Use of smaller apertures than f/11

<p>I was wondering if I am missing something about small apertures like f/22 and f/32. From what I've seen, most lenses reach their best resolution performance before f/11, so, other than very special situations where strong light might require stopping down that much, what use are these for? Is there any relatively common scenario where these would be useful, or is this just a remnant from an era when highest shutter speed was 1/1000?</p>

<p>Stopping down increases depth of field. The peak sharpness at the focal plane may diminish through diffraction effects, but the greater depth of field may mean that other areas of the picture become sharper.</p>

<p>Hmm haven't stopped down beyond F8 on my 16-85 - 16mm/F8 at hyperfocus gives tremendous dof. For my longer lenses, can't stop down much anyways - need the shutter speed.<br>

Alvin</p>

<p> Why don't you just go outside and <em>try </em> those apertures?</p>

<p>If you only use the images in a relatively small size (such as small illustrative images on magazine pages, books, or small web sized images) you can stop down to quite small apertures and obtain wider depth of field (i.e. f/11 on DX, f/16 on FX; f/22 might be ok in some contexts). The key is what is the display size and how badly you need the additional depth of field.</p>

<p>Today there is an alternative to using a small aperture: focus stacking. Most lenses in use were made for film and before focus stacking become a common technique. Therefore it might have been that for illustrative use the use of a small aperture was necessary whereas the best sharpness of the image had to be sacrificed. Today you have different options at least for static subjects, thanks to these new algorithms.</p>

<p>If you want to play with long exposures for creative use of motion blur, you may have to stop down to the smallest available apertures or use a very dense neutral density filter which may prevent the effective use of the viewfinder and autofocus. I have used f/22 on occasions for this purpose and still the exposure times were a bit too short (I shot on Ektar 100 at EI 50 since my digital cameras only go to ISO 200 as base ISO); but I didn't have a lens which allowed smaller apertures to be used at hand so I didn't quite get the results I wanted.</p>

<p>Diffraction, which is the cause of quality loss at small apertures, depends on the physical size of the aperture, not the f-number. A long lens has a physically bigger aperture at f22 than a short lens does. What this means is that long lenses like 200-300mm are usable at f22 or f32, but a 20mm lens would produce a significantly diffracted image at f22. Also the visible effect of diffraction is seen more easily on the magnified image, and if you are using that long lens on a large format camera, you won't be enlarging the image so much and all varieties of optical degradation are harder to see.<br>

As Luis says, give it a try and see what results you get. You might need a tripod to get a steady shot with very small apertures.</p>

<p>To go along with Brian, long lenses have a shallower depth of field, necessitating small apertures for the inclusion of a broad depth of subjects, whereas very wide lenses (like a 20mm) have deep fields of focus, whereby they do not require as small an aperture to achieve more depth-of-field. JR</p>

What Jeremy said. With a 300mm lens on an 8x10 camera, I get about the same depth of field at f/64 as I get at f/16 with a 50mm lens on a 35mm camera.

<p><em> A long lens has a physically bigger aperture at f22 than a short lens does.</em></p>

<p>This part is correct.</p>

<p><em>What this means is that long lenses like 200-300mm are usable at f22 or f32, but a 20mm lens would produce a significantly diffracted image at f22</em></p>

<p>It doesn't work like that. Typical high quality long lenses peak at around f/4-f/5.6 and the image quality degrades as you stop down more. Yes, due to diffraction.</p>

 

<p>Most lenses achieve their peak performance 2 or 3 stops below their maximum opening. Contrary to popular believe, the resolution of a lens does not automatically improve at smaller apertures. Diffraction aside, various optical design compromises tend to increase (aberations, coma, etc) tend to increase at smaller apertures. In most lenses, diffraction does not predominate until the aperture is smaller than f/11 or f/16.</p>

<p>Diffraction (the Airey Disc diameter) in a lens is proportional to the relative aperture (diameter/focal length) and the wavelength of light (q.v., <a href="http://en.wikipedia.org/wiki/Diffraction#Diffraction_by_a_circular_aperture">http://en.wikipedia.org/wiki/Diffraction#Diffraction_by_a_circular_aperture</a>) The focal length does not matter, in that its effect is incorporated into the diffraction equation - longer lenses have larger apertures but greater magnification. Consequently, few lenses for 35mm stop down smaller than f/22, and then only macro lenses in order to achieve a greater DOF.</p>

<p>Large format lenses usually go to f/32 or even smaller (q.v. the "f/64 club"). I have one lens which stops down to f/128, mostly for greater DOF (which is much less for LF film). This works because large format film requires less enlargement for printing, so the same uncertainty has proportionately less effect.</p>

<p >You can image even using a pin-hole substituted for the lens. A pin-hole images because it forms tiny image circles of light that represent points on the subject. It is the diameter of these imaging circles that we are concerned about. They will appear as a point without desirable dimension if viewed from a distance that is about 1000 times their diameter.</p>

<p > </p>

<p >Thus we can only obtain a sharp image if we are using a tiny pin-hole. The problem is the tiny hole size yields a dim image. If we enlarge the pin-hole the image gets brighter. Sorry to report that as we enlarge the pin-hole the image circles enlarge and overlap and the acuity severely degrades. </p>

