How is use of f128 possible/practical?

[Robert McClure]

I never grasped why the Apo-Nikkors, for example, were designed with ability to stop down to f128. What about defraction at that small of an aperture? Something to do with the graphic arts application? Attempt to imply greater control to the graphic arts cameramen? Something related to 1:1 shooting? Can someone please explain this to me? Never really understood. Thanks!

[Nick_3536]

I think the guy just kept going until he got bored -) I think they were always used at F/22 or F/32 depending on focal length.

[Robert McClure]

Thanks, Nick.

My curiosity still remains: if not useful/practical, why on earth an f128?

[Kevin Crisp]

Uh....my amp goes to "11"?

[Robert McClure]

Thanks, Kevin!

However,

I would trade a boatload of tongue-in-cheek remarks for one cogent explanation as to

WHY APOS STOP DOWN TO A SEEMINGLY UNUSEABLE F128.

Robert

[Kevin Crisp]

Since they were generally used for photographing things that are flat, requiring almost no depth of field, I have to say it stumps me too. Why do so many view cameras have swings and tilts so extensive as to be unusable? Maybe somebody else made one that would do that and the feeling was the feature had to be there so as not to come up short on features and capabilities? Tongue in cheek perhaps -- but I was making a point.

[GPS]

Graphic arts cameras are more "distortion free" than other cameras due to they special construction and use of the special glasses. They are constructed for use with a very small angle of view where they can keep their very high resolution under highly specified optical conditions .Therefore they can use smaller apertures that would otherwise kill the more "normal" camera lens. Such is my understandig.

[Nick_3536]

Thanks, Nick.

My curiosity still remains: if not useful/practical, why on earth an f128?

If you look at different brands some only go to F/32? Others F/45. Think about the other way to. Why do they open up so far? If none of them were used wider then F/22 why not have them set to F/22 wide open?

I still think it was just a reflection of the people making the lenses being able to do it more then any actual need for the lenses to have such wide F/stops.

I think a few people here have actually used the machines these things were designed for. Maybe one will pipe up.

[Ed Richards]

Just think of f128 as a built in pinhole camera.

[Patrik Roseen]

Lenses being designed that way...?
I guess it's more related to the available barrel into which the lens elements are mounted. The mechanics of the aperture in the barrel will allow the physical aperture-hole to reach a minimum size but the resulting aperture will depend on the focal length of the lens elements inserted into it.

It's no more strange than that the largest aperture would be f11 compared to other lenses that would start at say f5.6 and end at f45 in the same barrel.
The lens would still be used at say f22 as have been stated before.

[Ole Tjugen]

When using a high contrast film with a high contrast developer to produce negatives with only pure black and pure white, a certain degree of softness is often desirable.

As an experiment take a screen copy of a bit of text from this page. Import it into photoshop or whatever, and enlarge it 10x: Lots of blocky pixels, right? Go back to the original, convert it to 256 shades of grey, add a touch of Gaussian blur, the enlarge it 10x and convert it back to 1 bit black and white: The text is far more easily readable (unless you've used too much blur, of course).

For similar reasons many repro lenses have a "waterhouse slot". This is not for aperture stops, but for specially shaped apertures which can contribute to making a cleaner copy.Here (http://cgi.ebay.co.uk/Industar-11m-1-9-F-300-mm-620286-For-cameras-The-larg_W0QQitemZ7597968789QQihZ017QQcategoryZ30076QQcmdZViewItem) is an example of a repro lens with aperture slot, complete with a set of shaped apertures.

[Tom Duffy]

In a contact print, f128 is a usable aperture.

[Struan Gray]

The small apertures are for use when making halftone copies of continuous tone originals. The common way to do this pre-computers was to project the original using the repro lens onto a piece of graphics-arts film over which was placed a cross-ruled glass screen. The screen turned the projected image into a series of tiny dots, with each dot's shape being a copy of the lens' aperture. Fine screens require small dots, so lenses for these applications often stopped down much further than taking lenses or enlarging lenses.

FWIW, screened colour seperations required odd aperture shapes oriented to different angles for each ink, hence the Waterhouse slots with angular scales which are also often found on repro lenses.

[Brian Ellis]

Maybe the lens designers thought as I do that diffraction isn't a practical problem with 4x5 film until the enlargement factor gets up to at least 20x24, maybe larger. So f128 is certainly usable and in a given situation might be the only way to obtain the desired depth of field. Or even if the effects of diffraction were noticeable at f128, many photographers might decide that they'd rather suffer a minor loss of sharpness due to diffraction rather than a major loss due to inadequate depth of field.

