Diffraction. When does it really matter with LF?

[Alan Klein]

After reading through this entire (interesting) thread, I’m left with one seemingly simple reaction. After one has taken all the time, effort and expense with LF gear of choice to set-up and make an image on a composition (with needed movements and considering DOF and diffraction considerations), why not make the image at more than one f-stop (whenever possible given wind or other conditions)? Admittedly, I use 4x5 gear and not larger and more expensive film, but I usually take 2-3 exposures of most compositions and vary the aperture to be sure I get what I want as far as DOF. When it comes time to print, I use my light table and a good loupe to help me decide which film(s) to work with. I’m mostly doing landscape photography, so not often wanting to isolate focus narrowly with larger apertures than f22; but if I’m going to take more than one of exposure I usually start with the best aperture I’ve determined appropriate, and then stop down one and/or two stops (maybe also changing lens filters or other factors as experience suggests).

The problem with varying the aperture is that the DOF will change as well. When I bracket, for exposure not DOF, I change the shutter speed. That keeps the DOF the same on all shots.
I calculate what I think is the correct DOF that will get me the right focus range, then stop down one more stop for good measure. But the bracketing will be done with the shutter. Of course, there are no half shutter speeds, only full stops. You can get 1/3 or 1/2 stops changing aperture. But then the DOF changes. I've done this regularly with medium format shots as it's easy and cheap/ I haven't done bracketing with 4x5 though, except maybe one extra shot on the flip side of the same film holder.

[Vaughn]

Try stopping down instead on a couple images and check out if diffraction is an issue for you at f45 or f/64. Unless getting more DoF is a problem of course, usually it is not. Try printing the negatives to 16x20 or your max size the enlarger will enlarge a 4x5 to (exposing just a 8x10 sheet of paper in a 'typical' area) and see for yourself what f/stop works best for you. The feedback from your prints will help.

[Doremus Scudder]

Is this the "simple" table your were referring too towards the bottom of the link your provided? So if I understand, you measure the mm distance on the rail between near and far focus point to determine the f stop. If it falls between two, do you use the smaller f stop?

Do you set the focus point half way between the two or use some form of 1/3 - 2/3 ratio?

Could you explain the N value?

Table of optimal fstops (2)
This shows you the best value of the f-stop to use.
D(mm) F N
1 16.6 19.4
2 22.6 27.4
3 32.2 33.5
4 32.6 38.7
5 32.9 43.3
6 45.2 47.4
7 45.4 51.2
8 45.6 54.8
9 45.8 58.1
10 64 61.2

Alan,

Yes, this is the table I was referring to. The center column, headed "F" shows the optimum f-stop for a compromise between DoF and diffraction. The column labeled "N" shows what the author (QT Luong) calls the "admissible" f-stop, which guarantees a circle of confusion of 0.066mm on 4x5. The optimal column is what I use.

When the f-stop in the optimal column is larger (smaller number) than in the "N" column, the performance should be even better. When it's larger, you'll end up with more diffraction, but more depth of field.

Hope that's clear.

Doremus

[Doremus Scudder]

Heck, Doremus, I just love spending time looking at that opalescent image on a big ground glass, even if I don't even take a shot. And I'm damn nitpicky about how the frame specifically crops the composition. Moving the tripod an inch either direction might spoil it. When I see a promising spot, I generally stake it out, literally - I stab one of my walking poles into the soil at that exact spot. Then I walk around a little more, seeing if there's an even better spot. That's half the fun of it. The hunt is just as important as the kill.
Depth of field then becomes yet another compositional "hunt". And no rote formula can substitute for that "just right" feel on the ground glass itself, which is double-checked in critical spots with a loupe of course - those places in the composition I want the eye drawn into due to its acute focus, versus where I want the eye to slightly recede. Don't confuse that with gross selective focus; that's OK, but I often aim for a much more subconscious effect on the viewer. A good illusionist does not show his hand.

Don't get me wrong, Drew, I like viewing the image on the ground glass too. It's just that sometimes (more often than I'd like) circumstances keep me from being able to comfortably see everything at once on the gg (things like camera position, setting up in a precarious situation, lots of light coming from behind, wind, rain, crashing waves, whatever...). In those instances, I really like that I'm able to work quickly in those cases, just being sure that the composition I want is "in there" somewhere and then just focusing quickly and exposing.

I pretty damn nit-picky about my borders too. So much so that I almost always crop my enlargements to the exact dimensions and aspect ration that I want and that I think is determined (demanded) by the subject; I don't limit myself to the 4x5 aspect ratio.And, I rarely am able to get borders exactly where I want them from the camera position I choose; I'd need lenses in an infinite number of focal lengths for that. I just choose the next wider lens and make the shot, knowing that the composition I planned is within the borders. If I'm lucky, I don't have to crop too much, but sometimes, say when my 135mm lens is just a tad too long for a shot and the only shorter lens I have with me is a 90mm, and moving the tripod back would destroy the chosen perspective, I just make the shot with the wider lens and crop (and curse the fact that I don't have a 125mm lens with me...).

As for depth of field... My preference and my style is for as much as sharp as possible. Optimizing that with diffraction degradation is pretty straightforward. So, I'll stick with my "rote" method of working. I personally find a lot of out-of-focus areas aimed at "directing the eye" to end up doing just the opposite, i.e., drawing my attention there and distracting from the geometry and leading lines of the composition. That's not to say everything in my prints are razor sharp; I know how to make a foreground object pop out of a background by leaning the focus toward the near, but not enough to really defocus the background; just enough to give that 3D effect, etc. But those things are refinements.

