Nikkor 35/2.8 PC (late version)

[Airy]

as you may see from the "Crêpes" sign in the top right, coma correction is not perfect. Still, the lens is usable. Its contrast wide open increases the perceived sharpness ; the PC lens has a tendency to look dull in comparison, even when stopped down. That may well fit some scenes (e.g. the Christ shot, first page), but is otherwise not that seductive.

[Roland Vink]

Léon Trulin, taken hostage and executed in 1915. This is an attempt to use the PC lens at a wider aperture - do not expect wonders.

Rendering of backgrounds is smooth pleasant, no outlining or double-line blurs.

I wonder how the older version performs. It originates from 1968 and uses 8 elements in 7 groups. They are single coated and have the old styling with all metal barrel and scalloped focus ring. From 1975 the barrel was modernised and the optics were multicoated, although the optical design stayed the same. The newer (late) version is from 1980 and uses 7 element in 7 groups.

The overall optical design of the two versions is very similar, the main difference is that the cemented lens in the older version is combined into a single element. I'm sure there is much more to it than that, the curvature, spacing and glass materials were almost certainly updated, but the overall layout is very similar. Did Nikon manage to improve the performance with a simpler, more refined design, or was it a cost-cutting exercise?

[Roland Vink]

The place in front of the cathedral somehow delivers more piece of mind, at nighttime of course.

With 9 straight-edge aperture blades, the star-bursts are very nice!

[mxbianco]

Various building and boutiques, not far from the Opera house.

Your photos in this thread made me want to try my own sample of PC-Nikkor 35/2.8 latest version (=black knob).
I tried on three different cameras, a D2Xs, a D810, and a D500. My sample is currently unchipped.
I found out that, on account of the lack of communication between camera and lens, the f-stop reported (all three cameras) in the EXIF stays at 2.8. The situation does not change even if you change the Customize Command Dials... Aperture Setting between OFF (Aperture ring) and ON (Sub-command Dial). Maybe is it one of the magics that the DF operates with MF and exotic lenses?
I have another lens with preselect aperture (the 7.5mm/5.6), which I succesfully chipped. Passing the lens' f-stop to the EXIF can be achieved by setting Aperture Dial to ON (= subcommand dial), and then dialing in the actual f-stop setting on the lens. Not a big deal, as long as the f-stop is recorded.

So, the question is:
I don't have a DF to test myself, do you do anything particular on the DF to have the f-stop accounted for in the EXIF?
I assume your lens is unchipped, otherwise the lens type would be reported in the EXIF field "-nikon:lens" (instead on your DF photos it is a dash "-", on my DSLRs an unchipped lens is reported as a "0mm/0")

Ciao from Massimo

[Airy]

My lens is unchipped, so I dial in the FL and max aperture to identify it 35 2.8
I meter in stopped-down mode and consequently identify the lens as AI, so I do not have to handle the aperture dial in addition to the rest. Even so, it is easy to forget some manoeuver...

Of course, It is also possible (I just did it, Df-style) to
- declare the lens as non-AI
- set the diaph control to match the preset aperture
- set the exposure while wide open and no shift (instead of stopped down and no shift)
The benefit is, the preset aperture gets recorded. On the other hand, I do not care much - the shift value is equally important, and never gets recorded...

[mxbianco]

Thanks, Airy!

If I decide to use this lens, I probably am going to chip it, so it will work even with cameras without a non-CPU lens menu... One less thing to remember!

I'll have to look if the shift value gets recorded in more recent lenses (I have the 24mm PCE and the 85mm PCE micro)

Ciao from Massimo

[Airy]

That's seemingly not the case (I do own the 24 PCE...).

[mxbianco]

That's seemingly not the case (I do own the 24 PCE...).

I'll take your word for it, I could not find any field in EXIF as a potential candidate for holding the shift (and Tilt) value. For example,  the GPS fields ( -Gps:GPSLatitude and -Gps:GPSLongitude), albeit empty, are present in the EXIF even if no GPS is connected to the camera.

Ciao from Massimo

[Cyril]

Un ami / une amie Lillois? I really miss that city, and you photographed places I would walk about every day.

I really love the star shaped flares in the pictures you've posted. I already have the 35mm DX 1.8 G Nikkor. But not a big fan of it either...
Cheers from a fellow countryman. 

[Kenneth Rich]

Airy, your comments re using this lens with the Df interest me; I find this a difficult lens to use, and so do not use it. I'd like to avoid using the "front wheel" if possible.  Can you set me on the most simple road to success with this lens and my Df?

