ULF Minimal Focusing Discussion

[Michael Kadillak]

With the permission of Corran who made the image and posted it in an earlier discussion, I wanted to share some thoughts on employing minimal f stops in concert with optimal focus plane selection in an effort to keep diffraction to a minimum (even though some will argue that contact printing is indifferent to diffraction - which I do not believe is the case) and discuss the topic.

As an 8x20 photographer myself I contend that the image made by Corran on 8x20 with a 305mm lens with a modest from tilt could have been made at f32 and maintained sharp focus throughout as there are no complicating components of the proper film plane in the photograph.

The question therefore becomes do we ascribe to the minimal focus criteria in concert with judicious film plane selection as I feel is the desired case or do we overly employ increasingly (and I would comment unnecessarily) small f stops and let diffraction creep into our ULF photography? I have never made an 8x20 photograph beyond f45 in a long time and hope that I never have to in the future. ULF film costs and a desire to optimize the process inherently leads me to keep produced resolution as high as possible without sacrificing on image quality. Interested to see where the body intellect comes into on this topic.

[Mark Sawyer]

Resolution loss due to diffraction is small compared to resolution loss due to poor focus. If contact printing, as is often the goal with ULF negatives, diffraction is negligible. Remember that Edward Weston used pinhole apertures for his close-ups of vegetables and shells, and no one complains about the image quality of his 8x10 contacts...

[Oren Grad]

Interesting question.

I did one experiment, quite a few years back now, with a 270 G-Claron to find out whether I could see degradation of image character from stopping down, in contact prints (8x10, in this experiment). On burying my nose in the prints I decided that everything was fine through f/64, but that the print was visibly worse (subjectively unpleasant) at f/90. Note that I'm not talking about resolution per se, but about overall subjective image character. This will be a function not just of resolution numbers calculated on the basis of idealized optical theory, but also of the design tradeoffs of real lenses made of real glass, mounted in shutters with real physical diaphragms with (in my late-model Copals) non-circular shapes.

The results might be different if I were to do such an experiment with an Apo-Sironar-S, which renders OOF backgrounds and focus transitions differently from the way G-Clarons do, and which I generally prefer when working conditions allow. (In some situations the coverage and/or the size/weight advantages of a G-Claron or a Computar are compelling.) But the conclusion, that f/64 is OK but beyond that might be iffy, seems reasonable enough to me, and in most circumstances I'm not inclined to go past f/64 anyway for reasons of shutter speed, so it's convenient for me just to go with that as my rule of thumb.

Also: in my experience and to my taste, tilts and swings are a cure that's usually worse than the disease. On the one hand, I rarely photograph things where the configuration of the subject lends itself to a clean tilt or swing - ususally there are important things (trees, especially!) sticking out of any tilted/swung focus plane, however positioned. And on the other, even where that's not the case, tilts and swings usually just rearrange the OOF stuff in ways that are counterintuitive and disconcerting to my eye, rather than making it go away.

So I am resigned to the fact that in larger formats and with scenes focused at midrange, which is most of what I care about, I will not achieve anywhere near pan focus. Sometimes I don't even care, if the lens I happen to be using draws beautifully in that respect, or if for any other reason I want sharpness to fall off. But regardless, I'm stuck with it.

So my practice is simple: when depth of sharp rendering is important with big formats (say, whole plate or larger) I stop down to f/45 or f/64 if the light will allow, and I try to position the point of focus so that the falloff fore and aft results in a reasonable rendering overall. (I don't always get it right!) And that's all. I can't remember the last time I used a tilt or swing.

[Tin Can]

Key words perhaps.

Perspective control

Hyperfocal

Angular and rectangular objects vs forest

[Corran]

To summarize my thoughts that I shared with Michael:

The rock in the foreground is closer than some may realize - probably about 5-6 feet in front of the camera. I did use some tilt. In my experience with a foreground + trees situation, it can be problematic to go overboard with tilt, lest the focus on the tops of the trees in the image go way outside of the DOF "slice" as oriented by the use of tilt. The common practice of focusing 1/3 into the scene for me has often caused the foreground to be insufficiently sharp, in my opinion.

I think it's important to remember that "DOF" is relative - there is only one true point of focus, and everything else is in focus only approximately within the DOF attained with the aperture. I have generally had many more situations where my DOF was insufficient in an image, than situations where diffraction was noticeable or objectionable. f/64 and f/90 are generally my go-to apertures in 8x10 and 8x20 for maximum DOF with foreground/background compositions.

