Lens resolution question

[felipemorgan]

I am trying to plan for an upcoming photo assignment in which I will be photographing a landscape subject about 3 to 4 miles distant. The subject will be about 3.5 miles of wooded hillside with some houses. The final result will be film images that will ultimately be scanned and combined with GIS data and renderings. The scanned image needs to be able to clearly resolve houses and the houses need to be sufficiently clearly rendered on film to allow details like windows, rooflines, etc. to be apparent (drum scans are in the budget). The filmstock will most likely be Provia or the new Velvia, and I anticipate using a high-quality polarizing filter if necessary.

I'm used to thinking about lens performance in terms of published specs like line pairs per mm, etc. But in this case, I don't know how much performance is enough. I would prefer to use medium-format for this assignment, and I believe that since the images will be scanned, stitching them together is a strong possibility.

So where is the "sweet spot" of focal length/lens resolution/film format that meets my criteria? Although this is my first foray into high-resolution photography, I am aware of the basics like the importance of rock solid camera support, good lens shading, etc.

Thanks for any input,
--Philip.

 

[Dan Fromm]

Hmm. You'll have to separate details perhaps two feet apart at a distance of 4 miles. You want to capture a scene 3.5 miles wide. All with one shot. On medium format, probably nominal 6x6.

Draw the triangle. Opposite leg = 1.75 mi. Adjacent leg = 4.0 mi. Tan(half the angle) = 1.75/4. So the angle your lens needs to capture to do it all in one shot is 2 arctan(1.75/4) = 47 degrees. Basically you'll need a normal lens to capture the scene in one shot. 80 mm if you're shooting 6x6.

Resolution needed? 2 feet is to 4 miles as 2 feet is to 21000 feet is as x mm is to 56 mm. So x = 112/21000 = .005 mm. 1/.005 = 200. So you need a lens-film combination that will resolve 200 lp/mm if you're going to do the job with a 6x6 camera.

You're cooked, even before we start worrying about atmospherics.

You'll need a considerably larger piece of film, sharper film too. You can't count on getting more than 50 lp/mm across the field, should be conservative and plan on no more than 40. So you'll need a piece of film 5 times as big as 56x56, i.e., 11" x 11". 8x10 will about do, with a good sharp 300 mm lens that will cover the format at infinity, clear and very still air. Or 11x14, with an 18" lens. Shoot TMX.

Good luck, have fun, learn basic arithmetic,

Dan

 

[df cardwell]

Philip, my friend, go try the shot as soon as you can. Stop down 2 or 3 stops, shoot on a cool day, when there is shadowed light on your target.

See what you get. Theory is one thing, this is going to be about how much the rules can be ignored.

 

[jimgalli]

Dan hinted in his usual kind and understanding way about atmosphere. It really is your biggest problem. Where I work we have to do focus runs on our telescopes. On a good day you may get 15 minutes in the AM and 15 minutes in the PM when your best lenses resolution is meaningful. On a bad day you may just get skunked. Get some very high powered binoculars so that you can watch the heat waves. When they're gone, that's when all the rest of the resolution starts to play. Good luck. Watch Half.com for a book called "Image Clarity" by Williams.

 

[felipemorgan]

Thank you all! Dan: thanks for boiling down the math to a comprehensible form. Here's what I got from your math:

Lens/Film system resolution: flpmm (50 lp/mm)
Smallest resolvable detail on film: fresolv=1/flpmm (0.02mm)
Film Long dimension: fdim (56mm)
Scene Long dimension: sdim (5631.5 meters)
Smallest resolvable detail in scene: =(fresolv/fdim)*sdim (2.01125 meters)

So it seems to me that cranking up the focal length of the lens to 150mm on 6x6 would (on the theoretical plane) create a situation in which I am photographing a scene .75 miles wide (at 4 miles distant) and with 50 lp/mm of film/lens resolution available, I could resolve objects .43 meters wide. Does this sound right?