[Akira]

I have always been curious how or if the chromatic abberation of magnification (CAoM) of the images shot with the lenses shifted could be corrected in the software.

If I understand correctly, CAoM can be corrected simply by adjusting sizes (magnifications) of images of the blue- and the red channnels which do not coincide with the image size of the green channel, so long as the optical axis coincide with the center of the image.

On the other hand, there is no way for the software to know where the optical axis of the lens is when the lens is shifted and its amount of shift is not recorded, which make it impossible to correct CAoM using the same method.

[Airy]

Un ami / une amie Lillois? I really miss that city, and you photographed places I would walk about every day.

I really love the star shaped flares in the pictures you've posted. I already have the 35mm DX 1.8 G Nikkor. But not a big fan of it either...
Cheers from a fellow countryman. 

Cheers. By the way, "Airy" is male.

[Airy]

I have always been curious how or if the chromatic abberation of magnification (CAoM) of the images shot with the lenses shifted could be corrected in the software.

If I understand correctly, CAoM can be corrected simply by adjusting sizes (magnifications) of images of the blue- and the red channnels which do not coincide with the image size of the green channel, so long as the optical axis coincide with the center of the image.

On the other hand, there is no way for the software to know where the optical axis of the lens is when the lens is shifted and its amount of shift is not recorded, which make it impossible to correct CAoM using the same method.

That's correct. I remember having read an article, possibly by Ken Rockwell, explaining how to correct CA using layers in photoshop in such cases. However I did not care - for downsampled internet output, CA is no issue with this lens. Anyway I cannot remember any significant CA problem with the PC lenses so far (the 28/3.5 excepted).

[Akira]

That's correct. I remember having read an article, possibly by Ken Rockwell, explaining how to correct CA using layers in photoshop in such cases. However I did not care - for downsampled internet output, CA is no issue with this lens. Anyway I cannot remember any significant CA problem with the PC lenses so far (the 28/3.5 excepted).

Thank you for sharing your experience.  I agree that CA is not something to be worried about when the image is for sharing on the web.  But I'm still curious.

I usually correct CAoM anyway.  I like to watch how it is corrected by magnifying the edge of the image 100%.   

[Airy]

Airy, your comments re using this lens with the Df interest me; I find this a difficult lens to use, and so do not use it. I'd like to avoid using the "front wheel" if possible.  Can you set me on the most simple road to success with this lens and my Df?

I'm afraid one just needs to get used to it. My routine (handheld) is:

0 - pre-shoot : set lens characteristics to 35/2.8 AI (implying stopped down metering); disable Auto ISO, choose a viable ISO, put Df in Manual mode, choose decent speed given the circumstances, pre-set the desired aperture
1 - set aperture to f/2.8 (wide open)
2 - find a subject and orient the lens shift axis according to framing (vertical or horizontal)
3 - focus
4 - close diaph
5 - correct speed and/or ISO, so exposure gets right
6 - re-frame (usually, I try to keep the camera horizontal) and shift (without departing from the horizontal, the only difficult part) - here you'll soon learn how to correlate your finger-twisting with shift millimeters, not to exceed limits
7 - release shutter and go back to 1

I considered changing the routine by treating the lens as a non-AI one, setting the exposure before closing the diaph. But that would mean setting the aperture wheel of the Df to the same value as the preset aperture on the lens, with the risk of forgetting the one or the other. And it does not even simplify the whole procedure, since steps 4 and 5 would merely get swapped.

The nice thing is, the Df has a decent dynamic range (I used to shoot slides and adjusting all relevant areas in a +/- 2 EV range...) and the stopped-down lens usually has enough DoF, so steps 1-5 can often be skipped when shooting the same subject (some street or some room). Sometimes I adjust the focus distance by looking at the distance scale (old-fashioned, again)

Bottom line : even with some training, it is a slow process, and a good cure against itching right fingers.
Before using shift, it is helpful to learn using the lens as a manual lens in stopped-down mode. I can also exercise with my cherished, Leitax-mounted Summicron-R 50/2, which is even worse, as there is no aperture pre-set mechanism and aperture must be "felt" by counting the clicks on stopping down... or just appreciating the darkening of the viewfinder. Even more fun, also considering that the rotation direction for focus and aperture is reversed, compared to Nikkors...

Shots do not need to be perfect "straight out of the camera" ; allow yourself to do minimal corrections, regarding exposure or convergence of lines. As Akira pointed out, strict parallelism is often not satisfactory anyway, and some residual convergence (towards the infinite) is advisable. The amount is best fine-tuned on screen.