8x20 (200 x 500 millimeters) is an oddball format, so to compare apples-to-apples, I would like to compare shooting this image with a "6x15" (56 x 140) camera (6x17 cropped basically). If I were to have shot this image with such camera and a 90mm, I would attain roughly the same composition. The equivalent aperture for DOF would be about f/25, which sounds reasonable to me for max DOF in 120 format. If shooting with an XPan (24 x 60) that would be the same as shooting f/11. Again, reasonable. The equivalent apertures at f/32 would be shooting at f/9 for the 6x15 shot and f/4 for the XPan - I think most would call this unreasonable for the needed DOF (yes I realize the XPan would not have movements)?

Finally, regarding diffraction - I believe 4-5 lp/mm is required for a perceptually "sharp print," as mentioned by rdenney on occasion. The diffraction limit of f/90 is 17 lp/mm from what I understand, so f/90 should still allow for a 3x enlargement at worst.

Oren, your comments on your 270mm is interesting. I will point out, since you mention aperture shape, that my 305mm GC is in an older shutter with many blades, creating a circular aperture. I do generally shy away from Copals that have 5 blades. However, at f/90, what was out of focus, to show the characteristics of such an aperture?

[ic-racer]

If you can see the effects of diffraction in your prints, then you can see the effects of diffraction in your prints. If you cannot see the effects of diffraction in your prints, you cannot see the effects of diffraction in your prints.
A complicated equation with many variables, including viewing distance, etc, can be used to answer your question for you, but maybe you can answer the question for us and let us know what you think about the prints.

[Corran]

For the record, I have not yet printed this image. That said, I have not done any explicit tests to say one way or another whether or not there was clear differences in DOF or diffraction when shooting 8x10 or 8x20 at these apertures.

[Michael Kadillak]

To summarize my thoughts that I shared with Michael:

The rock in the foreground is closer than some may realize - probably about 5-6 feet in front of the camera. I did use some tilt. In my experience with a foreground + trees situation, it can be problematic to go overboard with tilt, lest the focus on the tops of the trees in the image go way outside of the DOF "slice" as oriented by the use of tilt. The common practice of focusing 1/3 into the scene for me has often caused the foreground to be insufficiently sharp, in my opinion.

I think it's important to remember that "DOF" is relative - there is only one true point of focus, and everything else is in focus only approximately within the DOF attained with the aperture. I have generally had many more situations where my DOF was insufficient in an image, than situations where diffraction was noticeable or objectionable. f/64 and f/90 are generally my go-to apertures in 8x10 and 8x20 for maximum DOF with foreground/background compositions.

8x20 (200 x 500 millimeters) is an oddball format, so to compare apples-to-apples, I would like to compare shooting this image with a "6x15" (56 x 140) camera (6x17 cropped basically). If I were to have shot this image with such camera and a 90mm, I would attain roughly the same composition. The equivalent aperture for DOF would be about f/25, which sounds reasonable to me for max DOF in 120 format. If shooting with an XPan (24 x 60) that would be the same as shooting f/11. Again, reasonable. The equivalent apertures at f/32 would be shooting at f/9 for the 6x15 shot and f/4 for the XPan - I think most would call this unreasonable for the needed DOF?

Finally, regarding diffraction - I believe 4-5 lp/mm is required for a perceptually "sharp print," as mentioned by rdenney on occasion. The diffraction limit of f/90 is 17 lp/mm from what I understand, so f/90 should still allow for a 3x enlargement at worst.

Oren, your comments on your 270mm is interesting. I will point out, since you mention aperture shape, that my 305mm GC is in an older shutter with many blades, creating a circular aperture. I do generally shy away from Copals that have 5 blades. However, at f/90, what was out of focus, to show the characteristics of such an aperture?

I am in respectful technical disagreement with a number of the statements above and I will state my case individually item by item for the benefit of this discussion on topic.

Item #1 - The rock in the foreground is closer than some may realize - probably about 5-6 feet in front of the camera. I did use some tilt. In my experience with a foreground + trees situation, it can be problematic to go overboard with tilt, lest the focus on the tops of the trees in the image go way outside of the DOF "slice" as oriented by the use of tilt. The common practice of focusing 1/3 into the scene for me has often caused the foreground to be insufficiently sharp, in my opinion.

