Comparing Image Circle Specs

I am trying to compare image circle specs for two different lenses: Schneider

47mm SA (not XL) and Rodenstock 45mm APO-Grandagon. The Rodenstock data sheets

provide image circle data for f11, which is reasonable since it is the

preferred working aperture. However, the Schneider lists image circle data for

f22, which I would never use on such a short lens due to diffraction. Is there

a way to compare these two? Is the image cicrcle at f22 larger or smaller than

at f11?

First, the 47mm Super-Angulon is strange, at least to me. Schneider made a F5.6 and an F8 version early on. Further, there may have been two different F5.6 versions. I have an early F5.6 and have used it on 4x5.

 

It does not cover 4x5. It comes close, but you must consider other factors such as the kind of fall-off a lens has. I'm too dim to understand the specs given by manufacturers, but I know what I've used and seen. Better luck to you with the factor charts.

 

Frankly, I'm done with Schneider lenses. If I could afford it, I would repurchase nothing but Rodenstock today.

The image circle increases as you stop down. It is hard to design a lens with a very large image circle because some aberrations are worse off-axis. As you stop down, some aberrations are reduced and the off-axis image improves, so the image circle gets larger. (Eventually, if you stop way down, diffraction blurs the image, even in the center.)

 

Many times starting at some f-stop, as you stop down further, the image circle doesn't change much. But near wide-open the image circle is likely to change significantlly with aperture. So unless you experiment with the lens, or some user reports here, since you want the image circle at a wider aperture than Schneider has quoted the image circle for, you can't be sure what it will be.

 

What format do you want to use one of these lenses for? The non-XL 47 mm Super-Angulon is a much older design than the 47 mm Super-Angulon-XL or the 45 mm Apo-Grandagon, recommended for 6x9 cm (http://www.schneideroptics.com/info/vintage_lens_data/large_format_lenses/super-angulon/data/5.6-47mm.html). It was made long enough ago that there was a single-coated version. The 47 mm SA-XL and 45 mm Apo-Grandagon are much newer lenses with expanded coverages compared to previous designs.

Perhaps I am missing something, but I thought the effect of diffraction was dependent only on the f-number and the degree of magnification. For a fixed format, with those fixed, it should not depend on the focal length. (It does also depend on the wavelength of the light, but usually one takes a standard middle range value for that.) See Jacobson's Optics Tutorial elsewhere at photo.net. Is there some subtle effect I don't know about?

Leonard, does not diffraction depend upon the physical size of the aperture? F22 on a 47mm lens is physically smaller than F22 on a longer lens. So it would seem that an aperture at F22 on, for example, a 300mm lens would be considerably larger, less susceptible to diffraction than F22 on a 47mm.

The f-number determines the effects of diffraction on the image, independent of the focal length. As an example, suppose both a 47 mm lens and a 470 mm lens are set to f22. The 470 mm lens will have a larger aperture diameter, which will cause the light rays to diffract by a smaller angle. In fact the angle will be 1/10 of the diffraction angle of the rays in the 47 mm lens. But the light rays in the 470 mm lens have to travel 10 times farther to reach the image, spreading out over this distance, so the linear size of the diffraction effects ends up being the same. And it's the linear size that tells us how much the image is smeared.

Diffraction in angular arc for a telescope depends on the diameter of the objective. Thus a diffraction limited 12" F8 reflector can split a double star better than a 6" F8 reflector, about twice as good.<BR><BR> For shooting photos at F8, both would have the same line/mm on film in theory, if diffraction limited, no vibration, perfect conditions with image dead nuts on axis, with perfect film.<BR><BR>Youn really want your optics to be diffraction limited, and not have abberations which maken the spot size larger. Its baffling why diffraction on photo.net is mentioned like it is a bad thing. Its like no taxes, perfect health, the ultimate goal of a good lens design. Imagine your neighbor complaining about having good health, complaining about low taxes, its bizzare.<BR><BR>A Diffraction limited lens at F11 resolves twice a diffraction limited lens at F22. Because of abberations usually there is not a 2x gain. <BR><BR>On an optical bench with artifical star one can take a Tessar or other lens and examine the spot size at the film plane, and vary the fstop. Wide open the spot size is large, then it gets smaller , then larger again as the lens is stopped down to F22. The minumum spot size usually is between F5.6 to F11 for many lenses. A BETTER lens will have a smaller spot size for a given focal length, and have its minmum spot size a faster fstop. Tsting line this is jsut a quick and dirty method of finding gross missalignment errors, ie lens tilt, focus shifts. What one gets actually on real film is radically worse than viewing an aerial image. In laser work one can be concerned with just one wavelength of light, and a simple one element lens can be diffraction limited on axis at a slow f number. The spot size can be measured by traversing a razor blade across the focus with a micron stage, and plotting energy/light versus razor blade distance moved. In optical recording one might be concerned in the diameter of the beam in microns at the 25, 50, 75 percent of energy levels.<BR><BR>In older ancient specs of LF lenses the "coverage" can be many times based on illumination, NOT resolution. If you are enlarging a 4x5 negative to 16x20", the enlargement might be 4.5 with a slight crop; 18 lines per mm on film might map to 4 on the print. 6 on the print would require 27 on film.

