PDA

View Full Version : Resolution of different focal lenghts of same lens



Richard Årlin
18-May-2006, 15:11
Take any lens that comes in different focal lenght for different formats, eg 4x5, 5x7, 8x10, would they have the same resolution regardless ? I am thinking of "compressing" the number of lenses you carry or own. A typical 8x10 wideangle, say a 158 mm Cooke wide-angle anastigmat series VIIb would it be as sharp as a 82 mm on 4x5 and could it be used as a normal standard lens replacing for example a 150 Apo Symar made especially for 4x5 without loosing quality. It would be nice to have a minimized series of lenses usable for both formats like (82, 108 ) 158, 240, 360 and maybe a 450 fuji c that could be bought selling some then unnescessary lenses.

Cheers, Richard

Brian Vuillemenot
18-May-2006, 16:26
You can use any lens with a big enough image circle to cover a given format for that format, as well as all smaller formats (any lens capable of covering 8X10 will cover 4X5 and 5X7). Resolution is a function of the given lens, and not the image circle. Any modern multi coated lens from the big four will have a resolution not easily destinguishable from any other modern lens under almost all real world working situations. If a lens for 8X10 has a low resolution on that format, so will it on 4X5 and 5X7, although the loss of resolution may be amplified somewhat by the smaller film size and greater enlarement factor required.

The main drawback with using 8X10 lenses for 4X5 is that they are physically much larger, many in Copal #3 shutters, and the rear element may be to big to mount on your 4X5 camera. 4X5 cameras in general are not as beefy as 8X10s, so the front standard may not be able to be kept firmly in place. A huge Copal #3 lens on a lightweight 4X5 camera will also shift the center of gravity forward, and make it suseptible to tipping over. In addition, many lenses for 4X5 will not quite cover 8X10, for example the 210 mm Apo Sironar S. I've used a large 8X10 lens for both 4X5 and 8X10 (300 mm Apo Sironar S) for quite a while, and it's worked fine, although you may be better off going with something slightly smaller (300 mm Nikkor M and Fujinon C are both in copal #1 shutters and will cover 8X10 with minimal movements).

Dan Fromm
18-May-2006, 16:28
Um, Richard, if you want, say to use a 158 Cooke Ser VIIb as a 4x5 normal lens instead of a 150 Apo Symmar, the thing to do is shoot the two lenses as you would normally -- same apertures, ones that you usually use -- and then ask whether both sets of negatives will give you "good enough" prints at the size you normally make. And if you want to replace an 80 shot on 4x5 with the 158 Cooke on 8x10, do the same. Your standards are the ones you have to meet, and none of us can know to what standards you work.

paulr
18-May-2006, 17:00
If you don't have the opportunity to test, it's worth noting that lenses for larger formats tend to perform worse than lenses for smaller formats, all else being equal.

Within the same line of lenses, you'll see performance gradually decline as the focal lengths (and image circles) increase.

Brian Ellis
18-May-2006, 19:46
"Within the same line of lenses, you'll see performance gradually decline as the focal lengths (and image circles) increase."

That's interesting, I don't think I've ever heard that performance declines with longer focal lengths in the same line of lenses. I'm not arguing, just curious - what's your basis for that statement?

paulr
18-May-2006, 21:21
Take a look at the published MTF curves from Schneider or Rodenstock. Schneider's are at schneideroptics.com; I believe Paul Buzzi has hosted Rodenstock's on his site. Look at the curves for any of their normal lens lines (like the apo symmars that Richard mentions). It's a steady downward progression with increasing focal length.

In that last post I suggested this would be true with all lens lines, but I just browsed through and it looks like it doesn't hold true (at least not as consistently) with wide angle lens designs.

Scott Rosenberg
18-May-2006, 21:27
isn't it pretty well established that resolution deteriorates as one moves away from the center of the lens? that said, would one therefore want to use the lens with the largest IC - maximizing the amount of image formed by the 'sweet spot'? assuming a good compendium is also employed to reduce as much non-image forming light from entering the bellows as possible. of course assuming the resolution of all lenses comprable with equivalent rates of deterioration as one moves from the center, which we also know not to be the case.

paulr
18-May-2006, 21:44
isn't it pretty well established that resolution deteriorates as one moves away from the center of the lens? that said, would one therefore want to use the lens with the largest IC - maximizing the amount of image formed by the 'sweet spot'?

There is some truth to this. The trouble is that in general, when you move towards a wider image circle and more even performance from corner to corner, the performance near the axis declines. At the extreme you end up with perfectly even but poor performance from corner to corner.

