Which 90mm f/4.5 or f/5.6?

[Jack]

The Schneider regular SA f5.6 has an identical spec IC to the Nikkor f8 and f4.5 at 235mm. The 90 XL spec is 24mm larger at 259. The regular Schneider SA f8 90 is specd at 216, 19mm less. Nikon specs their f8 lens at f22, while they spec their f4.5 lens at f16 -- a significant difference and why they spec the same total (smaller apertures generate larger IC's).

Now note you need 154mm IC to fully cover the 4x5 frame. So with the regular Schneider f8 SA, I can shift a total of (216-154)/2 or 31mm; the Nikkor can theoretically go (235-154)/2 or 40mm -- a 9mm difference if the spec isn't overly optimistic. At the end of the day, I never shifted more than maybe 15mm tops with my 90, so never even came close to running out of IC on my Schneider, but YMMV... Personally, if I knew I was going to need gobs of shift, I'd consider the XL to insure I never ran out! (Actually, I'd buy the 72 XL instead if I needed that kind of movement, but that's another story .) Seriously now, the 90 XL will generously cover 5x7 with about 30mm of rise capability, and 5x7 is where this particular set of lenses is going to be challenged on IC, not at 4x5...

Cheers,

 

[SCHWARZZEIT]

Actually yes. Note that I am measuring line-PAIRS per mm or LPmm, which is half the measurement of lines PER mm or L/mm.

When I was writing lp/mm I also meant line pairs per mm. So you need at least two pixels to cover one line pair of information.

The BL I had at maximum res (9000 interpolated x 12000 actual pixels) could discern around 56 LPmm on my test target. In some cases --- the Rodie APO S and Schneider APO L, I had a few lenses at a few apertures outresolve (or at least match) the camera -- notably the Schneider 120 APO L, 150 APO L and 210 APO W; in Rodie, the 135 APO S and 150 APO S did as well. My Sinar badged 90 SA MC and Linhof badged 65 SA MC both made 52, as did one of the Schneider 110 SS XL's I tested. However in most cases, you could see the lenses fall off before the sensor, usually around 50/52 LPMM for a great lens, 45/47 LPmm for a good lens, and 40/42 or so for most normal 4x5 lenses. Most of the older lenses only managed mid through high 30's, and the worst I tested managed a 26 or 28 -- but that was an old turn of the 20th century Tessar.

The problem I think with your test could be that the Betterlight or any other image sensor needs a certain minimum contrast to discern detail close to its Nyquist frequency. That's where the sensor MTF drops to zero. The alignment of the pixel grid with the lines of the test pattern could also vary and affect contrast.

The Kodak Imagelink HQ microfilm I used for my own tests has an MTF of 100% at 50 lp/mm, 80% at 100 lp/mm and 45% at 200 lp/mm according to its data sheet. I used a high res drum scanner to analyse my results at 8900 ppi and from all the lenses I tested I could get more than 80 lp/mm at some aperture with some like the 110 XL close to 100 lp/mm. When I had the chance to look at some of the negs under a microscope I found that even my drum scanner wasn't able to resolve all the detail on the film. In fact the resolution of the 110 XL at f/8 was between 118 and 139 lp/mm with the lower value being clearly resolved and the higher value barely showing contrast when enlarged x100. When I tested my scanner with a high contrast lith target it was able to resolve more than 160 lp/mm. But the actual contrast on film from LF lenses at their limiting resolution is simply too low for the scanner to resolve.
Dr. Hubert Nasse from Zeiss told me that the resolving power of a lens is always limited by diffraction. The difference between a so-called diffraction limited lens and a lesser lens is that the aberrations of the latter already reduce the contrast at lower spatial frequencies in a way that their contrast modulation becomes so low that it becomes impossible to detect detail at higher frequencies by most image recording media, thus limiting their effective resolution often far before the theoretical diffraction limit.

But keep in mind, for a 16x20 print, all you really need is about 25 LPMM on a 4x5 neg for as sharp an image as you can see with your naked eye on that print...

