Macro lenses for a digital back

[shlomibe]

Hi,


I have a digital back (56x36mm) and I would like to use it to shoot products - tabletop and macro.

I have a Schneider Digitar 120 N lens which should cover tabletop.

But the question is what to do about macro: 1:2 - 1:1.

I would like to close aperture as much as possible - Schneider engineers told me that they approve aperture of up to 8 in 1:1, but of course I need more. So I would like the lens which handles diffraction best.

Regarding the question of digital or non digital lenses - some people tell me there is a big difference, other people say the old lenses are just fine for a digital back.

The lenses I am thinking of are:

Schneider Digitar 120 M
Schneider Macro Symmar HM
Rodenstock Apo Macro Sironar Digital 120
Rodenstock Apo Macro Sironar 120
Nikon 120 AM ED Macro

Any opinion or experience would be much appreciated.

[Brian Ellis]

I don't think many people here use a digital back, in fact some don't even use a digital camera. You'd probably get more and better responses if you posted these questions in a forum more oriented to digital cameras/backs than this one is.

[Bob Salomon]

The digital macro lenses will perform best on digital backs and roll film.

Non-digital lenses do not deliver the best results as they are designed for film at smaller apertures then digital lenses, have lower resolution, greater distortion and their color curves are not as good. Additionally analog lenses were made for a medium that sags, digital sensors do not sag. Film has several layers of emulsion, digital is one pixel thick. Lastly the light rays that pass through a digital lens strike the pixels on a digital sensor differently then the rays through an analog lens so the edges of the pixels are more evenly illuminated with a digital lens.

Also bear in mind, the analog lenses that you listed reach optimal aperture at f22 while that small an aperture will be well into diffraction on your size sensor. That is why the digital lenses reach optimal aprerture at f8 to 11. So using film lenses at optimal apertures on digital backs will always lead to degraded images due to diffraction.

The Rodenstock Apo Macro Digital is corrected for 1:5 to 2:1 which menas that this is the ideal lens for table top to 2:1 and will outperform what you are using now.

[Oren Grad]

TAlso bear in mind, the analog lenses that you listed reach optimal aperture at f22 while that small an aperture will be well into diffraction on your size sensor. That is why the digital lenses reach optimal aprerture at f8 to 11. So using film lenses at optimal apertures on digital backs will always lead to degraded images due to diffraction.

That's right, but Shlomi has already told us that he knows he's going to need to go there to achieve the required depth, and he's asking whether there's any difference in how lens performance degrades when the situation forces you to stop down past optimal aperture.

[Bob Salomon]

That's right, but Shlomi has already told us that he knows he's going to need to go there to achieve the required depth, and he's asking whether there's any difference in how lens performance degrades when the situation forces you to stop down past optimal aperture.

Yes, he will be well into diffraction on that format. That is why digital lenses, including macro digital lenses, are optimized for f8 to 11 rather then f22.

[shlomibe]

That's right, but Shlomi has already told us that he knows he's going to need to go there to achieve the required depth, and he's asking whether there's any difference in how lens performance degrades when the situation forces you to stop down past optimal aperture.

Pretty much.
More accurately - I know there's a difference (in how different MF/SF lenses handles the same apertures.)
I want to know which current or past LF lens has the best diffraction resistance - from user experience, not from manufacturer data, which I have already read.

I know some MF lenses have very good diffraction resistance, and I would like to compare them to the best ones from LF.
In the past I've tried Digitar M, and it was not any better than the comparable Mamiya.

The answer: "best sharpness is in f/8" does not answer the question: "which lens has the best sharpness at f/22?"
Schneider told me: "all lenses are crap in f/22", which in my experience is not true.

BTW I will take film experience as well, if someone tried more than one lens and minded the diffraction performance.

[Bob Salomon]

Pretty much.
More accurately - I know there's a difference (in how different MF/SF lenses handles the same apertures.)
I want to know which current or past LF lens has the best diffraction resistance - from user experience, not from manufacturer data, which I have already read.

I know some MF lenses have very good diffraction resistance, and I would like to compare them to the best ones from LF.
In the past I've tried Digitar M, and it was not any better than the comparable Mamiya.

