Story of a 35mm: Nikon Canon Zeiss

[flavio81]

Hi, sorry for the clickbaity title, but I want to show two interesting things.

1. The tale of the 35mm compact manual focus lenses. These lenses seemed to have gone through many redesigns; in the case of Nikon the 35/2.8 went through 4 optical versions, in the case of Canon, through two optical versions. Those variations are often unknown to Nikonistas and Canonistas.

2. How optical configurations are adopted from one manufacturer to another.

I started researching the mysterious Canon FD 35/3.5. I say "mysterious" because few people are aware that this lens went through a redesign in 1975, losing 1 lens element and adopting a different configuration (5/5 instead of 6/6). Many sources of info (including the Canon Camera Museum) omit this fact and state that all versions are 6/6, while not only I have a Canon catalog that shows the 5/5 version, but I also have the diagram.

Now, the following image explains something very interesting about the evolution of such designs.

We have the 2nd version of Nikkor 35/2.8 lens, which is very similar in design to the MIR-1b 37/2.8 russian wide-angle of 1958 and there are many lenses with the same design. This was manufactured by nikon since the early 60s (note that the image says 1959 but in 1959 Nikon used a different 35/2.8 design, this is covered in "The thousand and one Nights").

But what happened in 1968? Canon released its first 35mm "compact" wide-angle, the FL 35/3.5 in 1968, part of a computation that introduced some higher performing lenses like the FD 55/1.4 II (same design as the famed FD 50/1.4) and famed FL 55/1.2 (ditto with the FD version). Canon explained in some leaflets that this was part of a new series of "lenses with higher performance" and it's obvious that this was done in preparation of the release of the FD lenses and professional F-1 system (1971).

I don't know if the design was inspired of borrowed by a preceding german lens, it's possible but so far yet i haven't found it. The preceding Leica Elmarit-R 35/2.8 (1964), and elder Distagon 35/4.0 for contarex (1959?), all have different designs.

This 1968 Canon design was to be used also in the FD 35/3.5 that was introduced in 1971 until its revision in 1975.

Now, in 1974 Nikon released a lens that closely follows Canon's FL 35/3.5 configuration (see image)

But the more interesting thing is that in 1973 Zeiss (West) releases the 35/2.8 for the Rollei mount, with a 5/5 design.

This Zeiss design with less elements (5/5) is then used by Canon for the FD 35/3.5 (2nd optical version) of March 1975.

And then in 1977, Nikon follows suit and adopts the same design, although for a faster f2.8 design of course. In 1979, the Canon New FD 35/2.8 will have a similar design but with 6 elements in 5 groups.

An image will follow.

 

[flavio81]

The attached image explains the evolution, please take a look.

Source of the Nikkor images: Marco Cavina (great italian blogger)
Source of the Canon images: Canon catalogs.

Have you found more lenses that follows these configurations? I've seen some Minolta 35mm lenses that also use the 5/5 configuration.

 

[flavio81]

Smaller version:

 

[Auer]

Pentax made some excellent 35's, like the https://www.pentaxforums.com/lensreviews/SMC-Pentax-K-35mm-F2-Lens.html

 

[Christiaan Phleger]

I went thru a similar journey a few years ago based in the Nikon system. At the time I was coming across and finding used gear and Nikon bodies that had different versions of this lens so I tested and kept what I liked and sold the rest. None of the Nikkors were 'bad' but the oldest single coated 'S' version could be considered the 'lowest' performing of the versions but even then it had it's own charms and rendering. I didn't care for the Series E version (2.5? Perhaps from the Nikonos line) but found the 'K' optical arrangement with the thick elements to my own liking. It seemed to me like the 'K' version gave a more round and full look to the image and had a really robust color look which was quite nice on a saturated print film or punched up digital capture. Interestingly it seems soon after this lens was released Nikon moved away from this style of lens designs, shifting to thinner but more numerous elements especially for a consumer oriented lens design. Other 'thick' designs include the 20mm f/4 and the 28mm f/2.8 Non-CRC both of which I also prefer.

 

[flavio81]

When reading this, I was wondering whether patents played any role in the optical industry as manufacturers routinely copied each others designs throughout history. Feels like the difference between inferior vs superior lenses today mainly comes down mainly to materials (glass, coatings), quality assurance and branding.

Many many lenses were patented and some designs were unique to some manufacturers.

Quality assurance is very important, of course.

