PRINTING-NIKKOR 105MM F2.8

[mike_halliwell]

With the digital revolution and image stacking, does anyone know if this fully manual lens is any fun?

[ilkka_nissila]

The Printing-Nikkors are stuff of legends and may be excellent for macro. With stacking - of course, why not? You may want to search nikongear.com for detailed information, some of the people there have used them. I don't have personal experience of these lenses, unfortunately.

[jose_angel]

As a great macro 1:1 lens, I think it is perfectly usable.

Another topic is price, or if it is of any use for you, assuming that the provided mount fit your system.

Anyway, I think that all (good) lenses are fun, specially this odd, exotic ones; never used this one but others from Rodenstock or Schneider which I pleasantly keep, just for fun (although I rarely use them).

My last macro shooting were on yesterday, so I just took a common, fast and easy to use Micro-Nikkor, as usual. I also tried it with a Fuji, but prefered the Nikkor images.

Edited May 5, 2020 by jose_angel

[CvhKaar]

Here is a write-up of someone who loves it . (BTW i like this whole "Red-Book" as a free time reading material..).

 

Printing Nikkor 105mm F2.8 Far Eastern Traditional Weekend

[BeBu Lamar]

If you can mount it on a bellow it would be a great lens.

[rodeo_joe1]

Personally, I wouldn't get too excited until you've tried it Mike. The issue with process and copying lenses that weren't designed for the 35mm format is that the over-large image circle can result in a lot of camera body flare and concomitant low contrast.

 

For example: I have a 135mm Schneider Repro Claron that has excellent correction of almost every aberration going. I can mount it on a Nikon bellows and it'll focus from infinity to better than 1:1, but its contrast is just awful. And you can see why by just looking into the bellows. There's so much excess image circle that the bellows pleats light up like a landing strip.

 

I'm sure the Printing Nikkor will have better AR coating than my old Claron, but I really wouldn't expect it to outperform something like a 105mm Micro-Nikkor in terms of image contrast.

Edited May 7, 2020 by rodeo_joe|1

[orsetto]

Author's summarized supporting data for the excellent PDF linked above can be found in several Luminous Landscape posts, such as:

 

The Printing Nikkors Compared

 

The Printing Nikkors for Close-Up Work

 

The Printing Nikkors: Images and Range

 

They are very interesting hunks of glass with unique qualities, but due to their original rarity, original astronomical industrial prices, and difficult mounting requirements they are not generally cost-effective to acquire or deploy today. Most photographers would get more from a premium digital-era Zeiss or Voigtlander macro lens designed for normal DSLR and mirrorless mounting.

[jose_angel]

I'm sure the Printing Nikkor will have better AR coating than my old Claron, but I really wouldn't expect it to outperform something like a 105mm Micro-Nikkor in terms of image contrast.

Agree, it has also been my experience.

As said, I haven't used this specific lens, but I also tested some "high magnification" process and enlarging lenses from Nikon, Rodenstock and Schneider, some better than others, but neither of them clearly outperform my "modern" Micro-Nikkors (say, 105/4, 105/2.8, 105AFS, etc.) in both overall performance and convenience, in (D)SLR cameras.

Edited May 7, 2020 by jose_angel

[mike_halliwell]

Thanks for all the info everyone!

 

I was eyeing up the Venus Optics 100mm 2.8 x2 macro for focus stacking and came across the Printing Nikkors. A seller in China has 5 of the 105mm for sale on Fleabay for ~£700.... each.

 

I'd noticed quite a lot of OOF CA effects on some of my images and was seeking a solution.

[conrad_hoffman]

Not everything is worth the hype or selling price. That said, if you look at the back end of a Micro-Nikkor 55 mm f/2.8, you'll see a sheet metal piece that's the shape of a 35 mm frame acting as a light shield for the interior of the camera or bellows. There's no reason you couldn't fabricate something similar for the back end of a repro lens to kill some of the flare and increase contrast. Heck, you got the thing mounted; that should be a piece of cake. ;)

 

Don't forget the front end either. My macro work always seems to involve lights shining nearly into the camera. Sometimes I use lenses with no easy way to use a proper lens hood. A tube of black construction paper taped in place can do the job. When somebody looks at the image, they can't tell it was made with baling wire and duct tape. I once took a nice shot of a planet transiting the sun with the front of the telescope held up by a garbage can because it was handy and the right height.