The answer: "best sharpness is in f/8" does not answer the question: "which lens has the best sharpness at f/22?"
Schneider told me: "all lenses are crap in f/22", which in my experience is not true.

BTW I will take film experience as well, if someone tried more than one lens and minded the diffraction performance.

Apo Sironar-S

Back in the early 80s or late 70s Modern Photography devoted an issue to comparing lenses from all medium format cameras and also included a Super Technika V, 23b with 3 Rodenstock lenses; 65mm Grandagon-N 4.5, Sironar N MC 105mm and 180mm.
Compared to all other MF cameras; Hasselblad, Mamiya, Pentax, Rollei SL66 and 6006, etc. The lenses on the Linhof outperformed them all. In many cases significantly. The Apo Sironar-S will be even better.

[shlomibe]

Thank you.

I am seeing "Rodenstock 100mm f/5.6 Apo-Sironar-S", but not a macro.

Is there one?


Regarding what you said - Hasselblad and Mamiya have made tremendous progress in the past 15 years in the quality of their glass, and specifically sharpness at all apertures, so I'm not sure the results from the 1970's are relevant.

[Bob Salomon]

Thank you.

I am seeing "Rodenstock 100mm f/5.6 Apo-Sironar-S", but not a macro.

Is there one?


Regarding what you said - Hasselblad and Mamiya have made tremendous progress in the past 15 years in the quality of their glass, and specifically sharpness at all apertures, so I'm not sure the results from the 1970's are relevant.
No they would not be relevant. But there has been no other article or test like it since that compared test results like this in a U S publication. Color Foto Magazine in Germany did one called Sharp Shooting several years ago. The results were very much the same.

[Bob Salomon]

No, the Rodenstock analog macro lenses are the 120 Apo Macro Sironar and the 180 Apo Macro Sironar. These replaced the earlier 210 Makro Sironar and the 300 Makro Sironar several years ago. All would be greatly inferior optically on your back to the 120 Apo Macro Sironar Digital.

[Nathan Potter]

If I were really serious about highest quality with a digital back for macro, tabletop photography I'd use a digitar lens optimized for parallelism of exit rays and at a maximum of about f/8, preferably less. I'd get the required DOF by using some of the newer focus stacking software and hardware.

Nate Potter, Austin TX.

[shlomibe]

When you do a 500 item catalog, it's not always possible to make every image from 100 captures.

At 5.6 the result would be the sharpest, but it would take a lot of time even with an electronic rail.

[Nathan Potter]

Indeed, you'd have to automate the distance scan using a stepper motor and sync each advance to a shutter actuation. 100 actuations could be done in less than 100 seconds per image. However, a 500 item catalogue would eat up 50,000 actuations on the DSLR. I suspect it takes longer to do the setups. Such DOF may be irrelevant for your purposes though.

Nate Potter, Austin TX.

[Bob Salomon]

When you do a 500 item catalog, it's not always possible to make every image from 100 captures.

At 5.6 the result would be the sharpest, but it would take a lot of time even with an electronic rail.

None of the analog or digital lenses will be sharpest at 5.6, nor would they have their maximum coverage.
Assuming the camera is designed for digital so it can make the small movements digital requires and that the proper exposure is used the sharpest results will be at the diffraction limited f stops. F8 to 11 on most digital lenses and F22 on most analog 45 lenses.

[shlomibe]

None of the analog or digital lenses will be sharpest at 5.6, nor would they have their maximum coverage.
Assuming the camera is designed for digital so it can make the small movements digital requires and that the proper exposure is used the sharpest results will be at the diffraction limited f stops. F8 to 11 on most digital lenses and F22 on most analog 45 lenses.

Schneider say explicitly their best sharpness is at 5.6.

[Bob Salomon]

Schneider say explicitly their best sharpness is at 5.6.

Ask them again.

[darrelll]

Pretty much.
More accurately - I know there's a difference (in how different MF/SF lenses handles the same apertures.)
I want to know which current or past LF lens has the best diffraction resistance - from user experience, not from manufacturer data, which I have already read.

I know some MF lenses have very good diffraction resistance, and I would like to compare them to the best ones from LF.
In the past I've tried Digitar M, and it was not any better than the comparable Mamiya.

The answer: "best sharpness is in f/8" does not answer the question: "which lens has the best sharpness at f/22?"
Schneider told me: "all lenses are crap in f/22", which in my experience is not true.