Anyway, the post of mine you were responding to was directed at the beginner's mistake of spending a lot of time finding camera position by viewing the ground-glass image (which Alan doesn't seem to be making, so my post is moot anyway ). I do my best to find the precisely right place for the camera before I set up. Yes, I have to scoot the tripod around an inch or so this way or that to get what I really want sometimes because I didn't really hit the mark setting up, but that's just part of the process.

I think we pretty much agree on all of this; I'm just addressing things to those struggling with the basics still, so I'm simplifying my information to that end,

Best,

Doremus

[Doremus Scudder]

That's what I've been doing using a micro 4/3 camera as a viewfinder in BW display if shooting BW and finding the spot before I place the tripod and camera. I also have been using the camera as a meter, but that's another story.

So a few hours ago, I was out shooting with a 75mm lens. The main object was about 4' to 80-100 feet. I focused on the far point using the asymmetrical line where the axis of the standard is. Then tilted the back standard backward to focus the near log up at the tip about four feet away. I set it for f32 and fired away. I just realized though that tilts probably aren't required with a 75mm. At f/32 I was covered from around 3 feet to infinity at coc of .1mm, with the hyperfocal at 6 feet. SHould I have used that and skipped tilts?

So that raises a general question about tilts. When shooting with wide angle lenses like a 75mm, it doesn't seem even required to worry about them. Just use the DOF tables What's the advise here? Also, the DOF chart I used uses .1mm COC. I could change that if another COC is better. Advise?

Alan,

Sorry if I assumed you were making the mistake of using your ground glass to choose the camera position. It seems you've got a good method. I hope I didn't come off as "tech-splaining" things to you.

As far as tilts go: they help if you are able to use them to get your aperture closer to the best-corrected one for your lens, usually f/22. So, if you can adjust your movements to reduce the focus spread so that you can use f/22.6 instead of f/45, then you've made an improvement in resolution and will be able to enlarge that negative more. Whether tilts (or swings) help in this regard depends on the subject. In your case, you may have been able to apply tilt and shoot at f/22, which would have been a slight improvement over your f/32 choice (very slight, really unless you plan on really big enlargements). The only way to really judge if the tilt is helping is to apply it and see if the focus spread improves. If you're already in a situation where the focus spread is really small, like 1mm or a bit more, then you don't have to worry about improving and don't need movements.

And, concerning your choice of CoC: you'll notice that Q.T. Luong, in the article I referred you to, likes 0.066mm for his CoC. He explains why in the article. Basically, it has to do with being able to enlarge more. If you only make 8x10s from 4x5 negatives (and don't crop much) you could likely get by with a much larger CoC.

Best,

Doremus

[Alan Klein]

Alan,

Sorry if I assumed you were making the mistake of using your ground glass to choose the camera position. It seems you've got a good method. I hope I didn't come off as "tech-splaining" things to you.

As far as tilts go: they help if you are able to use them to get your aperture closer to the best-corrected one for your lens, usually f/22. So, if you can adjust your movements to reduce the focus spread so that you can use f/22.6 instead of f/45, then you've made an improvement in resolution and will be able to enlarge that negative more. Whether tilts (or swings) help in this regard depends on the subject. In your case, you may have been able to apply tilt and shoot at f/22, which would have been a slight improvement over your f/32 choice (very slight, really unless you plan on really big enlargements). The only way to really judge if the tilt is helping is to apply it and see if the focus spread improves. If you're already in a situation where the focus spread is really small, like 1mm or a bit more, then you don't have to worry about improving and don't need movements.

And, concerning your choice of CoC: you'll notice that Q.T. Luong, in the article I referred you to, likes 0.066mm for his CoC. He explains why in the article. Basically, it has to do with being able to enlarge more. If you only make 8x10s from 4x5 negatives (and don't crop much) you could likely get by with a much larger CoC.

Best,

Doremus

I appreciate yours and everyone else's help. Just to mention, that in addition to using the micro 4/3 camera to spot the tripod and frame the shot and to "see" in BW, I also use its zoom to determine the LF lens I need. It all saves a lot of wear and tear. Plus franky, I don't like framing upside down. I can't see aesthetically that way.

[Nodda Duma]

Why isn't diffraction considered an optical aberration? It's missing from the conventional list of aberrations.

Late night pondering...

Because it is not a geometric effect.

[Nodda Duma]

Drew... LOL. I just.... What you saw was the effect of turbulence, integrated over the exposure time of your camera and the response time of your eye.

Diffractive effects over a mountain top are too small to be resolved by the eye by several orders of magnitude.

Mark - Diffraction affects different wavelengths differently. I discovered this long ago in reference to the setting sun over sharp ridges where there is extremely clean air, first at Comb Ridge in Utah, then in Kauai. At first I blamed my highly-corrected new lens for the color fringing, but then saw it with my own eyes. The rays were truly bent differently over sharp geological edges in the distance. Something analogous happens with very small apertures, especially in high-magnification optical microscopy. There are very expensive sophisticated methods for correcting this. The primary wavelengths are completely separated, then more precisely re-aligned through electronically controlled magnetic mirrors - add about another $40,000, plus digital output afterwards. There might be similar technology in the latest astronomical equipment - add 400 million dollars, I'd imagine.

[Vaughn]

... Plus franky, I don't like framing upside down. I can't see aesthetically that way.

Probably just because you have not learn to yet. It takes time, one just needs more time under the darkcloth.

[Nodda Duma]

Agreed, Drew, but which other aberrations are ever corrected to perfection, not just reduced or avoided?

All of the basic, third order Seidel aberrations can and have been corrected completely. The Cooke Triplet does that. Higher order / residual aberrations, however...