I have been working on my techniques for sharp focus using similar methods to those listed on the Large Format homepage. Reviewing my photos from the warmer months, I think the most common problem I've had is having the foreground and background in sharp focus, but the mid-ground is often less than sharp. I've often tried to squeeze an f22 setting when I probably should have used even an smaller f stop. This under the belief that the smaller f-stops would significantly reduce my image sharpness. I was rather surprized to read that John Sexton used f32 in 50% of his shots in Listen to the Trees and f45 in 20%. I believe that the technical nature of this question is covered in Q-Tuan's article on "How to Select the F-Stop". I guess I'm wondering what is common practice and how successful the results have been. I tend to use a 16X20 enlargement as my print size standard for sharpness. Thoughts appreciated.

-- Roger Rouch (rrouch@msn.com), December 04, 2000

Answers

I'm in the process of setting-up a home lens testing bench at the present time. Preliminary tests of my lenses (old Schneiders and an Artar from the 50's as well as brand-new Schneider G-Claron, Apo Symmar and XL's) show that the resolution averages for an aerial image is in the range of 60-75 lp/mm @ f/22, 40-55 lp/mm @ f/32, and 26-35 lp/mm @ f/64. The results on film will of course be lower. I have to build a better microscope because I hit a ceiling at 140 lp/mm at the larger apertures - the lenses deliver higher resolution but the viewing system I'm using can't resolve it. Most lenses deliver their best resolution and contrast one to two stops down from wide open. Once you hit f/16 the drop in resolution and contrast is startling to see. And it only gets worse. Sexton is a master in the darkroom, so he can adjust to the contrast drop. As to the drop in resolution, well his prints aren't very big, I think that he rarely even approaches 16x20 (so he's only magnifying 3x-4x).

Wayne

-- Wayne DeWitt (wdewitt@snip.net), December 04, 2000.

The previously posted numbers were for on-axis only, off-axis being too tedius to perform until I have the new microscope. BTW I am aware that the upper figure for f/64 exceeds the "theoretical" diffraction limits - I'm hoping that an improved microscope will clear things up. Also interesting were the focus shifts that I witnessed - stopping down 1 stop on a f/5.6 lens required refocusing or the resolution numbers dropped 20%.

-- Wayne DeWitt (wdewitt@snip.net), December 04, 2000.

Crap - the numbers posted for f/64 were really meant to be for f/45 (gotta stop posting at night).

-- Wayne DeWitt (wdewitt@snip.net), December 04, 2000.

Roger,

I wouldn't be at all concerned about using f32 which I use probably 75% of the time in my landscape work. I haven't noticed any significant drop in sharpness at f45 with the majority of my lenses, either. Ultimately the effect of stopping down to 45 or 64 will vary from lens to lens.

Your problem with soft midgrounds may very well be due to not stopping down far enough and that is the first thing I would do in trying to correct the problem you're having. If it continues in spite of this then you may be selecting "improper" near and/or far focus points.

-- (mwphoto@nwlink.com), December 04, 2000.

Roger: Technically, the best way to see how lenses perform as diffraction increases due to smaller openings is to study a series of MTFs for each of the openings. I was fortunate to obtain a complete series of these MTFs for the Apo Ronar. In summary, and from memory, MTFs peaked at f22 and decreased slightly one stop either way but only marginally so. This fits in with your quote about John Sexton. At f45 the drop was quite noticeable but in my judgement, still quite useable. On this basis I would not hesitate to use f32 on 4X5 specially on enlargements only to 16X20. You did not specify format size. Because the allowable circle of confusion increases as format size increases, smaller openings can be used on larger formats due to the fact that the increased diffraction is balanced to some degree by the larger COC. I do not mean to be patronizing but after all is said, we all seem to be overly obsessed with sharpness including yours truly. Then, in sober moments I think of the many great photographs taken with lenses that by today's standards would be not good enough for point and shoot cameras. F45 on today's films and 4X5 all considered, including John Sexton, seems reasonable enough when there are good reasons for it.

-- Julio Fernandez (gluemax@ora.auracom.com), December 05, 2000.

