Best F/11 normal lens

[Robert_M]

Advertisement (gone after registration)

Well you don't gain anything, but equally you don't lose much if you need the DOF. The old formula for stopping down to f/64 with a large format lens works because the hole even at f/64 is much bigger than the equivalent hole in a 35mm lens set for f/22. So diffraction (the bouncing of light off the edges of the shutter) is less (almost non-existent) with a large format lens than with a 35mm lens. But far too much is made of diffraction when the alternative (if the flawed logic is taken to its conclusion) is not to make an exposure at all. You either have the courage to do it at whatever aperture is appropriate, or your conscience listens to perfectionists who never make an exposure worth looking at and you just go home with your tail between your legs.

 

Steve

 

 

Steve, with all due respect, this statement about diffraction is wrong. As a physicist, astronomer and optical engineer I feel compelled to make a comment (sorry).

 

You are correct that the diffraction for an object at infinity is (inversely) proportional to the aperture. In fact, the diffraction (resolution) of the object at infinity goes as 1.2 Lambda/D where lambda is the wavelength and D the aperture. BUT, the lens in the system forms the image at the focal plane with a plate scale of 1/F (in radians per units of F) where F is the focal length of the lens. The net result of these two facts is that the Airy disk size (diffraction spot size) at the focal plane (sensor or film plane) is 1.2 lambda F/D which is 1.2 x lambda x f/ratio. The diffraction function is close to a gaussian with this value as the 1/2 power point - actually it is a (sin x/x)**2 function for those who are care.

 

The point is, it doesn't matter what the focal length or film standard is. The diffraction is the same physical size (in microns) and it is directly proportional to the F/ ratio. The M9 sensor is 6.8 microns if I remember right and this is typical for visible wavelength sensor systems. Fine grain film has a similar resolution (I think). F/8 will give one a diffraction smoothing function of 5.8 microns at the half-power point. This matches the sensor size very well. It is no mystery that about f/8 on all lens tests show the best resolution before diffraction starts to set in.

 

At f/11 the image will begin to show the effects of diffraction. The image is effectively a convolution of gaussian smoothing function with the image to reduce resolution even if the lens is perfect. In practice, this effect will not be too noticeable at f/11. Distortion will be more noticeable.

 

For large format photography, it always depends on the fact that the negative is not enlarged nearly as much. With a f/64 image (and even finest grain film) the image is heavily oversampled by the film grain vrs the diffraction resolution set by the f/64 stop. If one enlarged the image to the same extent as a 35mm frame, one would see the diffraction resolution severely limiting the image resolution. The image would, again, be much sharper at f/8.

 

I also do not understand why landscape photographers want to stop down to f/64. This may have been useful when film was coarse grained. But with modern film and digital sensors, about f/56. or f/8 will give one the best resolution. Of course, DOF is another story. That is a reason to stop down.

 

My 2 cents worth....

 

RM

[250swb]

 

I also do not understand why landscape photographers want to stop down to f/64. This may have been useful when film was coarse grained. But with modern film and digital sensors, about f/56. or f/8 will give one the best resolution. Of course, DOF is another story. That is a reason to stop down.

 

My 2 cents worth....

 

RM

 

Well landscape photographers want to stop down to f/64 because they want the DOF. It is exemplified by the F/64 Group of landscape photographers.

 

Now, it is all well and good spouting the theory behind diffraction, but the bottom line is that if the aperture blades are the same thickness and made from the same material in both the 35mm lens and the large format 8x10 lens, and the aperture opening is vastly larger in the large format lens, diffraction will not be noticable in a print under nearly all circumstances. But in any case I don't understand the point you are trying to make since I did acknowledge f/8 is better than f/11? I object to theoreticians scaring people into not using the full range of f/stops available presumably because its the only knowledge they have to impart about photography. For instance wouldn't it be refreshing if questions about apertures were answered with views about composition than textbook science?

 

Steve

[pgk]

I object to theoreticians scaring people into not using the full range of f/stops available presumably because its the only knowledge they have to impart about photography.

The really good bit about digital photography is that its incredibly easy to go out and test something for yourself to see whether it meets your own requirements. I have extensively tested some specific lenses at different apertures for actual high resolution requirements and my findings have been that although f/11 should be theoretically better than f/8 it is difficult to see real world differences on real world subject matter, however the benefit of increased depth of field was noticeable. So as far as I am concerned I am perfectly happy to use either aperture depending on the depth of field that I require. f/16 on the other hand can show some differences in fine detail compared to f/11, but again its a trade off between depth of field and fine detail so I use f/16 when I have to and when the subject matter allows. These are my findings for a specific requirement.