As I stated in my first post, the images are going to be scanned and I can stitch multiple images to obtain the final result. (The images must be color)

So I have another reality-check question: Are my chances better with stitched 4x5 or stitched medium-format? I'm concerned that it will be difficult enough to find a time in the location when wind and atmosphere are cooperative...

Thank you DF for the encouragement to just do it and see what happens!

--Philip.

 

[smieglitz]

Philip, Give Kodak (1-800-242-2424) or Lee Filters or Tiffen a call, talk to a technical rep, and see what sort of haze-cutting filters may be available for such an application. I used to own Wratten aerial haze filter gels HF-3 and HF-5 and something of the sort may still be available. I never used my filters and recently gave them away so I can't say for sure whether they would be satisfactory, but I think you'll need something of the sort to cut through the atmosphere as much as possible.

Joe

 

[JBrunner]

Size always helps. All other things being equal (frequently they're not) 4x5 has a ton more information for a given composition.

 

[MAGNAchrom]

The final result will be film images that will ultimately be scanned and combined with GIS data and renderings.

Since you are scanning, would you consider a stitched panoramic? Consider: six or seven 6x7 or 6x9 (oriented vertically) scans will give you approximately 10" width panorama. If this would work, then a good, sharp 250mm medium format lens might do the trick.

Or alternatively: rent a Fuji 6x17 with a 300mm lens. Might do the trick.

 

[felipemorgan]

Since you are scanning, would you consider a stitched panoramic? Consider: six or seven 6x7 or 6x9 (oriented vertically) scans will give you approximately 10" width panorama. If this would work, then a good, sharp 250mm medium format lens might do the trick.

Or alternatively: rent a Fuji 6x17 with a 300mm lens. Might do the trick.

I'm definitely considering this option. The chances of success seem higher...

The Fuji 6x17 is a tempting option but my local rental shop only seems to have a the 105mm lens for it...

--Philip.

 

[Helen B]

Do you need colour? Do you need full pictorial tonality?

Best,
Helen

 

[HerrBremerhaven]

One shot with an 80mm could be matched by four shots with a 150mm (allowing some overlap) on a 6x6 camera. Shoot during noon sun at no smaller aperture than f11.0. A Zeiss 150mm Sonnar would be a good choice, though I think looking at slightly larger Mamiya, like the RZ67 or the 7 II rangefinder lenses might get you nearly the same capability.

Resolution of colour transparency films are not as capable as for some B/W films. However, a somewhat interesting test from Leica Fotographie International (LFI) by Erwin Puts a few years ago had him getting just better than 90 lp/mm from two films. Those were Kodak E100G and Fuji Astia 100F, both of which are available in 120 roll film sizes. Both are also extremely small grain, so probably your better colour film choices for drum scanning.

As others have mentioned, atmospheric issues can make this tougher. Filtering and time of day can help, but this will still not be easy. Plan on needing several shots, then picking the best. Despite the distance, you might actually want to bracket focus just a little.

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

You might be interested in this test. Though using B/W film, it indicates some possible real world achievable limits. Zeiss sometimes claim the highest limits, even beyond some film test data from Kodak and Fuji. Erwin Puts claims real world results, but I would think that somewhere near 60 lp/mm to 80 lp/mm is more realistic than 80 to 90 (or 100) lp/mm.

Another idea is to shoot the scene with B/W film for the high resolution. If you really need colour, you can shoot another set of images in colour, scan both B/W and colour, then combine the lower resolution colour information onto the B/W high resolution images. Expect a few tries to get this correct. I hope you are charging quite a bit for this, since it will be a great deal of work.

Ciao!

Gordon Moat
A G Studio

 

[felipemorgan]

Do you need colour? Do you need full pictorial tonality?

Best,
Helen

I need to satisfy the client's needs. They want/need color... I do need full pictoral tonality also.

--Philip.