As a rule of thumb I agree and so does the printing industry by setting 300 ppi as a widespread standard. However, there have been experiments that many viewers could see a subtle difference in print detail quality up to 15 lp/mm. For example you will see a quality difference from a high quality 8x10" contact print to a drum scan of the same film printed to 8x10" at 300 or 400 ppi. John B. Williams in his book Image Clarity writes that you'll not perceive an improvement in detail in a print above 25 lp/mm.

In the end all these tests are more of an academic nature than having a serious impact on real world photography. But the process helped me a lot to understand the subject and optimize my own imaging chain where it mattered.

-Dominique

 

[Jack]

Hi Dominique:

My test was far less formal. My USAF test target has K, R, B and Y resolution targets oriented vertically, horizontally and tangentially, so I could measure all aspects. I could also easily measure center, mid and edge performance of the lenses. The BL scanning software has live focus, so I could confirm focus electronically, and actually can see slightly differing focus point for the R, G and B channels on most lenses, including the better APO lenses(!) I tested my lenses at f8, 11, 16 and 22, and not further because I could detect diffraction with this system kicking in usually between f11 and f16, certainly by f22. I did not use a Koran target or software and had to judge resolution to my best guess at around 20% contrast, which is certainly flawed. And yes, contrast is an issue in my format of test as was my visual measurement.

Anyway, my system was resolution limited to about 56 LPmm in the best case, though the best lenses I mentioned I hit that clearly, where the others did not, hitting the ranges I indicated. Moreover, I could easily see the resolution falloff from center to edge for all lenses.

I did test my lenses shifted as well (amount varied with focal), but found that most under 150mm lost resolution pretty quickly, though longer focals held longer (obviously).

FWIW, Mike Collette, the owner of Betterlight, also tested lenses for his backs. He found that extremely few LF lenses were capable of meeting, let alone outresolving his Super 12K scanning back. And his results supported my own findings, in that only a few Schneider APO L and Rodenstock APO S lenses delivered the goods -- specifically the 135 Rod APO S and the Schneider 120 APO L exceeded it, and the 150's from both were very close. Also, they only made it at f8 as you could see the effects of diffraction affecting results by f11. The 110 SS XL did not come close for him either, hitting the 50 or so LPmm I found.

FWIW only...

 

[Stuart Richardson]

I would imagine that the camera and shooting conditions are more important than the lens, assuming you are using a good modern lens. Most 4x5 cameras are not nearly as rigid as smaller format cameras...even the very good cameras tend to have some flex and play in them. It's just a fact of life in the materials used and the desire for camera movements, portability etc. In the real world when you are out in the landscape with a field camera and a even slightly breezy day, the vibrations from the wind and the flex are going to equalize a lot of lenses.
But isn't the point of large format that you don't need to pixel peep? You can use the lens that fits your criteria (focal length, size, coverage, aperture etc) and be reasonably sure that you will be able to print very large even if you are not using the latest, greatest lens ever. Technique and execution is more likely to matter in LF than in most of the smaller formats where the cameras do so much more for you (like stay plane parallel, focus automatically etc.).

 

[henningw]

The Schneider regular SA f5.6 has an identical spec IC to the Nikkor f8 and f4.5 at 235mm. The 90 XL spec is 24mm larger at 259. The regular Schneider SA f8 90 is specd at 216, 19mm less. Nikon specs their f8 lens at f22, while they spec their f4.5 lens at f16 -- a significant difference and why they spec the same total (smaller apertures generate larger IC's).

Now note you need 154mm IC to fully cover the 4x5 frame. So with the regular Schneider f8 SA, I can shift a total of (216-154)/2 or 31mm; the Nikkor can theoretically go (235-154)/2 or 40mm -- a 9mm difference if the spec isn't overly optimistic. At the end of the day, I never shifted more than maybe 15mm tops with my 90, so never even came close to running out of IC on my Schneider, but YMMV... Personally, if I knew I was going to need gobs of shift, I'd consider the XL to insure I never ran out! (Actually, I'd buy the 72 XL instead if I needed that kind of movement, but that's another story .) Seriously now, the 90 XL will generously cover 5x7 with about 30mm of rise capability, and 5x7 is where this particular set of lenses is going to be challenged on IC, not at 4x5...

