smallest f stop for landscape photography?

[jean_marie_dederen]

<p>I have always wondered why on the one hand Ansel Adams & Co. used the label 'f64' to express their love for detail and realism; whilst on the other hand many lens reviews stay away from from the smallest apertures when defining the 'optimal performance' of a lens. I use a hasselblad for b/w landscape photography (I prefer wide open views with 50, 60 or 120mm lenses).</p>

[scott_murphy_photography]

<p>Keep in mind that Ansel used 8 x 10 or even <em>11 x 14</em> view cameras. In a format this large, smaller apertures are required because of the long focal length of the lens and remember that f/ is simply a ratio of the diameter of the diaphagm and focal length. As the format changes, the relative aperture also changes in terms of diameter of the hole. The greatest effect would be diffraction around the aperture blades. But when you are using a 6" lens, it is a whole lot different than using a 50mm lens.</p>

[sfcole]

<p>Probably because the larger the lens and format the less noticeable the effects from diffraction. I regularly shoot my Mamiya 6 at f16 or 22 without noticing image degradation. It also depends on the size of the print.</p><div></div>

[alec_myers]

<p>Diffraction effects which rob pictures of their sharpness at small apertures depend on the absolute size (in mm or inches) rather than the f-number. As Scott says, with large format cameras the focal lengths of typical lenses tend to be much longer for a given angle of view therefore you can stop down to a much smaller f-number to get the same size physical aperture (and therefore equally acceptable diffraction effects) as with, say, 35mm photography.</p>

<p>Stopping down to very small apertures helps to make up for the decreased Depth of Field that the long focal length lenses involved in LF bring.</p>

[kelly_flanigan1]

<p>It depends on how much the original negative is enlarged F22 might be ok for a 4x6 print off 35mm; or a 16x20 off of a 4x5 negative.<br>

<br /> If one uses say F64 or F90 ; then one negatives probably will not be as sharp as using F22.<br>

<br /> Whether this *matters* in a print depends on how much it is enlarged. For contact prints and 2x prints one really can stop the lens down and the softer negative does NOT matter; since the print-may only resolve 7 line pairs per mm<br>

One has also the issue of the print's veiwing didtance too</p>

[jean_marie_dederen]

<p>Never thought of the effect of the view/large format camera... Thanks; that's one more mystery of my list. Two more related questions, if I may. Is the middle range of the f stops on a given lens, as a rule of thumb, the safest one to go with (for detail)?<br>

And: maximum DOF is seemingly no guarantee for good detail?</p>

[jean_marie_dederen]

<p>Scott. Love the use of the square in your picture. But the lens, being a 50mm, almost by definition gives good detail right through anyway, or not? What would the same shot, in the same circumstances, look like with, say a 120mm? </p>

[mark_farrell]

<p>At the bottom of this tutorial link is a calculator that for a given format and some other parameters, will indicate if a particular aperature might be diffraction limited.</p>

<p><a href="http://www.cambridgeincolour.com/tutorials/diffraction-photography.htm">http://www.cambridgeincolour.com/tutorials/diffraction-photography.htm</a></p>

[jean_marie_dederen]

<p>Mark. Beautiful stuff, thanks.</p>

[sattler123]

<p>Jean-Marie, stop sweating the diffraction issue and use whatever f-stop it takes to get the effect you desire. What benefit do you get by using the "sharpest" aperture when half of your pictures is out of focus because the DOF wasn't what you needed. Or controversy, if you want to isolate your subject and need a large aperture. The best or sharpest aperture is a theoretical value that is pretty useless in real-life shooting situations. I never worry about it.</p>

[jpo3136b]

<p>+1 for don't sweat it. If that setting is on the camera, and that's what you want to use, use it. </p>

[jean_marie_dederen]

<p>DOF scales on my old C lenses haven't let me down, I must admit.</p>

[jenny_jaques]

<p>As Ansel Adams was referred to above, ... hmm ... his images with extraordinary DOF were achieved not only by stopping down, but by adjustment of the plane of focus > with tilts?</p>

[sfcole]

<p>Jean-Marie,<br>

Thanks. The point I was trying to make was that because I was using a MF lens, the detail is still at an extremely high level, even stopped down. If I had used the equivalent 28mm on my Canon at F16 or 22, the effects of diffraction would have been more noticeable. It's difficult to appreciate on screen--you have to see the photo above as a print. And BTW--I shot at about 1/15 second for that one, hand-held.<br>

Scott</p>

[stp]

<p>Mark, that is a very informative link -- I learned a lot just by playing with it. Thanks for posting.</p>

[rossb]

<p>Each sensor or film size has a diffraction limit. This is not a precise method but basically you just take 1/2 of the diagonal of you sensor or film size in mm and you have the f/stop that you can use. For an APS-c sensor that is probably f11 and for a 35mm camera it is f16. Of course a little testing on your part will help out in determining what you think will work out best for you. For an 8mp APS-c sensor you may find that f8 really is the limit. An 8x10 format camera can easily handle f64. To determine the diagonal you just use the basic math formula a^ + B^ = C^. </p>

[jean_marie_dederen]

<p>Scott. 1/15th handheld, the beauty of the rangefinder; although that 50mm lens must be quite bulky? Would the compur type lens of the hassie have any additional advantages (in terms of diffraction) over the standard lens design (as in your mamiya I suppose)?<br>

