The GetDPI Photography Forum

Great to see you here. Join our insightful photographic forum today and start tapping into a huge wealth of photographic knowledge. Completing our simple registration process will allow you to gain access to exclusive content, add your own topics and posts, share your work and connect with other members through your own private inbox! And don’t forget to say hi!

Schneider-Kreuznach APO-Digitar 35mm f/5.6 L-88° on GFX

rdeloe

Well-known member
Schneider-Kreuznach sold the APO-Digitar 35mm f/5.6 in various housings, and in two models, the well known one being the later XL-102° in a Copal 0 shutter. Much less common is the L-88° version in a Schneider-Kreuznach electronic shutter that is screwed to a lens board for the Plaubel digital view camera system that was developed by Jenoptik, Plaubel and Schneider-Kreuznach and announced at Photokina in 2002.

UPDATE #3: The thread started with the assumption that the L-88 version and the XL-102 version were distinct. Then we put together evidence that suggested they might actually be the same. And then I found the data sheet for the L-88, and thanks to Rod's analysis, it's now clear that they are different optical formulations. To cut to the chase, jump ahead to this post in the thread and read from there: https://www.getdpi.com/forum/index....tar-35mm-f-5-6-l-88°-on-gfx.75891/post-904739 Long story short, the L-88 is close but not the same. It has a smaller image circle than the successor XL-102. However, the price for the larger image circle may be a very small reduction in sharpness in the XL-102, based on comparing the MTF charts posted below. The L-88 is still an excellent lens if you can get by with an image circle of 70mm.

Plaubel ad.jpg
Image from a Schneider-Kreuznach ad; thanks to Rod (@4x5Australian) for sharing.

There’s a good chance you didn’t know about the L-88 version because it’s not common, and doesn’t figure prominently in documentation. Everything I know about it I learned from Rod (@4x5Australian), who filled me in after I bought one thinking it was an XL.

The seller said it was an L-88° version in the ad; alas, the significance of the labelling simply didn’t register because I didn’t know there were two versions. Why does this matter? The L-88° version has a 70mm image circle, while the XL-102° that is beloved by many on this forum has a 90mm image circle.

The L-88 version is not a good choice for people using a medium format back that can take advantage of the 90mm image circle. However, on my F-Universalis plus GFX setup, I can only shift a 35 L or XL 8mm because the rear of the lens is inside the GFX mount. An 8mm shift needs a ~68mm image circle, so I couldn’t use the rest of the XL image circle anyway. That took the sting away from discovering that I had not purchased the lens I thought I had purchased.

The flange distance of this lens is very short, so the standards are close together. Nonetheless, there’s plenty of room for around 5 degrees of tilt and swing before the recessed lens board hits the frame. This set of images shows the lens centred at infinity (left), tilted as far as it will go (middle) and shifted as far as it will go (right). A copy of the XL I tried seems to have about the same flange distance, so it offers no advantage on a GFX and F-Universalis with Rotafoot setup when it comes to shift, tilt and swing.

Lens pics 2.jpg

The price was right for this copy because it’s an electronic shutter on a lens board for a digital view camera that doesn’t seem to have sold many copies. My plan was to remount the cells in a Copal 0 shutter. This is always a risky proposition because these wide technical lenses need to be adjusted properly for best results. I like a challenge, so I took the chance.

To remount it, I needed a shutter. The XL version is often found in a standard Copal 0 shutter. I didn’t want a Copal 0 because with a 35mm lens, the shutter cocking lever and the aperture overlap, making it fiddly to change the aperture. I chose a late Compur 0 shutter because the controls are spaced better. Fitting the front cell onto this Compur 0 shutter required removing a decorative plastic ring that surrounds the centre of the shutter.
Lens pics 1.jpg

#0 shutters are supposed to have a distance of 20mm between the mounting surfaces of the lens cells, but the design allows for -/+ 0.025mm on either side of that value. Some #0 shutters seem to exceed that tolerance. For example, my copy of the lens in a Schneider electronic shutter came with three shims that added 0.21mm. This was too much spacing for the donor shutter I used (a Compur 0 set up for a 100mm lens). Fortunately, a single 0.09mm shim produced optimum results wide open.

