Printing-Nikkor Adapter

[ChrisLilley]

By the way, what is this particular coin's story? (what details do you have regarding the coin)

I'm also curious - that's a very intriguing coin!

It is a cast bronze coin, or 'potin'. Gaulish, of the Remi tribe (from the area of France around the modern town of Reims).

The obverse is supposed to depict "three consuls". However, since the museum at Reims is full of statues of some sort of three-faced deity, I suspect that the coin referred to that.

There is a trace of a legend, RE[MI]. Gaulish coins were struck on flans that were smaller than the dies, so the full design does not show up on a single example and the design also tends to not be centred (this one is fairly well centred).

It dates to around 60 BC. (Gaulish and Gallo-Roman culture is a hobby of mine).

I haven't made a stack of it yet (or a decent photo really, need to get my lighting sorted out. The one posted was just a quick test in natural light when I first got the lens.

[Craig Gerard]

Thanks Chris,

Interesting reading.

Craig

[ray_parkhurst]

It is a cast bronze coin, or 'potin'. Gaulish, of the Remi tribe (from the area of France around the modern town of Reims).

Unlike your ancient from Gaul, most modern coins are pretty flat and unless the coin is not flat on the stage or the camera not mounted square on the stand I have not seen a strong need to stack. Critical sharpness is never at full aperture (at least not near the edges of the coin) so opening up the lens won't result in a better image when stacking is used. I've found that for most of the enlarger lenses I have tested the field is pretty flat at the "crossover point" where lens aberrations are about equal to diffraction effects, typically around f/7.1. Center sharpness may suffer a bit there but the edges are not much worse and depth of field is enough to give an overall equal level of resolution across the whole surface of a (fairly flat) coin. It's under these conditions that I've been comparing lenses to see which give the sharpest single-shot image.

[Oskar O]

So clearly, to me, stacking on this coin is merited and not a waste of time.

Agreed. However I would nitpick so much to point out that the exact flatness is not so easy to determine for the viewer, since the plane of focus doesn't seem to be parallel with the background and the right side of the coin more out of focus than the left. But I agree that stacking would benefit here

However, many coins are quite flat and wouldn't need stacking... Now that I think of it, I collected some coins as a kid, should get them out at some point...

[ChrisLilley]

However I would nitpick so much to point out that the exact flatness is not so easy to determine for the viewer, since the plane of focus doesn't seem to be parallel with the background and the right side of the coin more out of focus than the left.

That isn't nit picking. Its a correct observation.

That photo was taken as a quick test, to answer the question "does this lens work OK or does it have some flaw and I need to contact the seller about it". It wasn't taken to test coin flatness, I just happened to have the photo handy and felt it would contribute to the discussion.

Even so, the depth of the subject is clearly much greater than the acceptable DOF of the lens at optimum aperture (f/5.6) at this magnification (1x). With a circle of confusion of 16.5μm (3 * D90 pixel pitch) that comes out at 365μm.

However, many coins are quite flat and wouldn't need stacking... Now that I think of it, I collected some coins as a kid, should get them out at some point...

Yes, if the total depth of the front face is a third of a millimeter or less, it would not need stacking. If its greater, then there is a conscious choice between stacking to increase DOF, or selecting a smaller aperture to obtain more DOF, or increasing the circle of confusion (ie lowering the required quality) so that the subject all falls within the lower definition of "acceptably in focus". All of which are valid choices to make.

Note too that this depends on the optimum aperture for the lens, and also on the sensor. A lens which is optimum at f/8 for example would give 522μm. Change the D90 for a D3 or D700 with its larger, 8.45μm photosites and that becomes 803μm at f/8.

[ray_parkhurst]

Note too that this depends on the optimum aperture for the lens, and also on the sensor. A lens which is optimum at f/8 for example would give 522μm. Change the D90 for a D3 or D700 with its larger, 8.45μm photosites and that becomes 803μm at f/8.

But for the D3 or D700 the magnification must be increased to have same pixel count in the image for the larger pixels, and I believe this exactly compensates the increased DOF, doesn't it? At the same time the larger image circle causes more opportunity for field curvature and aberrations.

[ChrisLilley]

Note too that this depends on the optimum aperture for the lens, and also on the sensor. A lens which is optimum at f/8 for example would give 522μm. Change the D90 for a D3 or D700 with its larger, 8.45μm photosites and that becomes 803μm at f/8.

But for the D3 or D700 the magnification must be increased to have same pixel count in the image for the larger pixels, and I believe this exactly compensates the increased DOF, doesn't it? At the same time the larger image circle causes more opportunity for field curvature and aberrations.

Yes, this is true. To fill the sensor with the same sized subject, you need 1.5x the magnification.

The larger image circle (22mm instead of 15mm) isn't really a stress on the apo-rodagon-d, however, as it is designed for a 40mm radius circle.

[ray_parkhurst]

Yes, this is true. To fill the sensor with the same sized subject, you need 1.5x the magnification.

The larger image circle (22mm instead of 15mm) isn't really a stress on the apo-rodagon-d, however, as it is designed for a 40mm radius circle.

We'll see...I plan to buy one and compare flatness to the Printing-Nikkor and APO-EL-Nikkors. My 2nd 105AEN just arrived. It's the "new" model, much smaller than the older one. Supposedly it's corrected to a slightly longer wavelength. Coatings are a bit different. It's hard to tell the difference between them in final images, though the new one seems just a tad lower resolution.