[mike_halliwell]

A tube of black construction paper

I found if you put black velour paper into any decent browser you get lots of nice firm cards and papers, some self adhesive. Great for lining tubes etc. and although they won't be a black as genuine blacking, it's pretty good.

[rodeo_joe1]

I'd noticed quite a lot of OOF CA effects on some of my images and was seeking a solution.

A used Apo-Rodagon is probably going to be a lot cheaper than £700. My 50mm f/2.8 Apo-Rodagon is pretty good WRT LoCa. Not perfect, but an improvement on most other lenses.

 

BTW. Since that Printing Nikkor is designed solely for flat copying, there would be no need for LoCa to be perfectly corrected, since it only shows up outside of the plane-of-focus on 3D subjects. I'm not saying that LoCa isn't well-corrected in the Printing Nikkors, just that it doesn't have to be for their designed application.

[mike_halliwell]

there would be no need for LoCa to be perfectly corrected, since it only shows up outside of the plane-of-focus on 3D subjects

I'm not sure how the stacking software would 'see' such a problem?

 

If it's not absolutely sharp, it's transparent....:)

[Andrew Garrard]

Most LoCA, I believe, comes down to different wavelengths being focussed in different planes (I had a diagram on another thread somewhere); the in-focus point of one colour coincides with the circle of confusion caused by an out-of-focus cone in another. That explains different colour bokeh fringes in front of and behind the focal plane (usually chosen to be purple in the foreground and green in the background, since green fringes around trees don't look too weird). If you want fully sharp from a lens system, you usually need to pick one frequency - the old lenses with very high resolution figures tended to hit that only when you shone a laser through them. Obviously the offset by frequency for a well-corrected lens isn't all that big.

 

I believe there's a secondary effect: even if you have all frequencies in focus at the same point, that may have been achieved by cones of confusion of different angles (I think this isn't so unlikely, given the need to correct the split wavelengths). If all wavelengths aren't following the same path by the time they leave the rear of the lens, in theory it should be possible to get colour fringes around the bokeh either side of the focal plane anyway. In this case, the colours of the fringe should be symmetrical. This seems to be rarely seen, and in doodling it, it does appear that you'd see very small fringes this way. I'm curious whether there are lenses that correct for it, though.

 

In looking up what a printing Nikkor was, they do claim to be quite well colour corrected (and if they weren't, they'd be soft at some wavelengths). Exactly how well is another matter. By the time you've dropped to f/2.8, I'd have expected quite a few modern lenses to do well - the 70-200s aren't bad, I got the 150mm Sigma because it has very little LoCA (nor does the 200/2, but it's not exactly a close-up lens), etc. The Sigma 85mm sadly does have some LoCA, but not as bad as the Nikkors; I have a little envy of the new Canon, which seems very well corrected. The 55mm Otus certainly has some clever LoCA correction, although it's still got some at wide apertures. I simplistically assume that the more optical materials (with different dispersion) are involved, the more wavelengths you can correct, and that Canon's new material is therefore interesting.

 

If you want LoCA correction, there's the CoastalOpt 60mm. The Voigtlander 125mm APO-Lanthar is supposed to be very well-corrected too, although at least a while back the internet had noticed this and the prices went through the roof. The Laowa 105mm STF isn't bad, although the "STF" doesn't go as dark as I'd like (the Sony is better, but... LoCA).

[mike_halliwell]

Most LoCA, I believe, comes down to different wavelengths being focussed in different planes

Isn't that what APO lenses are supposed to fully correct for? .:D

 

Humm, seems the Sigma 150mm OS does a better CA correction..

 

Voigtlander SL 125mm f/2.5 Macro APO Lanthar Lens Test Review — Close-up Photography

[Andrew Garrard]

Isn't that what APO lenses are supposed to fully correct for? .:D

 

But then what would a superapochromat be for? :-) (I've a feeling that the astronomy has a slightly different definition for when things start to be achromatic, but then they only tend to have one focal length and the apertures usually aren't all that fast.)

 

Humm, seems the Sigma 150mm OS does a better CA correction..

 

So it does. (The popularity of the Voigtlander was, I think, around the time when photozone and Bjorn said nice things about it, and probably predate the updated Sigma.)