<p > </p>

<p >We can substitute a lens for the pin-hole and the situation is greatly improved. The lens acts to converge the light rays as they pass through the aperture and this action makes possible the use of a much larger aperture. Nevertheless the lens continues to suffer a loss of acuity at tiny apertures.</p>

<p > </p>

<p >This is because when light passes through a small aperture it does not strictly obey the laws of geometrical optics. Some of the light will be scattered (diffracted). This scatter light causes the image circles to be poorly defined. What is impairing the image is diffraction and interference. </p>

<p > </p>

<p >Seems, light rays streaming through a tiny aperture are not perfectly cut off. Perfect would mean light on one side and dark on the other. What actually happens is the edge is not clearly defined. In fact the shadow of aperture blades will takes the form of bands of light that decrease in spacing with distance from the blades. Moreover the bands reinforce and cancel each other, the result, tiny image circles that are scalloped at the edges surrounded by rings of light. The first thee rings make up 84% of total. Now we call the tiny image circles with rings Airy Disc. The result of the Airy Disc is degraded resolving power at the smaller apertures.</p>

<p > </p>

<p >Generally the best lens performance is 2 or 2 ½ stops stopped down. </p>

<p>". . . what use are these for?. . . "<br>

Less light; longer depth of field. We could just as easily say, what good is f/2.8? </p>

<p>Laurentiu, you should really experiment and may be surprised by the results. Some 35mm lenses can be quite good at F22 while others can be very bad. Interestingly the price of the lens is not always and indication of how it will perform. In general zooms perform worse than primes and some very good zooms (e.g. my old 80 -200 F4L Canon FD zoom which outperforms the FD 200 F2.8 prime at most apertures) can be terrible at small apertures.<br>

As to the use - well as others have said very large DOF and also the ability to take long exposures. I usually only carry one ND filter with me (3 stops). I have enclosed a recent example of where I had to use f22 as I only had a 3 stop ND and needed to get an exposure of around 1 minute. As you may be able to see (although the image is heavily compressed) the canon 16-35 f2.8 II is actually reasonably sharp at F22</p><div></div>

<p>"what use are these for? Is there any relatively common scenario where these would be useful"</p>

<p>Landscape photography. Macros. Anywhere you need a large dof. There are times bokeh is an inappropriate distraction. (Yes, I realize you can use focus stacking but some of us try to get the photos right in camera.)</p>

<p>Macros always have such shallow DOF ... question, since diffraction happens at f22, what does an image taken with something like the Sigma 105mm look like? I think that thing can go to f45? Thanks.</p>

<p>Laurentiu, you don't say what camera you use but with my Hasselblad Zeiss lenses I shoot f/32 more frequently than any other f/stop in my landscape photography. The depth of field drops off noticeably with every stop the lens is opened up except when shooting distant subjects. What I gain from the depth of field more than offsets the need for using a tripod.</p>

<p><em><strong>Macros always have such shallow DOF ... question, since diffraction happens at f22, what does an image taken with something like the Sigma 105mm look like? I think that thing can go to f45? Thanks. - Dale</strong></em><br /><em><strong></strong></em><br />Diffraction happens at at apertures much wider than f22 -- it's just not noticeable to the naked eye at wider apertures.</p>

<p>It all boils down to what your consider acceptable IQ, and a few other things. Will you be printing greater than 8x10 or is it just for web only? The larger your print, the more noticeable diffraction will be. Are you shooting with a full frame sensor (or 35mm film) or a DX-sensor DSLR?</p>

<p>I tend to shoot macros at f16, and sometimes at f22, and use flash. I've never shot at f45, but I would think that such images may only be acceptable on smaller prints or web viewing (and probably only marginally). Of course, there's always Photoshop :-)</p>

 

<p>Thanks to all who responded and my apologies for not providing enough context around my experience and the camera I use. I currently shoot APS-C and I usually shoot in the wide to short-tele range. Even in the rare case where I shoot above 75mm (35mm equiv), I don't do it in situations where I need larger DOF than what I can get at f/11.<br>

I've already shot at f/22, sometimes out of curiosity, sometimes because I wanted the longer exposure time for creating blurred images - (I have a couple of photos in my portfolio). I can see how for macro of static subjects, larger DOF is useful if resolution doesn't degrade much. I can also imagine scenarios where a large DOF would be needed for landscape - but how often do you shoot with a lens where that DOF can only be attained at f/22 (on APS-C)? I should indeed experiment more with my lenses to get an idea of their resolution performance at small apertures - it's just that I have little time available for such tests and they don't overlap with the settings I use when I shoot. I did do a test some years ago using a kit zoom (Pentax 18-55) and I've been disappointed by the resolution I was getting at apertures smaller than f/9 - DOF was more than I needed for my landscape shots but resolution was going down the drain, if I stopped down further than f/9.<br>

Here are some additional questions that I had while looking through all the answers that you generously posted.<br>

- Why are lens reviews (such as those at photozone usually) omitting small apertures in the resolution tests? How can I find out whether a lens will perform well at small apertures (other than by buying it and trying it out)?<br>