Unfortunately, there was an unexpected catch to closing up the spacing. With a single 0.09mm shim, the rear surface of the front cell is in the same plane as the shutter, meaning the shutter would not close. I don’t use the shutter in the Compur 0 housing, so this is not a problem for me. In my use, the shutter is locked open with the lever on the housing, and it stays open. However, someone who wanted to use this remounted lens with the shutter in the Compur 0 mount would not be able to do so.

The final step in this remounting process was calculating which of the positions on the scale that came with the lens was f/5.6. In the spirit of “close enough is good enough”, I did this by using my Pentax-A 645 35/3.5 to figure out the shutter speed at f/5.6 against a blank wall, switched to the 35 L and moved the aperture lever on the Compur 0 shutter until I matched the speed. The “temporary” label you see in the picture, above, is likely to remain on the shutter until it wears off… or I find a way to make a professional looking aperture scale.
 
Last edited:

rdeloe

Well-known member
I actually have a good 35mm lens already – the SMC Pentax-A 645 35mm f/3.5. I would not fault you for asking why I went to all this trouble to get another 35mm lens that has significant constraints on movements!

The main reason is that the Pentax-A 35/3.5, like other wide angle retrofocus lenses, suffers from strong distortion that takes a moustache form. I’m not an architectural photographer, so most of the time I don’t notice. But I do like symmetrical lenses that are free of distortion.

This pair of pictures tells the story. These are f/11 with 8mm of rise, Pentax then Schneider-Kreuznach. The distortion in the façade of this building in the Pentax image is complex and fixable only with a lot of manual work. In contrast, the APO-Digitar 35/5.6 L version is good to go right out of the camera.

SMC Pentax-A 645 35mm f-3.5 EV3 Facade.jpg

Schneider-Kreuznach APO-Digitar 35mm f-5.6 L EV3 Facade.jpg

I was also hoping to get much improved image quality, especially when shifting within the 8mm the lens can manage. Conventional wisdom seems to be that the APO-Digitar should easily outperform the old Pentax-A 645 medium format lens. In some ways it does, but – surprisingly – in most respects it doesn’t.

Here are three sets of images with 8mm of rise. All images were edited very lightly for tone, and I used flat field correction on the Schneider-Kreuznach images to deal with the strong light falloff. Don't get too excited about the processing. More work is needed to get these looking satisfactory! Full resolution versions of these images, and the two above, are available here: https://drive.google.com/drive/folders/1MOSXQy42rUqyqSlxQG3QzUP8JOuGvXh0?usp=sharing

Sample 4.jpg

Sample 2.jpg

Sample 3.jpg

I’m curious to know what other people see in these. Here are some takeaways for me, in no particular order:

* The Pentax is a bit wider. If the Pentax is actually 35mm, I’d say the Schneider-Kreuznach is at least 36mm. UPDATE: Data sheet says 36.4mm, which fits what I'm seeing.

* In terms of ability to resolve detail in the unshifted zone, the two lenses are very close. The Pentax looks a bit sharper in some scenes, while the Schneider-Kreuznach looks sharper in others. Differences could easily be due to slightly different points of focus.

* I recall one forum post where someone said the L version is of lower image quality than the XL. I’m not so sure. Prior to buying this L-88 copy, I had access to an APO-Digitar 35mm f/5.6 XL-102, and shot it alongside the Pentax-A 645 35/3.6. It was not significantly sharper than the Pentax. But see caveat regarding shifting, below.

* The Pentax shifts amazingly well. In some cases it looks better than the Schneider-Kreuznach in the shifted area; in others the Schneider-Kreuznach looks better. I think it’s a wash. This was surprising. Importantly, the Pentax can shift 12.5mm (15mm if you’re really pushing it). But the image quality is degraded with larger shifts, and the distortion is much worse towards the edge of the image circle on the Pentax.