[rodeo_joe1]

Isn't that what APO lenses are supposed to fully correct for?

Briefly - No!

The 'apo' designation isn't rigidly defined, but is usually only applied to lateral colour aberrations. Lateral colour affects the plane of best focus, due to different light frequencies coming to different focii, or giving different magnifications or having differing distortion parameters or varying degrees of astigmatism - think colour-fringing that affects the edges and corners, but not the centre of the image.

 

Longitudinal colour - LoCa or Spherochromatism - OTOH only affects out-of-focus parts of the image. Therefore a lens designed to have a flat field and designed to copy a flat object, doesn't need to be highly corrected for LoCa, because no part of the subject will be rendered out of focus. It needs excellent lateral colour correction -hence the 'apo' label - but in theory its longitudinal colour correction will never be an issue; provided the lens is only used for flat copy and kept in focus.

[mike_halliwell]

RJ, how do you think stacking software is gonna ignore the nice magenta/green bits? They should be OOF and very soft at f2.8.

 

I sold my horendious 105mm VR or I'd try it

[Andrew Garrard]

Briefly - No!

The 'apo' designation isn't rigidly defined, but is usually only applied to lateral colour aberrations. Lateral colour affects the plane of best focus, due to different light frequencies coming to different focii, or giving different magnifications or having differing distortion parameters or varying degrees of astigmatism - think colour-fringing that affects the edges and corners, but not the centre of the image.

 

Really? The Wikipedia page talks mostly about longitudinal (axial) aberrations in the context of apochromatic lenses, and uses lateral as a synonym for transverse aberrations. By this definition, longitudinal chromatic aberration (what I call "LoCA") affects whether the image of different wavelengths from the same point in the subject fall at different distances from the lens, and lateral/transverse aberration ("LCA") affects the magnification of the image at different wavelengths even if everything is in focus. LCA is relatively fixable in software (to an extent), since you're just scaling the colour channels differently; the issue with LoCA is that you don't have the 3D light field information once the image is captured, so fixing it has to be heuristic. I don't actually know whether Canon do anything about this with their dual-pixel capture ability. LoCA is usually highly dependent on aperture, whereas LCA is usually less so.

 

My understanding is that a simple lens would focus all frequencies at a different focal length (if 560nm green were in focus on the focal plane, 450nm blue might be focussed in front of it, and 700nm red might focus behind it); dispersion in optical media causes the frequencies to bend by different amounts, and if you only have one kind of glass there's not much you can do to bend the light and still get everything back in line. If you only want to look at a single frequency, though, this can still be fine. An achromatic doublet uses two kinds of optical medium with different dispersion properties, which means you can put a "bend" in the relationship between focal plane and wavelength: red and blue may hit the same plane, but green would be somewhere else. A lot of photographic lenses meet this definition, and it's common to have a green fringe behind the subject and a purple fringe in front because of it. Simplistically, it's like having a quadratic equation relating frequency and focal distance, rather than a linear equation.

 

The terminology typically seems to be that an "apochromat" can get three wavelengths aligned (meaning a cubic equation, by the same analogy), and typically this also means that all visible frequencies are pretty close to the focal plane. I believe marketing gets involved, and some lenses that just have very good correction (for example because the total amount of dispersion is small) get an "APO" label even if they don't strictly line up three wavelengths. I suspect this means you need three different optical materials with different dispersions involved, although I'm guessing at that point and don't claim enough knowledge of optics.

 

If you have four or more wavelengths corrected to the focal plane, the term seems to be "super-achromat", although again I suspect marketing would get involved. It might be nice of "order-n corrected" were written somewhere, but that's simplifying all the other aberrations that need to be corrected.

 

Longitudinal colour - LoCa or Spherochromatism - OTOH only affects out-of-focus parts of the image. Therefore a lens designed to have a flat field and designed to copy a flat object, doesn't need to be highly corrected for LoCa, because no part of the subject will be rendered out of focus. It needs excellent lateral colour correction -hence the 'apo' label - but in theory its longitudinal colour correction will never be an issue; provided the lens is only used for flat copy and kept in focus.