Cheers,

You're quite right in all of that, Jack. I do architectural photography and my observations and predilections are coloured by that. On the other hand, I did try both the f/4.5 Nikkor and the f/8 and found them to have essentially identical useable image circles.

I do use the 72XL a lot, and find that an outstanding lens for my work. The older 75/4.5's and 5.6's didn't have quite enough movement but I really don't run out of room with the 72. The 90 Nikkor has just enough movement for anything I want to do with a 90 whereas the 90XL has more than I need (and is huge, which causes problems with filters).

None of my clients are willing to have me shoot larger than 4x5 anymore, so I sold my bigger stuff a while ago.

Henning

 

[carstenw]

Well, the decision is made. I noticed that Jack had a 90mm f/5.6 Super Angulon Sinar select at some point, and I read a bunch of old/modern lens comparisons for LF, and decided that getting an XL or a 4.5 Grandagon-N was just not necessary. I am not in this game for the absolute last drop of resolution, almost the opposite. I want to move beyond all this tech, and just shoot, not worrying too much beyond my careful initial purchases. I will get a nice, simple, good kit, and then stop worrying.

I found a 90mm f/5.6 Super Angulon MC Linhof select running at around 360 Euro in the last minutes, and made my bid of 419 Euro in the last seconds of the auction. It went for 420 Euro... I was quite frustrated, but my current situation with my hands full of unsold equipment doesn't allow me to bid higher than that. Grumbling to myself, wondering if I would have gotten it had I bid 429 or 439, I distracted myself by reading through eBay's "perhaps you are interested in one of these other auctions" page, displayed after an auction is over. I looked through them, and spotted another 90/5.6 SA MC with a Compur #1 shutter (#1? I thought these were built with #0 shutters?). Grumbling about the Linhof select I just missed, I clicked on it. 399 Euro Buy Now, good condition, everything seemed hunky dory. Then I spotted it: clearly visible on the front lip, although unmentioned in the auction description, was the script Linhof! 0.1 seconds later it belonged to me

After the auction, I did notice something I am not familiar with. On the board on which the lens is mounted, there is a Prontor Professional shutter. What is that, and can I just remove it?

http://cgi.ebay.de/ws/eBayISAPI.dll?ViewItem&item=350312792015

Although I will certainly use it wide open in situations where I want small DoF and where the corners don't matter too much, I would still like to know how far I have to stop it down to get very sharp corners. Anyone?

 

[Jack]

I would imagine that the camera and shooting conditions are more important than the lens, assuming you are using a good modern lens. Most 4x5 cameras are not nearly as rigid as smaller format cameras...even the very good cameras tend to have some flex and play in them.

This is true too. But here again, I used among the most rigid 4x5 cameras for the test -- my Arca F-Metric or Ebony -- both mounted on either a 3 or 5 series Gitzo CF pod and my ground-floor studio with concrete floor. The most rigid camera the folks at Betterlight found was the Toyo monorail, the version *without* the base tilts. Next in line were the Arca Metric (not Monolith!) and Ebony.

 

[Jack]

You're quite right in all of that, Jack. I do architectural photography and my observations and predilections are coloured by that. On the other hand, I did try both the f/4.5 Nikkor and the f/8 and found them to have essentially identical useable image circles.

I do use the 72XL a lot, and find that an outstanding lens for my work. The older 75/4.5's and 5.6's didn't have quite enough movement but I really don't run out of room with the 72. The 90 Nikkor has just enough movement for anything I want to do with a 90 whereas the 90XL has more than I need (and is huge, which causes problems with filters).

None of my clients are willing to have me shoot larger than 4x5 anymore, so I sold my bigger stuff a while ago.

Henning

We are definitely in synch with this, and I understand your need for as much shift as practical with the 90 for interiors. My case was different as I was looking for color and tone consistency while having adequate IC, hence our different choices -- neither is wrong, and in fact both seem correct for our individual purposes :thumbs:.

 

[Jack]

Then I spotted it: clearly visible on the front lip, although unmentioned in the auction description, was the script Linhof! 0.1 seconds later it belonged to me

After the auction, I did notice something I am not familiar with. On the board on which the lens is mounted, there is a Prontor Professional shutter. What is that, and can I just remove it?

http://cgi.ebay.de/ws/eBayISAPI.dll?ViewItem&item=350312792015

Although I will certainly use it wide open in situations where I want small DoF and where the corners don't matter too much, I would still like to know how far I have to stop it down to get very sharp corners. Anyone?