* Because the image circle is only 70mm on the L-88, an 8mm shift is right up to the edge. Image quality is normally worse at the edge, and I think this is the case in these images. This is where the XL would be genuinely better; an 8mm shift on an XL is a long way from the edge of the 90mm image circle. UPDATE: See below.

* The APO-Digitar 35mm f/5.6 L-88 really is an apo-chromatic lens. There’s virtually no purple fringing, whereas the Pentax shows moderate to strong fringing in places. It’s fixable, most of the time. But it’s nice to not have to fix it.

There are some other differences between the two lenses that relate to functionality rather than image quality.

* Comparing the weight of all the bits and pieces needed to use the Pentax-A 35/3.5 on an F-Universalis (lens, lens caps, lens board, hood) to the Schneider-Kreuznach (lens, caps, board, hood), the Pentax-A 35/3.5 option is almost 500 grams heavier than the APO-Digitar 35/5.6. My back appreciates the weight reduction.

* Setting up and using the APO-Digitar is a bit faster. The Pentax-A 35/3.5 uses a close focusing design, which demands the lens is positioned precisely at its flange distance. And it has to be focused by the lens. I much prefer focusing by rail on the F-Universalis.

* An LCC frame and flat field correction are essential with the APO-Digitar 35/5.6. If I ever forget to shoot one, I can fix the image with gradients, but it’s much better to use an LCC frame. I don’t have a centre filter and won’t get one because it’s not necessary. I thought using LCC frames would be a major hassle, but it’s not.


It takes a long time to really get to know a new lens, so these are preliminary observations. However, I’m looking forward to getting to know this new lens much better and plant to use it a lot during this upcoming field season.
 
Last edited:

4x5Australian

Well-known member
Rob, that's a nice write-up.

Your method of establishing the aperture scale is a nifty one. To date, l've measured the aperture using a millimetre rule but it's not easy; your method sounds more reliable.

Rod
 

rdeloe

Well-known member
Rob, that's a nice write-up.

Your method of establishing the aperture scale is a nifty one. To date, l've measured the aperture using a millimetre rule but it's not easy; your method sounds more reliable.

Rod
Thanks Rod. However, it's based on the assumption that the aperture in the Pentax is correct! Robert O'Toole (he of the excellent Closeup Photography site) explained how to do it properly, but it's a bit of a process. If my scale is off, it's not off by much, so "close enough is good enough" really does work for me.
 

4x5Australian

Well-known member
Some background for readers:

Schneider Kreuznach (SK) introduced the focal lengths of 24mm and 35mm into its Apo-Digitar line at the Photokina trade show of September 2002. Previously, its widest lens was the Apo-Digitar 47 / 47 XL-100°. The 24mm introduced was the 24 XL-100° and the 35mm was the 35 L-88°.

The 24 XL-100° continued in the line for some time, however, by about August-September 2004 (* see note below) the 35 L-88° was replaced by the 35 XL-102°.

For the 35 XL-102°, the SK technical brochure gives the effective focal length as 36.4mm. Perhaps the focal length of the 35 L-88° is the same or similar.

For the 35 XL-102°, the SK technical brochure gives the field of view as 88° at its maximum aperture of f/5.6 and 102° at f/11. According to the SK data table, those angles of view provide image circles of 70mm and 90mm, respectively.

*The technical datasheet for the 35 XL-102° appears to be dated 26.08.04 and the SK technical brochure PDF for digital lenses, which includes the 35 XL, was created on 17.09.2004.

EDIT: Following Rob’s subsequent comment, below, I have deleted a sentence from here in relation to his provisional finding of a 70mm image circle, to avoid confusion.


Rod
 
Last edited:

rdeloe

Well-known member
Rob's post appears to be the first publicly available confirmation that the image circle of the 35 L-88° at f/11 is indeed the 70mm specified by SK. .
Let's hold that thought... The certainty implied by your statement triggered all my "check double-check triple-checked" instincts. I confirmed the 70mm circle this morning using a technique I've used for large format lenses, but adapted to this setup. This is an important question so I want to verify using another technique. Stay tuned.
 

rdeloe

Well-known member
While I'm looking for another empirical way to measure image circle, here's a mystery I could use help with.

