Its immediate ancestor, the motorised S2E shown in March 1957, had been made in two-digit numbers, and its 32 motor drives interchange with the S36. Then the market moved. The 鍋底不況 — the pot-bottom recession — arrived alongside the shift to single-lens reflexes, and prices were cut about thirteen per cent in March 1959, to ¥85,000, at the same time as the S4 launched. Nikon’s rangefinder output passed an aggregate 130,000 in 1959 and the company decided to end it, only to reopen the line because requests from press and magazine photographers were too strong to refuse: the 623xxxx cameras, built to order, usually with the new Double-Gauss 50mm f/1.4, and dated to 1964 by Japanese sources against a corporate history that contrives to recommend production in May 1964 and discontinue it in April 1963 in one sentence. How many SPs were made altogether is not settled either — 22,348 is the number everyone repeats, and Nikon’s own history says a little more than 21,300. Development did not stop with the F: a TTL-metered SPX and a zoom-finder SP II both existed, and neither reached production.

The finder was built once more all the same. Mito Nikon, the plant behind the Nikomat, the F3 and the FM3A, had begun asking in the spring of 1994 — at the height of the classic-camera boom Leica had set off — whether an S-series body could be made again at all, and when the work finally went ahead it went ahead on the S3, for the least romantic of reasons: the SP’s universal finder had already beaten the engineers once, and they had given it up. They were not content to leave it there. Data collection on the SP began in July 2001, a year after the S3 reissue started shipping, and that August two secondhand SP bodies were bought from a shop in Ginza to work from. The body came to 895 parts, and for the finder no assembly procedure had survived; the drawings gave nothing to measure from either, the engineers finding, in Nikon’s words, no approach to define a position as a reference simply by looking at them. They were stopped for half a year. What broke it was an optical-axis rig of their own design, a collimator and lasers dialling in each finder as it was assembled — after which the better modern prisms threw ghost flare off misalignments the old glass had been hiding, trimmed away in increments of two tenths of a millimetre. The project was approved in September 2002, trial manufacture began at Mito in February 2003 and volume production that July under the code M400, and the first bodies left the factory in March 2004, a full year before any customer received one.

Nothing had been kept from the fifties, so every die was cut new, press and die-cast alike; assembly was benchwork, done by hand alongside the FM3A line, with the lens and the finder optics made at Tochigi. Orders were taken from 14 January 2005 and the first camera was handed over on 18 March, at ¥690,000 — ¥724,500 with tax — for a body in black paint on cloth curtains, which Akagi Koichi (赤城耕一) objected to in public, the late originals having been titanium, though he found the rebuilt finder clearer than the surviving originals he compared it with. The 2,500 cameras were numbered 0001 to 2500 without a gap, each sold with a W-Nikkor 3.5cm f/1.8 carrying its own body’s number and a certificate tying the two together; number 0001, body and lens both, went to the JCII Camera Museum. And the reissue casts one last small shadow over the record: the dimensions and weight quoted everywhere for the 1957 camera — 136 × 81 × 43mm, 590g — are character for character the figures Nikon publishes for the 2005 reproduction, and no independently measured original has been published at all. Even on its own flagship, the documentation is thinner than it looks.