 

[Dan Fromm]

Thank you all! Dan: thanks for boiling down the math to a comprehensible form. Here's what I got from your math:

Lens/Film system resolution: flpmm (50 lp/mm)
Smallest resolvable detail on film: fresolv=1/flpmm (0.02mm)
Film Long dimension: fdim (56mm)
Scene Long dimension: sdim (5631.5 meters)
Smallest resolvable detail in scene: =(fresolv/fdim)*sdim (2.01125 meters)

So it seems to me that cranking up the focal length of the lens to 150mm on 6x6 would (on the theoretical plane) create a situation in which I am photographing a scene .75 miles wide (at 4 miles distant) and with 50 lp/mm of film/lens resolution available, I could resolve objects .43 meters wide. Does this sound right?

As I stated in my first post, the images are going to be scanned and I can stitch multiple images to obtain the final result. (The images must be color)

So I have another reality-check question: Are my chances better with stitched 4x5 or stitched medium-format? I'm concerned that it will be difficult enough to find a time in the location when wind and atmosphere are cooperative...

Thank you DF for the encouragement to just do it and see what happens!

--Philip.

Great! You've got the idea. If you can't go up in format you'll have to shoot at higher magnification, i.e., with a longer lens. Stitching, as you are willing to do, will save you from having to use huge gear. I think, though, that you may want to separate details that are less than .43 meters apart.

Ponder Jim's point about atmospherics. By all means buy a copy of John Williams' book, it will discourage the daylights out of you. I have it and every time I open it I lose heart.

Last point, getting the best that your lens(es) and film can give requires absolute steadiness too. Very solid tripod, cable release, mirror locked up, and so on.

Years ago Modern Photography magazine published a piece on whether it was possible in practice to attain 100 lp/mm. They concluded that it was just possible on Tech Pan with meticulous technique. Impossible with any of the color films available when they shot, including KM. Solid tripod, as mentioned above, also bracketed focus and shooting at the lens' best aperture. For the gear they used, "slow" normal lenses and normal length macro lenses on 35 mm cameras, best aperture was around f/5.6.

I think Don Cardwell's point about breaking rules is misplaced. You're going to the limit of what's possible. You may find that you can't quite meet your requirements.

Good luck, have fun,

Dan

 

[df cardwell]

DAN

Two points-

My advice had to do with the necessity of seeing for himself what the product of best technique was going to be. It's one thing to 'imagine it' based on numbers, it's another to see it. It would be handy to know, for instance, what time of day the sun will be shining on the target. And for a Portland photographer, it may be an interesting problem finding that elusive clear, blue sky. Is this a morning 'sweet light' shot in a foggy area, or not ? Handy to know, ASAP.

Secondly, breaking the rules is absolutely essential because the only hope of pulling this off is in the realm of ' the illusion of sharpness '; beyond theoretical resolution, what microscopists call 'empty magnification'.

From experience, I'll predict it's a toss-up whether 35, 120, or 4x5 is better for this job. Film flatness and focus accuracy will offset or fulfill the numerical advantages of big sheets of film. I have a clear glass reticle on a Linhof focus screen, with a modified microscope eyepeice for a loupe. Do normal people ? No.

Would 4x5 plates be an advantage ? You bet. So would being able to shoot in a vaccum. Or an immense oil immersion lens

Where the formulas can't predict results is in the specific manner the image degrades. Calculating the Resolution of an optical System is a cautionary experience that will forever change how one looks at telephotography, but interpreting it accurately demands more nuanced information. It's handy to know, for instance that Velvia 100 can deliver no more than 5o lp/mm at 50% contrast. Incredible as that is, no objective is going to do better than that, which results in an image with 25% contrast at 50 lp/mm. As we are driven from theoretical perfection, that contrast fiigure drops.

Does a Leica M system, with higher contrast at a given resolution target, greater focus accuracy, and less susceptibilty to lens movement during exposure, offer advantages ? You bet it does. Can a 4x5 system be prinked up to deliver close to its full potential ? Sure can.

Can you ever get past the reality that 35mm with shorter lenses, will be shooting at f/5.6 and 4x5 will be shooting at f/11 or greater ? Longer exposure is exactly that, longer exposure to heat waves, and image blur.