Oren Grad
18-May-2006, 22:15
>> I don't think I've ever heard that performance declines with longer focal lengths in the same line of lenses. I'm not arguing, just curious - what's your basis for that statement? <<

I've just checked this for two modern lens series. MTFs provided by Rodenstock for the Apo-Sironar-S line show performance in the center of the field at f/22 decreasing steadily but very modestly as the focal lengths increase. Remarkably, Schneider MTFs for the Apo-Symmar L line show essentially no deterioration in the center of the field at f/22 beyond the 180 focal length, though there is some deterioration at larger apertures as focal lengths increase.

Within a given focal length, lenses with bigger image circles do not necessarily deliver the same performance in the exact center of the field as those with smaller image circles. Whether it gets better or worse depends on the specific lenses, but in general among modern lenses the differences at the center of the field are not large at typical working apertures, and what happens as you depart from the center is much more important in selecting the optimal lens for a given application.

>> of course assuming the resolution of all lenses comprable with equivalent rates of deterioration as one moves from the center, which we also know not to be the case. <<

That's correct, it's not true in general.

Brian Ellis
18-May-2006, 23:04
So is it fair to summarize by saying that within the same line of lenses resolution deteriorates as focal length increases - sort of?

Emmanuel BIGLER
19-May-2006, 02:03
So is it fair to summarize by saying that within the same line of lenses resolution deteriorates as focal length increases - sort of?



Yes, I think so, with some caveats.

If there were no diffraction and only geometrical aberrations, the residual spot size for a given source point (and consequently the image quality) would scale exactly like the focal length for the same optical lens design, since all ray parameters that define a geometrical aberration-limited spot scale exactly like the focal length if all other parameters, curvatures,spacings and aperture diameters of lens elements are scaled accordingly.

For example in a classical textbook like "Modern Lens Design" by Warren Smith, many examples are shown with a conventional focal length of 100 mm. It is only when you want to compute the mixed diffraction/aberration MTF that you have to take into account the absolute values of curvatures, spacings, and apertures, other simulated data on gaussian lens parameters and geometrical aberrations can be scaled easily from the focal length value.

To this naive analysis we can object

- that diffraction does not scale with focal length, for a given relative aperture (f-number) and a given wavelength range, the diffraction spot size would be the same regardless of the focal length. To me this is detrimental to very short focal lengths used, say, for amateur digital cameras with a sub-centimeter detector size. However the fact that the larger the format and focal length, the smaller the aperture can be stopped down, is directly related to the balance between residual aberrations, scaling like the focal length and format, and diffraction, which depends only on the f-number. For example a compilation of recommended best apertures N_best for standard lenses used in 35MM, medium and large format suggested that the best f-number follows the "trend" law : N_best = f(in millimetres)/8
This behaviour, IMHO, can only be explained by the balance between geometrical aberrations and diffraction.

- manufacturers have no obligation to scale exactly their designs sold under a similar lens name for covering different formats...

Richard Årlin
19-May-2006, 02:08
So is it fair to summarize by saying that within the same line of lenses resolution deteriorates as focal length increases - sort of?

That was kind of my general thought. Using smaller format like 4x5 you nead a better performance from your lenses since the negative is normally enlarged, whereas a larger negative that is often contact printed would not require the same perfomance. It would seem logical to calculate a series of lenses of different focal lenghts to be optimized for the shortest in the series.

Richard Årlin
19-May-2006, 02:17
Um, Richard, if you want, say to use a 158 Cooke Ser VIIb as a 4x5 normal lens instead of a 150 Apo Symmar, the thing to do is shoot the two lenses as you would normally -- same apertures, ones that you usually use -- and then ask whether both sets of negatives will give you "good enough" prints at the size you normally make. And if you want to replace an 80 shot on 4x5 with the 158 Cooke on 8x10, do the same. Your standards are the ones you have to meet, and none of us can know to what standards you work.

Certainly you are right. Only the query was a rather academical one and it seems to have inspired some information to surface that was unfamiliar to me.

Oren Grad
19-May-2006, 07:55
So is it fair to summarize by saying that within the same line of lenses resolution deteriorates as focal length increases - sort of?

I think that's fair, with the caveat that the performance fall-off may show up in different parts of the performance "envelope" depending on the manufacturer's design preferences.

To add a couple of additional data points to this discussion, I've just looked at data for the Apo-Sironar-N and Apo-Symmar series as well. The pattern is comparable to what I reported above for the Apo-Sironar-S and Apo-Symmar L series.

Note that the tradeoffs for wide angle designs may be different from those for plasmats intended as near-normal focal lengths. I'll try to take a peek at some of those a bit later.

paulr
19-May-2006, 09:03
I
Note that the tradeoffs for wide angle designs may be different from those for plasmats intended as near-normal focal lengths. I'll try to take a peek at some of those a bit later.