[Oskar O]

Just remembered that there is also the repro nikkor 85/1, which is exotic and optimized for 1:1. Just sayin...

[ray_parkhurst]

Just remembered that there is also the repro nikkor 85/1, which is exotic and optimized for 1:1. Just sayin...

Talk about narrow depth of field...any idea how good this lens is?

[dickb]

Bjørn Rørslett likes his sample:

http://www.naturfotograf.com/lens_spec.html

I'd like to test this lens myself, but given the asking prices on eBay I doubt that will happen anyday soon..

[Oskar O]

Talk about narrow depth of field...any idea how good this lens is?

No, and there are very few users out there... see this http://homepage2.nifty.com/akiyanroom/r ... o1pon.html

Realistically though, I dont' think you'll really beat the printing Nikkor. If you truly want higher quality, moving to bigger formats should be a more efficient use of time and resources than trying to find something better than a printing Nikkor. You could of course consider stitching several images together.

[ray_parkhurst]

Realistically though, I dont' think you'll really beat the printing Nikkor. If you truly want higher quality, moving to bigger formats should be a more efficient use of time and resources than trying to find something better than a printing Nikkor. You could of course consider stitching several images together.

At this point I think the small differences I'm seeing between the exotics are proving this statement correct. What bothers me though is that even the Printing-Nikkor has a narrow range of aperture for critical resolution across the field, even at a field of 19mm (Lincoln Cent). This is telling me that when folks say the camera is the limiting factor in resolution it is not really correct. The less-exotic lenses I've been comparing don't even come close to providing full resolution at the 16MP/DX level, and the Printing-Nikkor and APO-EL-Nikkor are just at the edge of taking full advantage.

Would I do better with an FX or large-format camera? If the pixels are the same size, just with more of them, my conclusion would be "no". Going for high mag to fill the sensor would expand the image circle and cause more problems at the edges. In other words, looking at the image at 100% would be worse for the larger sensor. If the pixels are larger, I would say "maybe" if the improvement due to larger pixels (lower lp/mm) outweighs the degradation due to the larger image circle. Does anyone know which would win? It seems to me that the optimum solution is a very small, high pixel count sensor that can take advantage of the higher lens resolution at the center of the image circle.

[Oskar O]

At this point I think the small differences I'm seeing between the exotics are proving this statement correct. What bothers me though is that even the Printing-Nikkor has a narrow range of aperture for critical resolution across the field, even at a field of 19mm (Lincoln Cent). This is telling me that when folks say the camera is the limiting factor in resolution it is not really correct. The less-exotic lenses I've been comparing don't even come close to providing full resolution at the 16MP/DX level, and the Printing-Nikkor and APO-EL-Nikkor are just at the edge of taking full advantage.

You have to consider MTF and thus local contrast too (if you're unfamiliar with it, there's a nice article at Zeiss' site about it...I could dig up a link if needed). Resolution is usually quoted for where MTF is 10% or MTF 50%, but with a high MTF you'll get the best separation of detail with minimum noise. Resolution at MTF-10 for a lens might be high, but if the MTF curve is not overall high, then you'll have to increase contrast to bring out the detail and that will increase noise.

Would I do better with an FX or large-format camera? If the pixels are the same size, just with more of them, my conclusion would be "no". Going for high mag to fill the sensor would expand the image circle and cause more problems at the edges. In other words, looking at the image at 100% would be worse for the larger sensor. If the pixels are larger, I would say "maybe" if the improvement due to larger pixels (lower lp/mm) outweighs the degradation due to the larger image circle. Does anyone know which would win? It seems to me that the optimum solution is a very small, high pixel count sensor that can take advantage of the higher lens resolution at the center of the image circle.

With larger pixels, you'll be emphasizing the lower frequencies of the formed image, which (in good lenses) have higher MTF anyway. My practical experience is that there are superb lenses that cover 6x6 or more (such as the Apo Rodagon D mentioned), so the large format should win by sheer size. My Schneider Macro Symmar is rated for 12k resolution at optimal magnification and it covers roughly 6x6, so that should in theory give a 144 megapixel image of rpetty good quality (the quality on a 12 mpix APS-C sensor is pretty good).

The only caveat is that I haven't really tested this in practice, since I haven't bought into digital medium format (yet). But it seems to work out going from APS-C to full frame, medium format images with select lenses on film look good and the theory is sound so...

[ray_parkhurst]

Talk about narrow depth of field...any idea how good this lens is?

No, and there are very few users out there... see this http://homepage2.nifty.com/akiyanroom/r ... o1pon.html

Well, I tested an 85mm Repro-Nikkor and found it a very odd lens. It's NOT a camera lens! It wants to be mounted with no extension whatsoever beyond the lens mount, and in fact placement against the lens mount was too far from the sensor to achieve nominally 1:1. There would be no way to set it up for less than 1:1 magnification without devising a mounting scheme that threads the lens into the camera body. The documentation available for this lens states that the magnification range is 0.9 to 1.1, which seems to be set by practical considerations of glass-to-sensor distance and glass-to-object working distance. It is just not possible to set the lens up to have magnification outside this range. Being a symmetric design, working distance from lens to subject was <<2 inches with the lens against the camera lens mount.

The few images I was able to take show a fair amount of CA. This is NOT an APO lens. I did an aperture sweep and found pretty much what I expected, optimum in the center is around f/5.6 and the edges improve at f/8.

It's a beautiful lens but the narrow magnification range, large diameter, and short working distance make it extremely impractical. It also has large and odd mounting threads.