Were it me, I'd dig up an aero camera ( see Peter Gowland's camera below ) use some sticky tape in my Linhof holders, and be sure the film plane is exactly right.

And I'd shoot my Leica as a backup.

Increasing the effective mass of the camera buy bolting it to a piece of brass or steel , THEN to the tripod is handy. And bags of shot are always handy.

And some practice. Theory suggests it can't be done. He'll have to see what he can get away with.

.

 

[Dan Fromm]

Don, we're sort of in agreement. But since the original poster is open to stitching, what he wants to accomplish is quite possible although impossible with the setup he first described.

Fellow I used to work with was an odd sort of professional photographer; he looked like a pilot, walked like a pilot, flew like a pilot, but the planes he flew were SR-71s and, later, A-11s. Basically very fast-moving cameras. His work product wasn't flying a plane from one place to another, it was good-quality images of some of the ground he flew over. The cameras he operated would easily do what the OP needs. Unfortunately they're not available to the OP.

The question the OP needed answered was, what gear can he can rent or buy that will do the job for him. The point of the calculations I showed him was to guide him to a better understanding of the problem. From there, working what will do isn't hard. I reckon he'll need at least a 300 if he shoots 6x6, and there are some Zeiss lenses that long, longer too, that are probably good enough.

Getting the shots, though, is going to be very difficult because of steadiness problems and atmospherics. You're right that he should survey the situation, consult climatic tables, and so on to get a better idea of what he's up against.

Cheers,

Dan

 

[df cardwell]

I suppress my foto-geek tendencies as well as I can.
But when the rules don't fit the mission,
change the mission.

I'm thinking a Bell 500 and KenLab stabilizer.

The Sr71 is overkill.

 

[David A. Goldfarb]

This camera seems designed for this task--

http://www.linhof.de/english/kameras/metrika/metrika.html

Maybe Bob Salomon could tell you if anyone rents it in the US.

 

[phfitz]

HI there,

I think David found your answer, now for a heavy metal plate to attach it to. For films check:

Kodak Aero Films

df cardwell: "The Sr71 is overkill."

I take serious exception to that, it looks right , proper and correct to me. Actually, it's why we never had WW III, you only fear what you do not know and with this, we knew EVERYTHING.

 

[df cardwell]

David's got the answer.

As for the Blackbird, well, OK. How much J-7 a gallon now ? That's the client's expense.

As for WW III, perhaps that's a little beyond the charter for THIS forum.

 

[naturephoto1]

David's got the answer.

As for the Blackbird, well, OK. How much J-7 a gallon now ? That's the client's expense.

As for WW III, perhaps that's a little beyond the charter for THIS forum.

That may be true, but have to admit that the SR71 was some kind of a phenomenal airplane.

Rich

 

[df cardwell]

never disputed it

 

[phfitz]

"As for WW III, perhaps that's a little beyond the charter for THIS forum."

Why? Large format, high resolution FILM photography saved the world several times, the SR-71 was the camera and it looked damn good doing it.

Smile

 

[Dan Fromm]

Um, er, ah, before we go overboard patting each other and film on the back, be aware that digital image capture has pretty well displaced film for military aerial photography. At least in the first world, i.e., NATO. Why do you think so many Agiflite cameras have been sold off as surplus?

 

[Sparky]

This camera seems designed for this task--

http://www.linhof.de/english/kameras/metrika/metrika.html

Maybe Bob Salomon could tell you if anyone rents it in the US.

Actually - the lenses on those puppies aren't THAT much to write home about resolution-wise. They just have a known (and specified) distortion. That's all you need to do metric work.

 

[Sparky]

Um, er, ah, before we go overboard patting each other and film on the back, be aware that digital image capture has pretty well displaced film for military aerial photography. At least in the first world, i.e., NATO. Why do you think so many Agiflite cameras have been sold off as surplus?

Yeah, it's weird huh? How do you think they're getting the resolution? Maybe a betterlight back and they do separate aerial passes for each of R G and B...!