[Jim Jones]

I can't find the citation at the moment, but believe Edward Weston had an inexpensive Rapid Rectilinear with the iris modified to stop down to f/256. In some of his contact printed macro photos, depth of field may have been more important than high resolution on film.

[Jim Galli]

You could tilt the table to correct other problems like converging lines that shouldn't have been. Stop to f128 to get the dof. You still have about 10 lppm at f128. 10 is the magic number to fool the human eye, especially if the viewing distance is greater. Likely not used all the time but useful once in a great while. Like the other 185 horsepower we really don't need and almost never use.

[Kirk Fry]

It has a special use in with screens for printing in graphics arts. I can't remember the details, maybe someone who did this for a living could explain.

Kirk

[Emmanuel BIGLER]

a piece of graphics-arts film over which was placed a cross-ruled glass screen.
Struan, can you elaborate on that : do you mean that the crossed gris was placed at a certain distance from the film, and that an actually defocused "gridded" image was recorded on film ?
My feeling is that if the grid was just pressed against film without any gap and if the focus was made on film I do not see how the image of the iris could be projected.

BTW I have made myself, in an attempt to get better xerox copies, some "gridded/binary images from a halftone negative, with a conventional enlarger, simply by laying a speclal mask (made by Agfa, probably a slliver halide copy of a master mask) on top of a steep-graded B&W paper. The mask transformed very simply gray levels into dots of variable size, but the mask itslelf was halftone. The periodicity of the dotted grid in his process was simply imposed by the grid on mask and the lateral size of each dot imposed by the input density level.

So if I understand well the process described by Struan involves some kind of high contrast/no halftones grid instead of the halftone grid I used under an enlarger ?

[paulr]

struan's answer is intriguing. i'd be curious to hear more.

there might also be a simpler reason to make a lens that stops down too far ... because you can. it's not like making a wider maximum aperture, which requires redesigning everything and making a million optical compromises. A smaller minimum aperture just means letting the blades close more. Doesn't cost anything, sounds cool, and lets the end user deside how small a hole is too small.

[GPS]

Speaking about process lenses Sidney F. Ray states - "Small working apertures are also necessary to give a practical depth of field at the copy holder and to avoid the setting of impossible mechanical tolerances on flattness and alignment". Also - "a process lens need resolve no more detail than about 16lp/mm" (from Applied photographic optics, Focal Press 1988 p. 297).

[Eric Woodbury]

Just guessing, but it is probably just a matter of that lens in that shutter and that's what you get. Not that anyone wanted f/128, but the shutter could go that far, why not mark it?

[Robert McClure]

So, could it be said, therefore, that unusually tiny apertures (like f128) had true practical value only when it came to graphic arts applications?

Further, could it be said that most large format lens manufacturers (arbitrarily) decide to not go past, say, f90 because they feel the "image degradation" will generally be unacceptable to most users (even when contact printing the negatives)?

One local used photo equipment dealer had once told me that the idea of re-working a lens to allow stopping to f128, for example, even though it was said that Weston did it, was utterly foolish.

"The manufacturer has already determined in testing that diffraction would be far to severe at those tiny apertures!" he chided. If recall feeling like a real putz for even considering the idea.

I later observed that my 355 G-Claron, though having markings on its Copal 3 aperture ring only to f64, can easily allow stopping down to probably f360. I guess I will have to be the one to decide how much diffraction I can live with!!!

Thanks everyone for your great and informative input. Very, very interesting and stimulating!

Robert

[Struan Gray]

The cross-ruled screen acts as an array of pinholes, so each halftone dot is a pinhole image of the aperture. For optimum results you need to space the screen the right (small) distance from the graphics arts film.

Everything I know about this I picked up from posts by Jean-David Beyer and Richard Knoppow in the usenet group rec.photo.equipment.large-format. Credit where credit is due, and more detail, here: http://tinyurl.com/qdbq3.

[GPS]

So, could it be said, therefore, that unusually tiny apertures (like f128) had true practical value only when it came to graphic arts applications?
Robert

That's all what there is to it, I'm afraid - except if you have the same purpose with another lens.