I've never been very concerned about diffraction with LF lenses and prints up to 16x20, i.e. only a 4x mag factor. I doubt that it's a major factor with even larger prints but 16x20 is the largest with which I have personal experience. Diffraction is a big deal with 35mm photography because of the relatively large mag factors required to make even an 8x10 print from a 35mm negatives, much less 11x14 or 16x20 prints. That's why, as I've understood it, 35mm lenses don't usually have apertures smaller than f16 or f22. In other words, they stop where we start.

 

I often photograph with lenses ranging from 80mm to 300mm and stop down to f45, sometimes f64, and haven't noticed any adverse effects from diffraction though I've never done a real test by making the same photographs at a series of different apertures. IMHO the effects of too little depth of field caused by using too wide an aperture (because of diffraction fears) will be much more noticeable than the minimal to non-existant effects of diffraction with LF lenses and prints, at least in the 16x20 range and maybe larger.

<i>But the light rays in the 470 mm lens have to travel 10 times farther to reach the image, spreading out over this distance, so the linear size of the diffraction effects ends up being the same. And it's the linear size that tells us how much the image is smeared.</i><p>

<b>How do telephotos relate to the aperture to film distance part of the problem?</b>

<p><i>"Its baffling why diffraction on photo.net is mentioned like it is a bad thing."</i> People are warning about stopping down very far when there is no need. One careless approach is just to always stop down to f64, in which case you will needlessly lose some sharpness to diffraction (and also increase your exposure times). It isn't a baffling piece of advice. I agree that a lens that is diffraction limited at a wider aperture (smaller f-number) is a better lens. I also agree that some make too much of this issue, and as Brian says, it is generally more important to get the depth of field that you want for your photograph.</p>

 

<p><i>"The minumum spot size usually is between F5.6 to F11 for many lenses."</i> Probably this is true of small format lenses, or on-axis, or with an aerial image. I have made test photos using modern, state of the art LF lenses and found their best performance, considering the ENTIRE image area, to be at larger f-numbers. You lose slightly on-axis stopping down, but get decided improvements off-axis. I've also found the image to soften on the film at the smallest stops, but this might not matter unless a large print is made.</p>

Michael, I agree with point two. Many of my LF lenses are best at F16 even, compared to the F5.6 to F11 numbers I mentioned. Its really a mixed bag, often the corners require stopping down abit more.. With my LF digital scan back I can cheat and do some trial scans, and tweak the focus a microgrunt, and get rid of the +/- 0.007 inch GG to film plane tolerance issue. In this case my results my tend to be a stop more wide open to be used than when a film holder is used. Some of this too migh be focus shift from focusing wide open to actual working f stop.

Michael? What about the distance of the aperture from the film? While not LF, you can look at the sharp Nikkor 300mm F4.5 lens which has the aperture blade assembly completely outside of the lens cluster, close to the lens mount.

 

And a telephoto lens necessarily has a shorter physical length than the long lens.

 

Does your principle still apply? In these cases?

 

Does it really apply at all? Does the Airy disk's first minima in fact degrade to insignificance because of the distance?

Concerning the diffraction issue, you might recall my article in the January/February 2004 issue of View Camera Magazine in which I set up a practical "experiment" to evaluate the effect of camera lens diffraction on actual prints. I have a link to it on my web site: http://www.paulwainwrightphotography.com/biblio_by_me.shtml -- scroll down a bit and you will see the link.

 

Although I have a PhD in physics and should know better (I was an experimentalist and not the sharpest student when it came to theory), I began my experiment under the assumption that shorter lenses should exhibit more diffraction than longer lenses at the same f/stop, but my results showed the effect from diffraction was independent of lens length. And I was deluged with e-mail pointing out my lack of knowledge of the theory of optics.

Longer focal length lenses have more lateral color/error. In visual astronomy you can get acceptable results with a simple achromat, your eyes at low light levels seem to not detect the color error compared when the lens is used for daytime images. Exotic glass to reduce color shifts is sometimes used in a high end photo lenses that are long focal lengths, and lessor in the shorter ones. A lens such a 50mm F2.8 Tessar might be diffraction limited at a faster Fstop than a 300mm F4.5 Tessar, some ofd this effect just due to color errors.

I would be interested in an analysis of diffraction off the center of the field based on physical optics theory. Similarly for the effect of tilting the image plane with respect to the aperture. Does anyone have any references to such treatments? Less technical treatments would present less of a challenge to the old gray cells, but in principle I have the background necessary to understand a complete treatment.

Leonard,<p> it's definitely too much for my grey cells but maybe you'll understand it: "Chapter 11" on Fourier optics in:

Eugene Hecht "Optics" 4th edition, Addison Wesley Longman. Those more than fifty pages are dealing exhaustively with theory, analyis and practical consequences of light wave transformation caused by passing an aperture. <p> Uli