 

What they illustrate is broadly in line with theory but that theory has to be tempered with the subject matter's requirements. For landscapes I usually shoot at f/11 when I want to maximise depth of field, preferring to retain as much fine detail as possible, and this is also the aperture I usually choose for buildings too. When in doubt just go and try things - it has no cost other than time and effort and you can decide for yourself what works best for you. Theory can be used as a guide but there are many factors which influence a final image so you need to be aware that theoretically based suggestions are inevitably somewhat simplistic as they only take the specific influence upon which they are based into account and ignore the (inevitable) others.

[01af]

Some pictures need the depth-of-field to be narrow, others need it wide. If you're shooting the latter kind (which usually includes landscapes and architecture) then by all means stop down to f/11 or f/16 (or maybe even to f/22 if the lens permits), to get the DOF you want, and don't bother worrying about diffraction. If your image's composition depends on wide DOF then it simply will look blurred in most parts when DOF is too narrow. The (slightly!) improved peak sharpness in the narrow part that's in focus won't help in this case.

 

Photographers should be aware of the issue of diffraction, but they also should properly appraise the position and significance thereof in the pan-galactic context of all things (i. e. don't overestimate it).

 

EDIT: What Paul said.

[giordano]

Here is a link again to two very recent images that were shot "blind" and at f16. You can download them at full resolution for sharpness and also see where I placed focus and how it falls away. (And the numerous quite tiny dark spots from sensor dust that don't show up at f8.) One is with a 35 year old Nikkor 35mm PC and the other is with the current 24 TSE. I don't see why the Leica lenses should not do as well or better when at f11.

 

Gallery Password Required | Alan Goldstein

Password: Alan

 

Alan, I hadn't looked at your pix because I haven't drunk the Kool-Aid and still know that nine times out of ten diffraction is less important than getting the right depth of field - and because I've had my Nikkor 35/2.8 PC for about thirty years and know what to expect. When I did download them I was really impressed with the results from the Canon 24 TSE.

 

So many good lenses, so little money...

[pico]

[...]The diffraction is the same physical size (in microns) and it is directly proportional to the F/ ratio.[...]

 

F-Ratio? Is diffraction (the aperture) not more related to the absolute (physical) size of the aperture? I think 6mm in diameter and smaller is where we begin to see diffraction. 6mm on a large format lens, for example a 12" lens, is a much greater (numeric) F-Ratio than the same F-stop on a 50mm lens for 35mm. I could measure it exactly if you wish. I have a 14.75" lens in storage. I'm guessing that 6mm on a 14.75" lens is about F-64.

[pico]

Advertisement (gone after registration)

Well landscape photographers want to stop down to f/64 because they want the DOF. It is exemplified by the F/64 Group of landscape photographers.

 

True. F/64 was also chosen to separate the movement's aesthetic from the painterly-like, dreamy soft focus images of the romanticist movement in photography.

[Robert_M]

Well, apparently I did not get my point across. The diffraction effects at the focal plane are dependent on the f/ratio NOT the physical size of the aperture. I tried to explain this with a few simple formulae. I was not trying to snow the readers or confuse them. One can find a decent explanation in the following wikipedia article which includes an example for cameras. It explains the same bottom line: The f/ratio determines the diffraction size at the focal plane.

 

Airy disk - Wikipedia, the free encyclopedia

 

One can also find further information by doing wikipedia searches for "point spread function" or "diffraction" if interested. BTW, this is not some "theory" but real, factual constraints of physics and optics.

 

The aperture (and wavelength) will indeed determine the resolution (diffraction limit) for the lens (in angle). And, the scale of the angle per mm in the focal plane is determined by the lens focal length. Combining these two facts will explain why the resolution (diffraction spot size, point spread function, Airy disc size - take you choice of terms) is dependent on the f/ratio (not the physical aperture size).

 

In my last post, I was not addressing use of DOF or any photographic composition issue. I was only trying to correct the mis-statements being made about the physical aperture determining the diffraction size. There is no reason to attack me for trying to explain the facts.

 

RM

[Guest #12]

Well, apparently I did not get my point across. The diffraction effects at the focal plane are dependent on the f/ratio NOT the physical size of the aperture...

 

...no you're right, "the hole is bigger" did not explain things across formats. But when you're explaining how the diffraction term balances the aberration term at the optimum aperture...it seems natural to say that for a particular lens, the diffraction effect depends on the physical aperture.

 

Thanks for the info...

[NZDavid]

I am a big fan of the Elmarit 24-ASPH. It is usefully wide without giving you huge foregrounds, and managing perspective distortions is easier than superwides. Otherwise, I'd go for the 28. Yes, you can correct distortion later (I use LR3) but it is nice to get the picture as right as possible beforehand. A tripod and spirit level will help. With wider lenses of course there is plenty of DOF at a lot wider than f/11.

 

David

http://www.davidkillick.co.nz

[Shade]

At F/11, overall edge to edge sharpness, sharpness in the centre, and all of the frame?

 

The 35/2.8 Summaron-M.

[SJP]

I never realized my lenses could go beyond f/8, interesting.