[Jim Galli]

Robert, I have an ancient Wollensak f10 process lens of 19" (protar 8 elements 2 groups type) that easily covers the 12X20. I have stopped it down to f90 1/2 often but that seems to be a practical limit for me. Beyond that and I consider even the contact prints too soft from diffraction.

[Capocheny]

Maybe the lens designers thought as I do that diffraction isn't a practical problem with 4x5 film until the enlargement factor gets up to at least 20x24, maybe larger.

Brian,

So, I've been worrying about this for nothing all this time! :(

Since 16x20 is the largest I normally enlarge to... this is even less of an issue!

Thanks for this great bit of information! :)

[BTW... you missed a fun workshop! :) Next years is slated for Sequim/Port Angeles... in your neck of the woods I believe. :)]

Cheers

[Robert McClure]

Pretty cool to learn from you guys. I will content myself with the "arbitrary" aesthetic of "tack sharp" images. That said, some of Jim Galli's images were pretty awsome.

Robert

[Emmanuel BIGLER]

Thank you Struan for the reference to the historical crossed-grid technique to make printing plates from a halftone image. Fascinating ! And certainly very difficult to properly adjust.

[Roc Chan]

The subject of small F stops on Graphic Art Lens came up several times over the years.To controll contrast of graphic art photo sensitive materials,a flash,non subject(white light) exposure is use ot "fog the film".This exposure is use to boost one zone to the next without affecting the adjecent zones.This Pre,Post flash exposure are small fraction of the subject exposure,thus small F stops are use.In later Agfa Repromaster ,the flash exposure occures in the bellow between the lens and the film.

[Michael Graves]

Aside from all of the other practical considerations mentioned by various people; isn't it possible that f128 came about for the simple reason that the manufacturer used the same diaphragm assembly for a 12" lens as they did for an f64 24" lens?

[sanking]

Robert, I have an ancient Wollensak f10 process lens of 19" (protar 8 elements 2 groups type) that easily covers the 12X20. I have stopped it down to f90 1/2 often but that seems to be a practical limit for me. Beyond that and I consider even the contact prints too soft from diffraction.


I have observed this as well. In theory stopping down to f/128 should give resolution of 10 lppm, and even at f/256 you should get 5 lppm. Since 5 lppm is considered adequate for contact prints it should be possible to effectively use apertures of f/127 and f/256, if the final goal is a contact print. But as a practical matter I find that stopping down beyond f/90 gives contact prints that to my eye are too soft.

Sandy

[Mark Sampson]

25+ years ago I worked in a custom lab; among other things, I worked on the process camera making copy negatives and such. (No graphic arts work.) One of my tasks was to make b/w internegatives from color trannies; for this we used (ASA 200) Super-XX film. For some reason I can remember that the exposure was 4 seconds at f/45. Smallest f/stop I ever used with any regularity, now or since. Everything else was set between f/11 and f/22, probably to get convienient exposure times. In those days I was ignorant of things like diffraction- but following this thread makes me want to shoot a photo with the lens stopped all the way down, just to see what will happen.

[Alan Davenport]

I'm feeling lazy right now, so I'm not gonna do any math. But:

1. Diffraction is present in every photograph ever made. It rears its ugly head by making (what should be) points on the film become blurry discs. The blur circle of a point in subject space is called, on the film, an airy disc.

2. The size of the airy disc is the same at a given f/stop, regardless of focal length. In other words, the physical size of the diffraction disc on the film will be constant for any lens at f/128. IOW, the diffraction blur -- on the film -- of any lens, anywhere, when at f/128, will be the same size.

3. As long as the airy disc is no larger than the acceptable circle of confusion, diffraction is not a problem.

4. It's generally accepted that the allowable CoC becomes larger with increasing film size. This is both because there is less enlargment required, and because the viewing distance tends to increase with larger prints. A commonly used CoC value for 35mm film is something like 0.03mm, while on 4x5 film a common value is 0.1mm. Larger films can tolerate more blur because they don't have to be enlarged as much.

Therefore: at some point on the scale of increasing film size, there should be an acceptable circle of confusion which is larger than the airy disc produced at f/128. You can use f/128 on that size film without degradation!

[Jim Jones]

The one 11x14 enlargement I have that's made from a 5x7 negative shot with a decent lens at f/128 is unpleasantly unsharp. So are many transparancies or enlargements from 35mm shot at f/16 or f/22. Much depends on the subject, though. Images with severe diffraction limiting might be satisfactory when the subjects are clouds or nudes.