 

[Auer]

I forgot, five elements.. https://www.pentaxforums.com/lensreviews/SMC-Pentax-K-35mm-F3.5-Lens.html

 

[flavio81]

Pentax made some excellent 35's, like the https://www.pentaxforums.com/lensreviews/SMC-Pentax-K-35mm-F2-Lens.html

Yes I love Pentax lenses. But the 35/3.5 design by Pentax is very different to the ones depicted and predates all of them. It traces back its lineage to one of the Voigltlander 35mm lenses for the Bessamatic. Of course the 35/3.5 S-T and SMCT lenses are very good and compact.

Interestingly it seems that Pentax was the first japanese maker to bring out a retrofocus wideangle. It was also the first to release a 50/1.4.

I am not covering the f2.0 fast lenses here, and I did take a look at the evolution of the Pentax 35/2 designs. I had the Pentax-M 35/2.0 but promptly sold it, i didn't like it at full aperture. The K version is reportedly much better.

 

[flavio81]

I went thru a similar journey a few years ago based in the Nikon system. At the time I was coming across and finding used gear and Nikon bodies that had different versions of this lens so I tested and kept what I liked and sold the rest. None of the Nikkors were 'bad' but the oldest single coated 'S' version could be considered the 'lowest' performing of the versions but even then it had it's own charms and rendering. I didn't care for the Series E version (2.5? Perhaps from the Nikonos line) but found the 'K' optical arrangement with the thick elements to my own liking. It seemed to me like the 'K' version gave a more round and full look to the image and had a really robust color look which was quite nice on a saturated print film or punched up digital capture. Interestingly it seems soon after this lens was released Nikon moved away from this style of lens designs, shifting to thinner but more numerous elements especially for a consumer oriented lens design. Other 'thick' designs include the 20mm f/4 and the 28mm f/2.8 Non-CRC both of which I also prefer.

This is a very interesting observation, thanks. I wonder if Richard Haw's web page has sample images with the K version?

Since I own the 35/3.5 and 35/2.8 PC-nikkors i never felt the need for the non-shifting lenses. And I have the AI 35/2.0 which I think is great.

On Canon i have the FD 35/3.5 (both versions) and the FD 35/2.0 (first version with radioactive elements), plus the FD new 35/2.8. They are all great lenses, perhaps the 35/2.8 is the most convenient.

On Pentax I stay with the 40/2.8 pancake, and in M42 i have the Carl Zeiss Jena Flektogon 35/2.4 MC; it is a very good lens in terms of rendering.

 

[Christiaan Phleger]

Yes the 'thick era' for Nikkor Wide Angles seemed to be a continuation of thought from the early 1970's with a thick element located usually one element in front of the aperture; the 35 and 28 I mentioned were perhaps the most consumer oriented of these and Nikon made a reasonably large amount of them before they were phased out. I think the continuing development of glass meant that Nikon shifted designs from the few narrow and thicker elements to that of a thinner and larger diameter and more elements. For this reason I prefer the 24mm Nikkor N.C over the Nikkor 24mm Ai or Ai-s with its slightly altered optical formula change.

 

[flavio81]

I think the continuing development of glass meant that Nikon shifted designs from the few narrow and thicker elements to that of a thinner and larger diameter and more elements.

I think it had nothing to do with glass types, but a lot to do with the Olympus OM system. Small lenses were the new fad and it's evident how Nikon, Canon and Pentax started reducing their lenses from the mid 70s onwards. Most Nikkor AI (1977) lenses feature reduced size. All Canon New FD (1979) lenses are smaller than the predecessors. And the Pentax M system (1976?) is the most extreme example.

The comparison of some Pentax M optics to the previous, bigger K optics show size reduction can have a negative impact in optical quality in some cases.

 

[Bill Burk]

My favorite 35mm lens.

 

[Christiaan Phleger]

Oh there's no doubt the marketplace and all the camera makers at that point were influenced by Olympus OM (my first system) but I think considering what Nikon lenses were about to released (fast and large) and judging from just the increase in production numbers Nikon made a push into higher volume glass production. Would thinner and larger diameter of high refractive glass be more productive than thicker and narrow? Considering that Nikon from the K of the mid 1970's to the Ai-S of the mid 1980's made a big jump; 135mm f/2, 85mm f/1.4, the ED-IF series and the ultra-wides of 13mm and 15mm wouldn't have been possible without big steps in glass production as well as computational power.

 

[reddesert]

Newly available and consistent optical glasses, specifically high-index low dispersion glasses, and computer-aided optical design, certainly drove many improvements in photographic lens design in the 1960s and 70s and onward.