I have been checking lens sharpness for years and this test has yet to fail. Compose a page of grided lines, varing the lines in thickness from extremly thin to thick rectangles (11 x 17 computer printout works great.) Set the grid page up parallel to the film back, focus only on the grided page and at a distance that is what you normally would have in the fore ground. Then burn some film. Run the full range of f-stops, one f-stop for each piece of film. Use post-it notes with the f-stop printed in large marker so you don't get lost on which is which. With a 10x to 20x lope you will know what stops are sharp and what are not by looking at the thinner lines or the line edges where black meets white. If you only enlarge prints to 3x to 4x a 5x to 10x lope will work. So on and so on. Keep things simple; set up tests that relate to actual field use, lab use and final print size. It works. I photograph with an 8 x 10, there are many occassions when I plan on contact printing and absolute sharpness is not a factor (useful when extreme depth of field problems occur.) f-90 contacts well most of the time and enlarges like . . . . For enlargements, my 355mm G Claron doesn't even get sharp on infinity until f-32 where my 210XL goes to krap at anything past f-32.

As for having the foreground and background in focus with the mid-ground less sharp you are probably using too much tilt. Try finding the spot where the fore/back ground are in approximate focus as compared to the mid-ground focus and stop down to acheive sharp focus throughout. If you have to stop down to f90 to get both fore/back and mid ground in focus you probably have too much foreground in the frame or the tilt is off.

-- Anthony Howell (howell@zianet.com), December 05, 2000.

Roger,

If I understand what you've written, you're discussing a situation in which you are employing tilt. Otherwise, I don't see a way in which the foreground and background could both be sharp, and yet the mid-ground be out of focus. Soooo...

The problem may not be entirely with your selection of f stop. It may be related to the places in your composition that you chose when you are focusing. The one with the most effect on the mid-ground is usually your far focal point. At the risk of telling you something that you already know well, let me say that everything in the horizontal plane (assuming tilt) connecting your near focal and far focal points will be in focus. At the intersection of this plane and the plane containing your film, the depth of field will be zero. It increases with distance from this intersection. The rate of increase is more rapid than linear with distance, but to make the discussion easier, let's assume that it is linear, and that it extends equally above and below the plane of focus.

If your composition is a field of flowers and a mountain, your far focal point should not be the summit. If it were, the field of flowers might lay below the wedge defining the depth of field. Decreasing the aperture increases the depth of field at any given distance, and so might bring the flowers into focus by using f45 instead of f22. Alternatively, focusing on a lower point on the mountain would also bring the flowers into the wedge of focus. Because the mountain is further away where the wedge is wider, you might also find that the summit remains within the wedge, too. I would try something about 1/3 up the mountain's side as my point of distant focus.

I find this particularly frustrating is when my composition includes the ground at my feet, a distant mountain, and an intervening valley. The distant focus needs to be quite low on the montain flanks to keep the slope from your feet into the valley in focus. An aperture of f45 would not be unexpected.

Finally, let me say that I sympathize with your problem. On a recent trip, I found myself resisting using very small apertures and ended up with a number of compositions with the mid-range out of focus. It just doesn't seem reasonable that you should have to resort to f45 or more when you're tilting one of the standards. Well, if the composition contains significant vertical relief, you may have no choice. Or you can re-compose. Personally, I've done many compositions at f64, and only rarely has it been apparent with a 4x loupe, or when printed at 16x20. I have had editors make remarks when everything is not in focus because it is assumed that it will be because it's a 4x5 camera. Beauty is in the eye of the beholder.

Hope this helps, and that it wasn't totally off the mark.

Bruce

-- Bruce M. Herman (bherman@gci.net), December 05, 2000.

I'm using Schneider 90XL, 150XL, 210 apo symmar, and Nikkor 300M. The 150XL (and maybe 210) are the only two lenses that are sharp between f32-45. I'd give the edge to the 150XL..could be because its image circle covers 8x10, so I'm using the "sweet spot" all the time w/ 4x5. The 90XL is quite good at f22 or less (don't shoot it enough at f32 since it has tons of DOF). The 300M is terrible below f16, best at f22, acceptable at f32, and dismal otherwise. I realize a 300/5.6 plasmat would give superior sharpness since I'd be using the center of the image circle where the MTF is 0.9, but the weight...