The fact that the foreground rock is 5-6 ft in front of the camera is PERFECT for LF camera focusing. I go back to the fundamentals of selecting a focusing plane (Howard Bonds Article Photo Technniques 1998 is perfect primer for this application). You pick a background spot 1/2 to 2/3rd's up into the background trees and focus on that point. Then you focus on the foreground rock. If you have to extend the front standard to get the foreground rock in focus tilt the front standard more. Iterate on this back and forth until the selected part of the tree in the background and the foreground rock are perfectly in focus and there is your film plane. Now lets be sensible. There is no such thing as too much forward tilt Why? You are working with the 8" proportion of your format and you know the lens covers the 20" lateral so tilt away as necessary because you have enormous latitude in this regard. If the front tilt goes down 30 degrees so be it. Be disciplined on where you select your film place. Now you have 1/3 DOF into the subject matter in front of the selected film plane and 2/3rds in back of the film plane and your subject matter is all covered adequately for focus. Stop down 2 stops while you are watching the GG and you will see the entire image area snap to attention in focus. My point is that this process begins and ends with visually and purposefully designating an optimal film plane and one tilts as necessary to make that happen. I would lay money that this image could be easily made at between F22 and F32 all day long with the proper film plane election and the image area would be as sharp as a hand full of razor blades. Done it countless times myself to know this is in fact the case.

Item #2 - I think it's important to remember that "DOF" is relative - there is only one true point of focus,

Again I respectfully disagree. There are countless points of focus with a view camera and its movements and all of these are uniquely controlled by the LF photographer. That is why the process of focusing a view camera is an iterative process. DOF is therefore uniquely related to the selection of this film plane. When you optimize the designated film plane and judiciously use camera movements to the degree that they compliment your vision and not fall back to stopping down unnecessarily (which is at the core of why I am bring up this topic in the first place) you have the opportunity to only stop down minimally, make sharp high resolution images and then you can leave stopping down as a last resort option that rarely if ever gets deployed. But this whole process of optimization starts with a disciplined approach to selecting the optimal film focus plane.

Again, I am not in any way shape or form jamming on anyone in this discussion. I know that 8x20 film is expensive and all I am interested in is ensuring the best possible results from the field efforts are obtained. Only stop down minimally to grab the optimal center of the resolution sweet spot of your optics and make some gorgeous prints.

[Corran]

There is no such thing as too much forward tilt

Disagree. Yes, the 8" vertical makes things less worrisome with regard to the tops of trees, but is still a problem to be considered.

With regard to point #2 - you're right, I should have said plane of focus, not point.

One final point that I think is tangential but I do often consider - reciprocity, and more specifically its broader effect on contrast. The image in question was shot in early morning, with little directional light. Normally this situation would call for N+ development to increase contrast. However, when correcting for a large amount of reciprocity failure, you can attain a higher contrast due to the inherent buildup of exposure in higher exposure areas which get concurrently less reciprocity failure. I use this to my advantage often in these situations to get higher contrast but not have to change dev times.

On diffraction: I have a friend who used to shoot everything at f/64, with 4x5 film. A slight softness was apparent in prints of 16x20 size. This is the only time I have ever directly observed softer images from apparent diffraction. I generally do not stop down beyond f/32 on 4x5 for this reason.

Also, I have to wonder about the differences in apparent diffraction at the same actual aperture size for a given lens focal length. What I mean is, I believe there is some difference in diffraction when using longer focal lengths, due to different size in the actual aperture, which may be why diffraction becomes more apparent in the smaller formats.

I think at the end of the day, I'd rather be safe than sorry, especially on a format that I will only be contact printing.

[Vaughn]

Key words perhaps...

And I'll toss in: Flim flatness

Oren, thanks for your words of wisdom. The only time I go down to f90 is when I have a lens that has one...tho if possible I use F45 or f64. I always use all the movements, tho often I set them at their neutral position. A habit I picked using a 4x5 for years with no indents -- I don't assume that neatly zeroed out movements are what I need, or don't need to make an image. It's all on the GG.

After learning photography under the redwoods, I am always amazed how different it is in the 'grand landscape', such as my month in Zion. Instead of ferns 5 feet away and redwoods back there a 100+ yards away, the closest foreground might be 20 feet or more, f22 seems to work just fine and most of my meter reading are about the same. Plus the legs of my Ries slide easily.

11x14 Carbon Print -- dang fern in the foreground moved in an unfelt breeze. Would have been fairly sharp otherwise.