BTW I will take film experience as well, if someone tried more than one lens and minded the diffraction performance.

Well, if you are not going to take the word of the Schneider engineers... why would you take the word of any of the members here? You are just going to have to conduct your own tests since you are not going to take the word of anyone here or at Schneider.

[Nathan Potter]

Specifically at f/22 all the lenses the OP mentions will show about the same degree of diffraction which will be around 30µm. Any slight variation due to a particular optical design will be insignificant IMHO. More variation in resolution between the candidates mentioned will be found at the larger apertures where the optical arrangements (formulas) leads to more variation in the degree correction of the image degradation parameters.

You don't need to do any lens testing to know the situation at f/22 insofar as diffraction is concerned.

Nate Potter, Austin TX.

[shlomibe]

Well, if you are not going to take the word of the Schneider engineers... why would you take the word of any of the members here? You are just going to have to conduct your own tests since you are not going to take the word of anyone here or at Schneider.

Schneider engineers have their commercial agenda.
They've made a decision to optimize their digital lenses to f/5.6, and therefore have no interest to help me find the best lens for f/22.
IMO their decision is aimed at amateurs and pixel peepers, and they aim to create lenses which will show the best possible sharpness, which due to diffraction can be achieved at f/5.6. It is useful for professional photographers - not very much.

Regarding conducting my own tests - I would need 2-4 of the pertinent pretty expensive non refundable lenses at my disposal, which is not easy to get.
At this point, considering the feedback and materials I got, I've arrived at the conclusion, based on impressions not on concrete tests, that MFD lenses are the best solution for me.

Regarding "why would you take the word of any of the members here" - the reason is that forum members mostly don't have a commercial agenda.

[paulr]

Schneider's MTF charts show their digital lenses performing best at f8 or f11. I haven't seen any that were better at 5.6.

I think you'll see some differences at f22, but they may not be differences worth paying for. At least not as far as sharpness is concerned. The digital lenses are mostly diffraction limited at this point. The better macro analog lenses will be close behind. At wider apertures the analog lenses can't compete.

If you're looking for people with experience with these lenses on digital backs, try the getdpi.com forum.

[shlomibe]

Specifically at f/22 all the lenses the OP mentions will show about the same degree of diffraction which will be around 30µm. Any slight variation due to a particular optical design will be insignificant IMHO. More variation in resolution between the candidates mentioned will be found at the larger apertures where the optical arrangements (formulas) leads to more variation in the degree correction of the image degradation parameters.

You don't need to do any lens testing to know the situation at f/22 insofar as diffraction is concerned.

Nate Potter, Austin TX.


In my actual experience there is very considerable difference in diffraction resistance between different lenses.

If Schneider bother to mention their Digitars are optimized for f5.6, and Rodenstock bother to mention their Sironar Digitals are optimized for f/8-11, then I would say they also believe that specific lenses are optimized for specific apertures.

[paulr]

I'm not sure what you mean by diffraction resistance. There is no difference in diffraction from one lens design to another; diffraction is a basic physics phenomenon of light passing over an edge. It's determined by aperture and nothing else. Possibly you're seeing a difference in optical aberrations between lenses. If a lens has prominant enough aberations, they will be visible alongside diffraction even at relatively small apertures.

[GG12]

The problem, at least as this poor mind understands it, is that the digital back raises the issue of the circle of confusion, which has to do with the size of the resolving component (the pixel dimension on the sensor, different than film grain). The smaller the pixel, the smaller this circle has to be to maintain sharpness. With the newest type backs (60, 80 mp), diffraction is a serious problem. and thus lenses are re-designed to optimize at f 5.6-f8, or 11 max. On an older back with larger pixels (40 mp or less), f 16 is still usable. Sounds like the shooter has a newer type back. There has been a lot of discussion about this on getdpi or LL forums.

[rdenney]

I'm not sure that one lens is more diffraction resistant than another, at least in the center of the image. Diffraction is affected by the optics, but it is caused by the edges of the diaphragm itself.

What Bob didn't mention is that in return for the additional performance of the digital lenses, they also have shorter focal lengths and much smaller image circles. Not a problem for smaller sensors, of course. But diffraction will affect them same as any other lens.