-- James Chow (dr_jchow@yahoo.com), December 05, 2000.

I believe the information each of you folks have posted so far has some usable tips. Thanks. Digressing slightly, I might add that the situations where this is most obvious is indeed where an intervening valley or depression separates the fore- and background, or where the background is especially vertical and at a moderate distance. And also where tilts are used to reduce the focus spread between near and far focus. By following the techniques I've learned, where the spread in focus from near to far is minimized using tilt and the F- stop is estimated by the near-far spread, I've found that a further adjustment seems to be required by then getting sharp focus on the mid-ground. This, in the types of situations I've described, increases the focus spread and thus the indicated F-stop. Only late in the "warm months" did I experiment with this final adjustment and I did venture into the f32 area. The results in these cases seemed improved. I've yet to try f45.

-- Roger Rouch (rrouch@msn.com), December 05, 2000.

An after thought - would it be that lenses with a larger image circle - as in greater coverage would in general tend to give better results at these small f-stops? It is interesting that the lens tests I've seen tend to stop at f22.

-- Roger Rouch (rrouch@msn.com), December 05, 2000.

In response to your last questions. The lenses with a larger coverage may give better results due to a flatter field when the image field is smaller than that for which it is designed ("sweet spot"). And as far as publishing lens test results for apertures smaller than f/22 there's really little point in doing that. By f/22 diffraction is the dominant abberation in any decent lens - publishing such lousey- looking numbers (which are an unavoidable fact of physics) doesn't make a good impression on the less-informed consumer, and really doesn't tell us anthing of use.

-- Wayne DeWitt (wdewitt@snip.net), December 05, 2000.

I have always wondered this: if f/16 or f/22 is the sharpest aperature-setting on most large format lenses, why was "f/64" adopted by Ansel Adams, Edward Weston, and certain other photographers of the "f/64 Group" in the 1930's as the ultimate expression of sharpness and resolution in a photograph? Or were the members of this group more interested in the extended DEPTH-OF-FIELD specifically, afforded by an f/64 aperature--and that this outweighed any loss of sharpness and resolution that may ensue from stopping down so far? Or is f/64, in fact, the sharpest aperature on most 8 X 10 lenses (as opposed to 4 x 5 lenses)? Or was it just deliberate exaggeration on their part, for literary effect?

-- Nick Rowan (nrowan@vf.com), December 06, 2000.

Nick - Those photographers used 8x10 cameras (as well as odd-sized formats that are no longer around) and mostly made contact prints. Once you start using the long focal length lenses common in 8x10 depth of field decreases dramatically - they chose the smallest f/stops usable that would still give acceptable sharpness. I've continued my testing and for the lenses that go to f/64 I have the following results on a 5x7 format (aerial resolution - results on film may differ). 150 Super Symmar XL: C/E- 23/18, 210 Apo Symmar: 23/19, 150/265 Convertible Symmar @265: 23/10, 305 G-Claron 20/20, 450 Fuji 12.5C: 22/22.

You can see that the resolution figures are for the most part the same - diffraction has done it's dirty work. Even @ 20lpmm though you have room to enlarge 2x-4x (depending upon your tastes) and get an acceptable print. Don't forget, these figures are for the plane of the focused image, once you induce the losses for divergence away from this ideal plane the resolution goes even lower. In response to Adams and his friends there was another group that called itself f/128 - wonder what happened to them?

Wayne

-- Wayne DeWitt (wdewitt@snip.net), December 06, 2000.

My solution to the depth of field problem springs from the saying "you can not be too rich, too thin, and you can't have too may film holders". If in doubt I just shoot an extra sheet of film at a smaller aperture (exp. f32). At worse you have wasted a sheet of film and learnt a little. At best you will be thankful later because the first sheet did not have enough depth of field. Let's face it, compared to the gasoline, cost and effort it took me to get there, one more sheet of film isn't that much to pay for a little insurance. Go ahead live it up a little. And yes, get a few more holders.

-- Pat Raymore (patrick.f.raymore@kp.org), December 06, 2000.