- Why do wide-angle lenses that provide great DOF still provide such small apertures like f/22? For example, is there any scenario where I would want to use f/22 on an 8mm lens (that, on top of that, is designed for APS-C)?<br>

- I get the point about larger format cameras benefiting more from these apertures, but I see these apertures available even on lenses that are exclusive to APS-C. Is this partially due to marketing, because people might expect to see that f/22 available even if the resolution is useless?</p>

 

<p><em>"Diffraction, which is the cause of quality loss at small apertures, depends on the physical size of the aperture, not the f-number."</em></p>

<p>This is absolutely incorrect. Diffraction depends on focal ratio, not absolute aperture. This error has been posted here about fifty times over the past thirteen years that I have noticed. It was wrong the first time and it is wrong this time.</p>

<p>The only sure way to increase depth of field is through lens tilt. Diffraction is an important factor beyond f8 in most lenses and that will become even more important (the f number will decrease to wider apertures) as pixel numbers increase in future cameras (see recent articles on diffraction in Photo Techniques of this year) </p>

<p>depends on the individual lens, and whether you're using film or digital</p>

<p>Ok, I'll bite on one:</p>

<p><em>Why do wide-angle lenses that provide great DOF still provide such small apertures like f/22? For example, is there any scenario where I would want to use f/22 on an 8mm lens (that, on top of that, is designed for APS-C)?</em></p>

<p>The closer you get to the object you're shooting in the foreground, the more dof you need. Example: I have a 24mm that has great dof, but it also has "macro" capability. What that really means is I can put the lens close to a relatively small object - like a mushroom. But I still want the mushroom's habitat to be reasonably sharp because it's part of the story. (If I wanted the mushroom by itself, I'd probably use my true macro, but it would give a completely different view of the subject.) Now if I'm shooting mountains using the same lens and I'm not <em>that</em> close to a foreground object, I can use the same lens to get everything in focus without resorting to f/22. But there are honestly times when only f22 or above will do, because by being so close, it's the only way to gain enough hyperfocal distance to capture the entire scene.</p>

<p>So the lens makers include f22 for people like me. ;) As for diffraction - I honestly haven't had many problems with diffraction.</p>

<p>I've looked a bit at the lens reviews on dpreview - unlike photozone, they do show resolution results for the entire aperture range. I've checked half a dozen prime lenses and a few zooms and all of them are what is usually qualified as "soft" at f/22. And the resolution becomes soft all across the image - from center to borders, whereas up to about f/11 it's much better in the center, if not all across. Sometimes, it's interesting to note that the f/22 flat resolution can be above the border resolution at f1.4, but the center one is a long shot from what can be obtained up to f/11.<br>

I guess I rarely needed that extended DOF so badly as to be willing to incur the softness penalty and risk motion blur on top of that.<br>

Deb: you answered my question about a particular lens by talking about a very different lens. Have a look at this resolution chart for the specific lens I had in mind: <a href="http://lenstip.com/160.4-Lens_review-Samyang_8_mm_f_3.5_Aspherical_IF_MC_Fish-eye_Image_resolution.html">http://lenstip.com/160.4-Lens_review-Samyang_8_mm_f_3.5_Aspherical_IF_MC_Fish-eye_Image_resolution.html</a>. At f/22 this lens is softer than at f/3.5. Even at f/8, which is the sharpest aperture all around, I am not amazed by the resolution of this lens - it's good, but not outstanding. For me, it makes no sense to stop down to f/22 with this lens to get a mushroom in focus - besides, it would have to be 1 foot tall to be noticeable because I can't focus the lens under 1 foot and at that distance a mushroom would look like a speck. I would say this is the case of adding a checkbox on the technical sheet of the lens, don't you agree?</p>

<p>Model railroading is another hobby of mine, and I've used photography to share my work with others for many years. Way back when I was using film cameras, to get exteme closeups of my work I used an old manual focus 28mm lens in which I'd added a pinhole aperture. I could take photos with items almost touching the front of the lens. I knew from taking photos with other lenses that the pinhole shots did not have as sharp focus. But viewed on their own, the pinhole shots were acceptable At least seven of my pinhole photos were used as cover shots on model railroad magazines.</p>

<p>Switching to a DSLR some years ago, I found the cropped image from my pinhole equipped lens did not provide the view or sharpness as it was on film cameras. So I've been using my Canon 28mm wide angle at f/22 to get acceptable results. The acquistion of a Tamron 11-18mm lens made things better. Again using it at f/22 I can get results that I like.</p>

<p>Here's a reduced in size and quality shot of the interior of a scale model engine house that I made from scratch, using the Tamron 11-18mm lens at 11mm and f/22. The structure is 10 inches deep:<br /><br />So using f/22 is sort of a "way of life" for some aspects of my hobby.<br>

<br /><img src="http://img.photobucket.com/albums/v409/railfan/Enginehouseinterior-1.jpg" alt="" /></p>

<p> </p>

<p>I guess the main takeaway point is that in some scenarios softness at very small aperture might become an attribute instead of an issue.<br />Bob - thanks for including your model photo - lovely detail in that model house.</p>