First off, congrats, you did good. Second, the Prontor Pro is an excellent shutter. The "stuff" that looks like it's on the board is actually part of the shutter -- makes adjustments easier from behind the camera -- only downside is the bit of added size and weight. You can remove it, but may need a standard cable release socket -- I cannot specifically recall on the Prontor. Regardless, I would mount it and shoot it as-is for a while before deciding...

On film even with a good scan, you should be able to use up to f32 without seriously degrading the image due to diffraction. Bottom line is there is a curve of tradeoffs between the added DOF and the added diffraction -- heck, you may even like the slight veiling, glowy diffraction effect from f45, so possibly worth a try!

 

[carstenw]

Thanks for the info, Jack. About stopping down, I was more looking for what my minimum aperture should be before the corners sharpen up. f/8? f/11?

I presume that the Compur-1 is just mounted instead of the usual 0 due to the Prontor shutter?

 

[carstenw]

Now I just have to decide if I spring for a Linhof Technika (IV/V/MT/MT2k), or wait for the Chamonix 45N-2. I will go read documentation for the Technika to get a better feel for what I might be giving up. Superficially, it has a lot less movements, but it looks like it is cleverly enough done that with a bit of swing or tilt on the back, one can actually get most things done regardless.

I don't anticipate a huge need for movements, but since this is partly a learning experience, I was interested in having them regardless. Most of the time, I can probably get away with a little front rise-n-fall, and a little rear swing-n-tilt, with the occasional front tilt also coming in handy. The Technika can handle all this.

 

[Jack]

I'd get an Ebony before the Tech. Heck, I'd buy a Shen Hao before a tech, but that's me ...

 

[DougDolde]

I'd go 110mm SSXL and 210 Apo Symmar.

 

[carstenw]

Doug, I guess you are coming from a more well-fed wallet state than I I am quite happy with the two lenses I have bought. Both have good reputations, neither is going to be beat easily by any lens, including the two you mentioned, unless you really get out the high-powered magnifying glass...

 

[carstenw]

Jack, I love the Ebonies (Ebonys?) but they are very pricy, and hard to find second-hand. I would also need new lens boards, yet another additional cost.

 

[Jack]

Jack, I love the Ebonies (Ebonys?) but they are very pricy, and hard to find second-hand. I would also need new lens boards, yet another additional cost.

Ebony uses Linhof Tech boards too Arca sells and adapter for Linhof tech boards -- in fact almost every camera out there can be adapted to Linhof tech boards.

 

[carstenw]

Ah, I thought that Ebony had their own. There aren't that many different (relatively modern) 4x5 boards that I can tell. I thought it was Technika, Ebony and Toyo, but apparently I mixed something up.

 

[Stuart Richardson]

I know you are keeping the budget down, but I would agree with Doug. I have two lenses -- 110mm SS XL, and 210 APO Symmar L. Both are fantastic. My camera is an ebony SV45Ti, which is also great to use. I think monorail cameras are much more pleasant to work with in the studio, but the ebony is so great to use out in the field and it is so small and light when all folded up.
However, I think you will really enjoy the lenses you got. That 90 looks fantastic!

 

[carstenw]

Look everyone, I looooove the Ebony cameras, and would loooooove to buy one, but the Technika is expensive, and the Ebonys are twice as much, so there is no chance in hell. Unless someone sells me theirs cheap

And yes, no matter which lens you get, there is something better out there, but honestly, there are strict limits to how much better a lens can be than "great". Here is food for thought, and keep in mind that my lens is sharper, has more coverage, and is faster than the 90 Angulon:

http://www.hevanet.com/cperez/test/AngSSXL.html

 

[Stuart Richardson]

Sorry if my post came off as trying to convince you to buy the Ebony and the 110 + 210, rereading it, it sounds like that. What I was really trying to say was that I love what I have, but I think you will also love what you are getting. I agree that a lot of times the differences between the super high-end and the very good is not so great as it is made out to be. Especially in the largest formats like 4x5.