Please do, I'm curious about this also. I had mostly looked closely at this with plasmats. I was surprised last night when I glanced at the charts for the super symmar xl line and super angulons ... performance looked pretty consistent from model to model, based on the few I glanced at.

John Berry
19-May-2006, 13:08
The circle of confusion can be larger as the format goes up. 8x10 lens on 8x10 camera will give comparable results as 4x5 for the magnification. I too would like to have 35mm specs in my large format lenses, but I would be inclined to think it might make you run for one of the new cooks as it would be so much cheaper. As for the drop off in preformance, when I put an 8x10 neg in the enlarger that was shot with my 24" red dot, it looks like I'll get by.

JW Dewdney
28-May-2006, 22:01
The circle of confusion can be larger as the format goes up.

Well, if you want to get really PRECISE about it. The circle of confusion AND diffraction increase linearly with effective focal length. That's why shorter F.L. lenses seem to measure better. The main reason anyway.

Arne Croell
29-May-2006, 02:46
Well, if you want to get really PRECISE about it. The circle of confusion AND diffraction increase linearly with effective focal length. That's why shorter F.L. lenses seem to measure better. The main reason anyway.
For diffraction that is only true if you assume a fixed diameter of the aperture. Usually the f-stop is kept constant in a lens line for most of the focal length, i.e. the aperture diameter increases proportionally with focal length, and diffraction will then be constant for a constant f-stop at infinity. Only for the longest lenses, due to the limitations of the shutters, does the f-stop go down occasionally.

paulr
29-May-2006, 09:10
This behaviour, IMHO, can only be explained by the balance between geometrical aberrations and diffraction.

interesting. i think i get the general idea of what you you're saying.

could this be related to the reasons the normal lenses seem to lose more quality with focal length than the wide angle designs?

Emmanuel BIGLER
29-May-2006, 10:17
interesting. i think i get the general idea of what you you're saying.
could this be related to the reasons the normal lenses seem to lose more quality with focal length than the wide angle designs?

Well, I do not know, since the design rules for wide-angle lenses seem so special to me... for a wide-ange lens probably the designers trade the absolute performance at the center against the overall homogeneity across the wide field. So a direct comparison between standard view camera lenses (6 elements, 70-75°) and wide-angle lenses (6 to 8 elements, up to 120 degrees !!!) is something I am not comfortable to discuss about...

The only thing for which I'm confident, simply from reading the brochures, is that modern ultra-wide angle view camera lenses of short focal lengths (e.g. Schneider lenses shorter than 58MM, or the Rodenstock 35-45-55 apo-grandagon series) should not be stopped down to f/22 in order to avoid too much of diffraction. For example the recommended f-number for the apo-grandagon series is 8-11 as explained with detail in the manufacturer's specs.
So the trade-off between aberrations and diffraction exists as well for wide angle lenses.
Since there is no apo-grandagon in focal length 150mm we'll never know... however for the grandagon-N series, from the 65 mm to the (discontinued) 155mm, the recommended f-stop shifts from 16-22 to 22-32, so the same ideas apply, namely that you can stop the aperture more on larger focal lengths provided that your final magnification factor will be smaller !

Oren Grad
29-May-2006, 11:13
So the trade-off between aberrations and diffraction exists as well for wide angle lenses.

Interestingly, the Apo-Sironar-Digital series are specified as being optimized for f/8-11, including the longer focal lengths through 180. Schneider also specifies image circles for the Digitars at f/11 rather than the familiar f/22.

Richard Årlin
29-May-2006, 11:31
Interestingly, the Apo-Sironar-Digital series are specified as being optimized for f/8-11, including the longer focal lengths through 180. Schneider also specifies image circles for the Digitars at f/11 rather than the familiar f/22.

Aren't digital backs smaller than sheet film, thus requiering shorter lenses and consequently deeper DOF ?

Gordon Moat
29-May-2006, 12:23
Aren't digital backs smaller than sheet film, thus requiering shorter lenses and consequently deeper DOF ?

The largest are near 36mm by 48mm, or slightly under 6x4.5 sizes. There also seems to be an issue of using too small an aperture causing problems with the resulting image files.

Ciao!

Gordon Moat

paulr
1-Jun-2006, 11:01
Interestingly, the Apo-Sironar-Digital series are specified as being optimized for f/8-11, including the longer focal lengths through 180. Schneider also specifies image circles for the Digitars at f/11 rather than the familiar f/22.

not too surprising ... as optical aberrations get corrected better, the optimum aperture shifts toward wider open. because diffraction is going to be a constant at a particular focal length. the optimum aperture is the point at which the lens the combined effects of diffraction and aberrations are lowest.