 

Again, that's not how I've understood it. If you have the same point focus at different distances from the lens in red and green, green may form a sharp point, but red would form a circle of confusion. In other words, if you're trying to focus on a plane and you have an achromatic (but not apochromatic) lens, you can at best get two frequencies in focus, and everything else is a little blurry. If you want to copy a monochrome image illuminated by a laser, you don't need chromatic aberration correction; if you want to use white light, you do. A side effect of different frequencies hitting different focal planes is coloured fringes: if a point of light expands to a cone at the lens (well, the light expands in a sphere, but the bit hitting the front of the lens is a cone) and then that cone is brought back to a different point at each frequency, the point is larger when it's out of focus, and stays larger as you get more out of focus (probably more so - the cones of light have different angles), which is why you get "coloured bokeh"; this is usually way more visible than the amount by which light misses focus. Fortunately an easy way to limit the amount of visible LoCA is to limit the maximum angle of the light cones - in other words, pick a small aperture.

 

Copy lenses do tend to be designed to limit distortion and LCA, but you need everything corrected if you want to take perfect colour photos even of a flat field. Of course, you can also put a colour filter on the front of the lens and deal with one(ish) wavelength at a time.

 

RJ, how do you think stacking software is gonna ignore the nice magenta/green bits? They should be OOF and very soft at f2.8.

 

I sold my horendious 105mm VR or I'd try it

 

:-) I don't know whether focus stacking software does this, but in theory nothing stops you handling different light frequencies separately. Just pull the images into separate RGB channels and stack them independently. A sensor filter covers a wide range of wavelengths rather than a single one (which is why you can't fix LCA perfectly in post-processing), but it might have a positive effect. I've so far concluded that focus stacking is difficult (also I need a MacBook with more RAM), so I'll let an expert report back...

[mike_halliwell]

Doesn't seem too bad.....

 

Nikon Printing-Nikkor 105mm f/2.8, A version

 

I've so far concluded that focus stacking is difficult

Indeed. There's a table somewhere, but I can't find it at the mo', which show's which method of picture taking works best for what scale and degree of 3Dness it exhibits.

 

A lot seems to depend on the difference between moving the focused plane by helicoid with a fixed camera, moving the whole ensemble on rack/slide/focusing rail or just the camera via bellows or somesuch, with a fixed lens.

[Andrew Garrard]

Does seem to review well, if I understand it. Now if only I had that much spare money and the desire to photograph flat things that are exactly 36mm long... (Philately will get you nowhere, as the saying goes.)

 

I've long had and pretty much never used a Manfrotto macro rail. I recently picked up a Neewer one, which has two axes of control(ish) and Arca clamps, although I think it only looks like it's screw-thread driven. Seemed okay for the aphids a few weeks back, other than issues with the breeze, but it's pretty bulky. I gather cheap driven macro rails aren't all that expensive. Lacking an obvious third axis control, I also picked up a little lifting bench, so the next time I'm inside with a suitable subject, I ought to be able to center it more accurately. This was mostly fun during the last black Friday deals, btw.

 

It strikes me that since foreground points which are out of focus will intrude on the image of more distant ones, it ought to be possible to calculate this in an image stack and reverse it - effectively giving you an aperture shaped like a cone with cut-outs from the foreground. I've no idea whether that effect is either significant or worth it. It might be interesting if I ever find the need to photograph tea inside a tea bag. :-)

[rodeo_joe1]

Really? The Wikipedia page talks mostly about longitudinal (axial) aberrations in the context of apochromatic lenses, and uses lateral as a synonym for transverse aberrations. By this definition, longitudinal chromatic aberration (what I call "LoCA") affects whether the image of different wavelengths from the same point in the subject fall at different distances from the lens,

That's a very simplistic view of spherochromatism. As the name suggests, it's a second-order aberration that defines how the point spread of out-of-focus image points behave WRT their wavelength. I.e. whether the spread point has an even illumination, or is 'hot' centred or 'hot' edged. Obviously, if different colours behave differently, there will be some colour fringing in out-of-focus images. Regardless of whether the minimum point spread at all wavelengths is co-incident.

 

So the simple explanation; that an apochromatic lens brings 3 wavelengths to a single plane of focus, doesn't really tell the whole story. And there may even be zonal changes in correction where residual field curvature changes sign at different wavelengths.

 

Nothing is simple where lens design is concerned! That's why it's a job best left to powerful computers crunching complex algorithms.