Image circle diameter = 2 * f * tan((total angle of view) /2)

To make the formula work, you have to convert total angle of view in degrees to radians.
  • 88 degrees is 1.553589 radians. Assuming the focal length (f) is actually 36.4mm, then the image circle is 70.3mm using this formula
  • 102 degrees is 1.78024 radians. Again, with f=36.4mm, the image circle at 102 degrees is 89.9mm.
These values correspond almost exactly with the published data for the XL version of the lens. So far so good.

So now the mystery... forget the 70mm I thought I measured this morning. I've already confirmed that the method I used is not reliable. If we do it just by the formula, is the 88 degrees printed on the lens barrel for my L version referring to the total angle of view at f/11 or at f/5.6? For the XL, it seems that 102 degrees is for f/11. If the total angle of view of the L at f/11 is 88 degrees, then it would have to be a lot less at f/5.6 on the L, right? That would mean I could shift 8mm at f/11, but not at f/5.6, because I need a 68.3mm image circle to shift 8mm. Except I can shift 8mm at f/5.6 without hitting the edge of the image circle.

Do you see where I'm going with this? If this lens design has a smaller total angle of view at f/5.6 than at f/11, which Schneider says the XL does, and if the L has a total angle of view of 88 degrees at f/11 (and correspondingly smaller at f/5.6), then there is no way I should be able to shift 8mm on the long edge at f/5.6 and not see the edge of the image circle.

Therefore, I'm now strongly suspicious that the L-88 is the same lens as the XL-102, and that the difference is the marketing people said, "Why did you pick the small number, 88 degrees, when you could have picked the large number, 102 degrees?? Rebrand this thing as an XL-102!"

What do you think?
 
Last edited:

4x5Australian

Well-known member
Rob, I wondered about that possibility, too. It would be the simple explanation for the lack of any reference to the 35 L in the SK technical brochures that we find online.

SK’s technical brochures give the angles of view for lenses in the Apo-Digitar line at their maximum aperture and again for f/11. For some lenses those two values are the same, while for others, such as the 35 XL, the value at f/11 is larger. On the lens barrel itself, SK routinely used the value at f/11.

Why the 35 L-88° was labelled the way it was is still unclear. Perhaps the decision to relabel it in 2004 as the 35 XL-102° was a standardisation or correction. Or perhaps there really was some modification made to the optical formula.

One obvious test is to compare the maximum shift that the two versions can sustain with good image quality. Another test would be to measure the exact FFD, which is less simple to measure accurately. My 35 XL can be shifted at least 15mm (with the IQ4 in landscape orientation) with excellent results. As we know now, achieving optimal image quality involves having a Copal 0 (or equivalent) that was built or shimmed well to the 20mm depth specification, and it seems that many weren’t.

Rod
 
Last edited:

rdeloe

Well-known member
To put a bit more math on the bones of this argument, if the image circle is 88 degrees at f/11, then it's 74 degrees at f/5.6. A 74 degree total angle of view on this lens means an image circle of 54.9mm, just barely enough to cover a GFX sensor. An 8mm rise on the long axis would be impossible. And yet here we are: 8mm of rise on the long axis at f/5.6...

R. de Loe GFXB4382.jpg

Image quality at the top of the building is crappy at f/5.6...

Top.jpg

But that's totally expected given what the MTF chart says (and the fact that we're at the extreme edge of a 70mm image circle in this part of the frame). I added the red and orange lines to Schneider's chart. The red line is the circle that covers the GFX sensor, while the orange line is the circle that covers an 8mm shift on GFX.

MTF for 35 XL.jpg
 

rdeloe

Well-known member
Rob, I wondered about that possibility, too. It would be the simple explanation for the lack of any reference to the 35 L in the SK technical brochures that we find online.

SK’s technical brochures give the angles of view for lenses in the Apo-Digitar line at the maximum aperture of f/5.6 and again for f/11. For some lenses those two values are the same, while others, like the 35 XL, are shown as having a larger angle of view at f/11. That value at f/11 is the one that SK routinely used on the lens barrel itself.

