Nikkor 55mm f/1.2, Coatings

[Roland Vink]

Agree with Mongo. Very interesting, Ron. The multicoating around 1972-1973 is also claimed to have a positive effect on the light performance, I read somewhere, a 1.8 actually being an 1.5 or 1.6? Is that perhaps the reason why Nikon did split the 85/1.8 into an 85/2 and a 85/1.4 ?

There is no way the 85/1.8 was actually 1.5 or 1.6 ... such a lens would require 53 - 57mm entrance pupil, no way to fit that inside a 52mm filter thread

I think the 85/1.8 was replaced my the AI 85/2 in 1977, following a trend towards more compact lenses. Many other lenses were also replaced with more compact models around the same time: 50/1.4, 135/2.8, 200/4 etc. I guess that also created a bit more "space" to create another 85mm model one stop faster...

[Bjørn Rørslett]

" such a lens would require 53 - 57mm entrance pupil, no way to fit that inside a 52mm filter thread "

Technically speaking it is of course possible, as the pupil is a virtual quantity.

To name an example, the 85 mm f/1 Repro-Nikkor has 86 mm entrance/exit pupil and filter thread 52 mm. Looking into the lens is an uncanny feeling as you literally see a hole bigger than the lens itself.

[simsurace]

Brian, Thank you for posting the Pop Photo test report excerpt.  Norm Goldberg's test reports of cameras and lenses blow away most so-called "tests" that are available today. 

Agreed, and also note the excellent design of the charts.

[null]

One can interpret the Nikon remark as the 35/1.4 being the first Nikkor lens with all surfaces being multi-coated. The 55/1.2 Nikkor-S had the rear groups multi-coated. I also remember this statement from working in a camera shop in the 1970s and discussing this with the Nikon rep. Olympus was one of the last companies to use multi-coated optics. I also remember that discussion with the Oly rep.

Manufacturers also did "test runs" and batches with little fanfare. I have Zeiss lenses from the same batch, one coated and one uncoated. The coated lens is not marked "T". Two years later, coated lenses were marked "T" and uncoated lenses were still in production. Perhaps a limit while the new equipment was being brought online? It happens. SN is not an absolute indication of revision level.

[richardHaw]

my sample looks amazing at f/1.4 by the way 
f/1.2 is "art" lens 

[null]

I always wondered why Schneider used "Arton" and "Tele-Arton" for their lenses. So it was not some guy names "Art" that formulated it. I need to post a thread on using the 90/4 Tele-Arton on the Df.

[richardHaw]

I always wondered why Schneider used "Arton" and "Tele-Arton" for their lenses. So it was not some guy names "Art" that formulated it. I need to post a thread on using the 90/4 Tele-Arton on the Df.

it's an ART lens 

[simsurace]

To name an example, the 85 mm f/1 Repro-Nikkor has 86 mm entrance/exit pupil and filter thread 52 mm. Looking into the lens is an uncanny feeling as you literally see a hole bigger than the lens itself.

I suppose, as the name suggests, the Repro-Nikkor is not designed for use at infinity. I suppose the effective aperture would be smaller than f/1 at far distance, since if you look into the lens from afar, you cannot possibly see a hole bigger than the lens. Or am I fooling myself for thinking that?

[Bjørn Rørslett]

Yes, you are.  The part is indeed larger than the whole which is the reason for this "uncanny feeling" I alluded to earlier. Remember the pupils are virtual entities. They can reside outside the lens as well.

The definition of nominal aperture is for infinity focus. As the entrance (and exit) pupil of the Repro-Nikkor is 86 mm, it is a true f/1 lens. At 1:1 magnification its effective aperture is f/2, which of course is only possible were it an f/1 lens. Compare this to most close-focusing lenses that may run to f/5.6 or smaller effective at the this magnification.

[simsurace]

Yes, you are.  The part is indeed larger than the whole which is the reason for this "uncanny feeling" I alluded to earlier. Remember the pupils are virtual entities.

The definition of nominal aperture is for infinity focus. As the entrance (and exit) pupil of the Repro-Nikkor is 86 mm, it is a true f/1 lens. At 1:1 magnification its effective aperture is f/2, which of course is only possible were it an f/1 lens. Compare this to most close-focusing lenses that may run to f/5.6 or smaller effective at the this magnification.

I guess I have to see it to believe it
I would have thought that you have to move your head around to see the entire hole from across the room. It's hard to imagine seeing a 86mm object through an opening of 52mm from afar.

[Bjørn Rørslett]

One of course has to move something, either lens or head, relative to each other.

Plug the data into the usual equation for effective aperture and find out for yourself that the pupil must be larger than the lens width in this case.

N_eff = N * (1+ M/p)

where N: aperture number, M= magnification, and p= pupil factor (exit/entrance). For the Repro, p=1 as the lens is perfectly symmetric. N_eff is 2, which one can easily verify with a light meter.

[simsurace]

Ok, that we have to move to see the hole (in other words, that the hole is not visible all at once from a certain distance) is reassuring. But if only part of the entrance pupil is visible, isn't the usable pupil effectively smaller? Or put differently, how can the light from one point cover the entire pupil when from the same point, part of the pupil is obscured by the lens casing?

Even in lenses where the entrance pupil is smaller than the front element, i.e. a 50/1.2, looking at the entrance pupil from well off-axis, part of the pupil is obscured by the lens casing (we see at cat-eye shape instead of a circle). This is -- so was my understanding -- part of the reason why we get strong vignetting on the periphery of the image circle with fast lenses. In a lens where the entrance pupil is bigger than the front element, I would think that the effective aperture drops in the far field because we have to be close to the front element to see the full pupil, even on-axis (I think you just confirmed this conjecture). Depending on where the entrance pupil is (how far back in the lens is the virtual image of the aperture, not the aperture itself, located), this may never be observed in practice when the lens is used as intended. But I still wonder whether the image would get darker when focused at a far distance, or if not, what would be the explanation.

At any rate, this lens is intriguing even if you don't have it

[Roland Vink]

I guess this is what we see when looking through the rear element of a wide or standard lens: the pupil is much larger than the rear element when looking at it from a distance. If you get your eyeball close to the lens (as the sensor would be when mounted on a camera) you can see the entire pupil.

[simsurace]

I guess this is what we see when looking through the rear element of a wide or standard lens: the pupil is much larger than the rear element when looking at it from a distance. If you get your eyeball close to the lens (as the sensor would be when mounted on a camera) you can see the entire pupil.

But from the front the situation is different. If I see only half the pupil, the sensor will only see half the light I emit. Ain't it so?

[Bjørn Rørslett]

Pupils are hard to understand. In any connotation of the word.

As the entrance pupil is "collecting" light on behalf of the lens, it does the job even though you don't see all of it.