You can look at a Nikon patent related to the 35/2.8 lenses. I found it from https://cameragossip.github.io/nikon-lens-patents.html . That site links to this patent: https://patents.google.com/patent/US3731989A (filed 1971, but lens production could have started before or after). The patent is for a 5-element 35mm f/2.8 lens (angle of view 62 deg) with a thick element in front of the aperture stop, and doesn't exactly correspond to the design of any of these 35mm versions, but gives some idea of the design considerations. Maybe the AI version is a descendant of it.

They give several example designs with tables of radii of curvature, indexes of refraction n_d, and Abbe numbers v_d (inverse dispersion) for the 5 lenses. One thing to note is that the second element (the thick one), the 4th, and 5th, are always made from glass with relatively high index 1.6-1.7 and low dispersion (Abbe v as high as 49 to 60). These glasses are lanthanide crowns, for example they specify in several places a glass with n=1.71 and v=53.9, which is pretty close to Schott N-LAK8. You can use Schott's Abbe diagram to explore glass types at https://www.us.schott.com/advanced_...chnical-articles-and-tools/abbe-diagramm.html (although Nikon probably didn't use much Schott glass, you can buy very similar glasses from Ohara, or maybe they made it themselves).

Anyway, the patent gives some information about the design considerations but you need a real lens designer to understand exactly the effect of having a thick lens there rather than an airspaced doublet. FWIW, I don't think size or weight was the main design consideration here. The 35mm K version with the thick lens isn't that big, it has a smaller front element than the earlier Nikkor-S 35mm 7-element version. I think they're all about the same physical size until you get to the Series E 35/2.5, which is smaller.

 

[Christiaan Phleger]

Yes; exactly my point. Much like Marco Canvina my own reading thru many of the Nikkor Patents gave some insight about the glass types being used in certain ways. In this case the thick but high index element is really not that large a piece of glass, and as I mentioned they made almost 100,000 of the 35mm 'K' 2.8 formula made so its far from a rare lens, and as a consumer lens Nikon must have found it to have a benefit from a cost/perfomance to both first incorporate it and then change it as consumer tastes shifted to zooms in the early 1980's. Interestingly I seem to recall that some of Zeiss's designs for Contax wide angles which were mostly computed in the early 1970's had a thick either single element or two thick elements cemented.

 

[Paul Howell]

Konica had 3 versions, 2 were 5 elements in 5 groups, one in 6 elements in 5 groups, Miranda had 2 versions both 5 in 4 groups. I pretty sure Pentax both M42 and K had evolved over time. The most interesting 35mm lens I have is the Konica, 35 F2, 9 elements in 7 groups.

 

[M-88]

Olympus had this 35 mm f/2.8 with 7 elements in 6 groups, incredibly compact, as small as their 50mm f/1.8, but in terms of image quality it's nothing to write home about. At least not when wider than f/5.6. People say 35 mm f/2 is way better, but I never had that one. It has eight elements. I take it Olympus design their own 35 mm independently from others?

 

[Christiaan Phleger]

Yes, Olympus did its own optical design.

The rapid photographic technology advance of the the 1970's cannot be overemphasized IMO. Take any one year of the 70's and skip ahead ten years and look at a few of the many photographic jumps; Zooms, wides and ED glass tele's in 1974 and then in 1984; Match needle metering in 1976 vs Matrix metering in 1986; state of the art viewfinders and focussing screen aids of 1979 vs the AF revolution of 1989. I am sure it is possible to think of a few more (not to mention what happened with film). As the market grew with the consumer bait of "new and improved" more resources allowed each company to indulge in design and accessories, and each company could design and produce a reasonably priced 2.8 35mm optic. Maybe a couple!

 

[blockend]

I had the Pentax-M 35/2.0 but promptly sold it, i didn't like it at full aperture.

Older lenses never perform well at full aperture, in my experience. Contrast this with the latest RF 50mm 1.8 for Canon mirrorless cameras. Except for the corners, this lens exhibits maximum sharpness wide open, even with the highest resolution sensor. If you opt for an aps-c sized crop, the lens is pin sharp across the frame at maximum aperture, remarkable for a "kit" lens. Whether this is due to optical design or improved coatings I don't know, but it would be interesting to see the formula of the new lens.

 

[flavio81]

Would thinner and larger diameter of high refractive glass be more productive than thicker and narrow?

I'd say that the glass type would be a more important factor. A special type glass over regular optical glass (also used in good lenses) might be orders of magnitude more expensive.