You might need to try something like focus-stacking if you need all the performance that those lenses offer at f/8, but the depth of field provided by f/22.

Rick "noting a lot of questions about macro these days" Denney

[rdenney]

More: Generally, lens aberrations reduce as the lens is stopped down, but diffraction increases. A lens is not "optimized" for f/8. It just behaves optimally at f/8, because that's where the lens faults are well enough corrected that the effects of diffraction cause degradation at smaller apertures.

So, a lens that performs at its best at f/8 (or f/5.6) would be a very good lens indeed. It will also perform as well or better at f/22 than a lens that achieves its optimum at f/22.

Remember, though, that large-format lenses are often optimal at f/16 or f/22 at the edges of the specified coverage circle. The performance will be better at the center, and indeed the lens may behave optimally at wider apertures if you are just using the center of the frame.

Catalogs have been photographed at high quality for decades using film the size of this sensor, and lenses that are not the latest and greatest designs optimized for digital sensors. Except for very short lenses with respect to the format, those lenses will perform as well on the digital sensor as they will on film. The digital lenses will perform better, especially at wider apertures.

Another thing to remember is that the better performance is usually measured in terms of higher MTF. If the low-MTF resolution is sufficient to your needs, you may be able to increase the apparent MTF using sharpening tools.

But all this for a catalog? The enlargement for an 11x17" full spread is roughly 5x. And the 133-line screen (for high-quality process color) will cut off micro-resolution in any case (though, at the same time, the dot pattern will increase apparent sharpness at the bright end of the histogram by creating false edges). I would think a lens that can put maybe 35 or 40 lines/mm at reasonable MTF on the sensor will achieve all that is possible for the biggest possible image printable in a catalog. Most catalog pictures are much smaller and consequently much less demanding.

Rick "who has done a little photography for moderate-quality process-color catalogs in the deep past with a lowly Mamiya C3, the normal 80mm lens, and a paramender" Denney

[Ken Lee]

With apologies if this has been asked before or is otherwise obvious: what about using a "digital" lens on film ? Would the curvature and inherent imprecision of film placement - or the thickness of the film itself - negate any benefits ?

[Nathan Potter]

With apologies if this has been asked before or is otherwise obvious: what about using a "digital" lens on film ? Would the curvature and inherent imprecision of film placement - or the thickness of the film itself - negate any benefits ?

Ken that seems to be a good question but I can see the answer being quite complex and very dependent on the lens under consideration.

As I understand it many "digitar" type lenses are designed to have exit rays somewhat more orthogonal to the sensor (film) than normal lenses. There are a couple of reasons for doing this when using a sensor.

First, off axis rays are scattered by the dead real estate between the pixels (that area containing the wiring traces and other components). Such scattering causes a variety of distortions especially chromatic effects. This is really serious with wide angle lenses at the periphery of the sensor. Ideally parallel rays create less crosstalk between pixels.

Second, the active area of the sensor pixel functions better when receiving rays more orthogonal to the surface of the silicon - that is the absorption depth of a photon in the silicon bulk is more uniform across the sensor surface.

IMO there would be an advantage (small?) in using digital lenses with film because they would be slightly less sensitive to variations in the film plane due to the slightly less off axis rays hitting the film.

All of this assumes that the inherent resolution of the digital lens is no better than the best non digital equivalent. But I think such an assumption is not valid for all digital lenses.

Nate Potter, Austin TX.

[Bob Salomon]

With apologies if this has been asked before or is otherwise obvious: what about using a "digital" lens on film ? Would the curvature and inherent imprecision of film placement - or the thickness of the film itself - negate any benefits ?

Ken,

Digital lenses, at least those from Rodenstock, are outstanding performers on roll film formats that they cover. BUT, with the Apo Sironar Digital HR and their replacements, the HR Digaron S the last element in the lens is the cover glass over the sensor. So if you are using those lenses then you need to install a special "Corrector Plate" that screws into the rear of the lens for film.
The other rodenstock digital lenses, Apo Sironar Digital, HR Digaron-W, HR Digaron-S/W and the Apo Macro Sironar Digital do not use a corrector plate with film.

Also note, while many of these digital lenses only cover 70 or 90mm circles at infinity many of them do cover 125 and 150mm circles.