Why the 35 L-88* was labelled the way it was is still unclear. Perhaps the decision to relabel it in c.2004-2005 as the 35 XL-102* was a standardisation or correction. Or perhaps there really was some modification made to the optical formula.
Of the two (standardisation vs modification to the optical formula), I'm leaning to the former based on the logic I've mapped out here. It just makes sense. Why would Jenoptik, Schneider and Plaubel come together to build and sell a lens that would just annoy people -- because that's what a lens with a 70mm image circle at f/11, and 55mm at f/5.6, would do on the camera in the picture you shared with me, and which I posted at the top of the thread.

One obvious test is to compare the maximum shift that can the two versions can sustain with good image quality. Another test would be to measure the exact FFD, which is less simple to measure accurately.
Given how rare these are, we may never see this test.

My 35 XL can be shifted at least 15mm (with the IQ4 in landscape orientation) with excellent results. As we know, achieving optimal image quality involves having a Copal 0 (or equivalent) that was built or shimmed well to the 20mm depth specification, and it seems that many weren’t.

Rod
15mm makes sense. That's an image circle of 88mm, so within the specified 90mm.
 

diggles

Well-known member
That took the sting away from discovering that I had not purchased the lens I thought I had purchased.
To quote Bob Ross "We don't make mistakes, we have happy accidents…"

Nice find. Your sample images look really good, Rob.
 

rdeloe

Well-known member
To quote Bob Ross "We don't make mistakes, we have happy accidents…"

Nice find. Your sample images look really good, Rob.
Thanks Warren. Those were from a quick run around campus on a workday to find some buildings. My usual subjects would not reveal the differences as well between the Pentax-A and the S-K because the distortion that bothers architecture folks is simply not evident in the woods.
 
Last edited:

anwarp

Well-known member
That was an interesting read! Thank you Rob!

So, does this lens join the bang for buck list?


Anwar
 

daz7

Active member
I think it is the same case as with 47mm SK lenses coming in two descriptions (XL and non-XL) but being the same lens.
 

rdeloe

Well-known member
That was an interesting read! Thank you Rob!

So, does this lens join the bang for buck list?


Anwar
Based on the arbitrary cutoff of $500 USD or $500 EUR, not quite. ;) I paid less than the cutoff for the lens itself, but I had to buy a Compur 0 shutter that had another lens in it. And then shipping and taxes. If I had a donor #0 shutter lying around, that would have saved me some money. All in all it's still a lot less than the going rate for XL in Copal 0 shutters.

I don't know how common these things are. Rod tracks S-K lenses very closely, and thinks the "L" version was only sold for two years. I've only seen two of the ones in the Plaubel board, I'm not sure if L versions were sold in regular shutters. Those might be a the best value.
 

John Leathwick

Well-known member
I think it is the same case as with 47mm SK lenses coming in two descriptions (XL and non-XL) but being the same lens.
I'm not so sure on that one as the 1995 SA brochure that I have lists the XL and non-XL versions as having markedly different weights and dimensions. For example, the rear element on the XL is nearly 20mm larger in diameter than the non-XL version, and the XL version is around 40% heavier. It appears that the non-XL version is a very close match in specs to the 47mm Apo-Digitar XL, and I've seen suggestions that the Apo-Digitar may be based on the non-XL 47mm if not a rebadged version of it.

John
 

daz7

Active member
I'm not so sure on that one as the 1995 SA brochure that I have lists the XL and non-XL versions as having markedly different weights and dimensions. For example, the rear element on the XL is nearly 20mm larger in diameter than the non-XL version, and the XL version is around 40% heavier. It appears that the non-XL version is a very close match in specs to the 47mm Apo-Digitar XL, and I've seen suggestions that the Apo-Digitar may be based on the non-XL 47mm if not a rebadged version of it.

John
I meant 47 digitars, not SA - SA XL and non-XL are two different lenses, but Digitars 47mm XL and Digitar 47mm (non-XL) are the exact same lenses, and in fact, from what I read, identical to SA 47mm (non-XL)
 
Top