The Leica glass 900403, their so called APO glass consists of no fewer than a dozen different ingredients, including the rare earth element lanthanum. One kilogram of this glass costs almost 60 times as much as a common optical glass such as BK7. Please keep in mind that this APO glass is twice as heavy as BK7. Subsequently the volume of glass in a kilogram that can be turned into lens elements is only half as great, making the actual price ratio between these two optical glasses approximately 120-to-1. This is the very reason why relatively inexpensive lenses rarely, if ever, contain any glasses of this quality.​

http://gmpphoto.blogspot.com/2012/02/by-heinz-richter-few-industrial.html

 

[flavio81]

Older lenses never perform well at full aperture, in my experience. Contrast this with the latest RF 50mm 1.8 for Canon mirrorless cameras. Except for the corners, this lens exhibits maximum sharpness wide open, even with the highest resolution sensor..

I meant that the Pentax-M 35/2 (ca. 1977) is universally reputed to be lower performing than the older K 35/2. And it was lower performing than the Nikkor-O 35/2 (1967) i had. And most likely my Canon FD 35/2 concave version (1971) is better, everybody raves about that one (i've yet to test it) being pin sharp wide open.

Regarding the Canon RF lens you cite, it isn't fair to compare a lens that will be over 40mm away from the film/sensor plane with a lens whose rear element almost touches the sensor plane. The latter will give optical designers far more optical freedom. That's why rangefinder lenses often outperformed reflex lenses of the same vintage.

 

[flavio81]

View attachment 263936

My favorite 35mm lens.

I had the chance to buy one but was put off by the size!

 

[Auer]

My Olympus Infinity Stylus's 35mm 1:3.5 is the best 35mm I own

 

[dynachrome]

Lenses like the 35/3.5 Canon FL and the later 35/3.5 FDs were not made specifically to be compact. They allowed someone to get that focal length for a lower price. I have a 35/2 SMC Pentax-M and find it sharp but I use it with film. I had a 35/2 Super Takumar for one day. There was play in the focusing mount so I returned it. My favorite 35/2.8 Nikkor? The 6 element K/early AI. The last two versions of the Minolta and Celtic 35/2.8 are also very good.

 

[reddesert]

I'd say that the glass type would be a more important factor. A special type glass over regular optical glass (also used in good lenses) might be orders of magnitude more expensive.

The Leica glass 900403, their so called APO glass consists of no fewer than a dozen different ingredients, including the rare earth element lanthanum. One kilogram of this glass costs almost 60 times as much as a common optical glass such as BK7. Please keep in mind that this APO glass is twice as heavy as BK7. Subsequently the volume of glass in a kilogram that can be turned into lens elements is only half as great, making the actual price ratio between these two optical glasses approximately 120-to-1. This is the very reason why relatively inexpensive lenses rarely, if ever, contain any glasses of this quality.​

http://gmpphoto.blogspot.com/2012/02/by-heinz-richter-few-industrial.html

This article has a number of things that are basically true but one has to discount everything by 50% for the particular company advocacy, ie GREATEST OPTICAL MANUFACTURER EVER puffery. This is not to criticize your conclusions, merely the biases of the article.

Most optical glass manufacturers are capable of making a wide array of glasses to suit lens designs for all but the most exotic lenses (excluding aspheric and molded/machined designs which are another realm of manufacturing difficulties). One cool thing is that Ohara gives us mortals a rough idea of what each of their glasses costs. Have a look at their chart: https://www.oharacorp.com/pdf/pricing-chart-2018-01.pdf They only give rough price ranges and production frequencies, to get real numbers one needs to request a quote. Nevertheless, they have nice price ranges relative to S-BSL7, which is the Ohara borosilicate crown glass equivalent to Schott's N-BK7. This is basically the lowest cost optical glass, so if you order a random lens or window from Edmund Scientific, it's usually made out of N-BK7 or the equivalent. Ohara says it cost ~ $10/lb in 2018.

The high-index rare earth glasses that we identified from the Nikon patent were lanthanide crowns with say index n ~ 1.7 and Abbe number ~ 50-60. From the Ohara chart color coding, these are mostly in the "medium" price range of 3.5-6 times the borosilicate crown, although there are a few that are higher. It's possible that in 1970 when these glasses were more recently developed that the difference was greater.

The small elements of a wide angle lens probably each weigh less than a ounce, although I'd guess each began as a blank that weighed maybe twice as much? So at today's prices, the bulk glass for each element in an ordinary lens might cost from $1 up to ~ $5 for a rare earth element as a very rough estimate. Certainly for mass production the price of the glass is a factor restricting the use of expensive glasses, but the fabrication and assembly costs probably count more than just the raw material cost.

Once you get to really large elements or flourides such as CaF2 or BaF2, all bets are off. There, the difficulties of casting and fabricating can be much greater and the glass really is a significant engineering problem.