I'm still not clear on the issue of DOF. Juergen didnt make much sense when saying: ignore diffraction issue; and then gives the example where the 'sharpest f stop' (the smallest) would cause 'half of the picture to be out of focus'? How do you prioritise DOF choice or selection without 'sweating' about f stop? <br>

Ross. Thanks. the mathematical formula, is that old Pythagoras?</p>

 

[q.g._de_bakker]

<p>A thing i believe hasn't been mentioned yet is that differaction is always present, and always increases, gets worse, when you stop down.<br />The effect of diffraction alone is to halve resolution about every two stops a lens is stopped down. That, the same when going from f/2 to f/4 as from f/22 to f/45.<br />So you can do the sums: going to f/64 from, say, f/4 will reduce resolution, i.e. the amount of detail you will be able to capture, to 1/16th. Always. No matter what format.</p>

<p>The effect of resolution limiting diffraction is usually 'masked' at the wider aperture range by residual lens faults, which limit the amount of detail you can capture even worse than diffraction. These faults decrease when stopping down.<br />Hence the slight improvement stopping down a bit produces, until there is either no more gain by reducing faults, or the limiting effect of diffraction gets worse than that of the lens faults.</p>

<p>At something like f/64, you can be sure that it is diffraction that is limiting what you can get.</p>

<p>The above is all about the maximum achievable performance at the plane of focus. We usually do not take pictures of flat planes.<br />So the "f/64" thingy is about maximising DOF, i.e. getting more of the 3D scene in focus.</p>

<p>Like trading the limiting effect of lens faults against that of diffraction, the DOF game is a give and take. You give away the amount of fine detail you can record against the increased proportion of the scene you get in (reduced resolution) focus.<br />We have to choose.</p>

[jean_marie_dederen]

<p>Thanks, QG. so, detail/resolution/sharpness is one thing; focus/dof is another? but of course the two must relate somehow? surely both 'out of focus' and 'lack of detail' are 'unsharp? I still dont quite get it. I have been taking landscape shots with 120mm (hassie lens), an orange filter and shooting at f45 (thinking that would give me the best detail)... And I used 3200, 400, and 100 asa film b/w. What results can i expect? And one more question if I may? Does loss of detail at the smallest aperture relate to the (reduced) amount of light that gets through the lens?</p>

[dan_k2]

<p>Not to throw a monkey wrench into the works, but here's another thing to think about.... Could be referring to U.S. f-stops that kodak used into I *think* the 20's or 30's. That would be f/32 for U.S.64. I have an 12" lens on a century that goes to U.S. 256 (f/64) I saw that and went *HUH*??? until I figured it out.</p>

<p>Dan</p>

[jean_marie_dederen]

<p>Dan, that one puzzled me too, until I read Weston's diaries and tried to figure out his long exposures.</p>

[alec_myers]

<blockquote>

<p>Does loss of detail at the smallest aperture relate to the (reduced) amount of light that gets through the lens?</p>

 

</blockquote>

<p>Loss of detail due to diffraction: it's a wave effect, and can only be understood by thinking about the effect of the lens on the electromagnetic radiation (light) that hits it. The lens performs a transform (a fourier transform) on the waves so that the intensity of the electromagnetic field (in the plane of the lens) varies across the aperture according to the spatial frequency of the incoming wave. </p>

<p>The highest spatial frequencies, that represent the sharp edges and tiny details in the image, pass through the outermost edges of the lens.</p>

<p>When you close down the aperture you filter out the highest frequencies and lose the sharpness in the image. It's very similar to turning down the "treble" tone control on a hi-fi. </p>

[ondebanks]

<p>Alec, sorry, but your physics is awry - or at least missing vital statements, without which the wrong impression is given.</p>

<blockquote>

<p>Diffraction effects which rob pictures of their sharpness at small apertures depend on the absolute size (in mm or inches) rather than the f-number.</p>

</blockquote>

<p>This is true in angular units, but since photographers measure sharpness in linear units (lp/mm), it's untrue for a discussion such as this. </p>

<p>Angular resolution gets better with larger lens diameter (strictly, entrance pupil diameter), D. In angular terms, the diffraction spot gets smaller. But if focal ratio f=F/D is held constant, focal length F increases in the same proportion as D does. And since F is directly proportional to image scale (object magnification), it amplifies that smaller diffraction angle right back to the same linear size it was before.</p>

<p>Example - Let's say you have a 35mm-format lens, 50mm focal length, shooting at f5.6; and a 4x5 large format lens, 150mm focal length, shooting at f5.6; both optically perfect i.e. both diffraction-limited at f5.6. The LF lens has 3x F and 3x D of the 35mm lens, since f=5.6=same for both. With 3x D, the LF lens resolves angles three times smaller. With 3x F, it magnifies angles to three times larger on the film. Net result, the resolution (diffraction spot size) is the same, in mm. Key result, diffraction spot size <em>depends only on f-ratio</em> and not on the focal length, lens diameter, film format, lens design, or anything else.</p>

<blockquote>

<p>with large format cameras the focal lengths of typical lenses tend to be much longer for a given angle of view therefore you can stop down to a much smaller f-number to get the same size physical aperture (and therefore equally acceptable diffraction effects) as with, say, 35mm photography.</p>

</blockquote>

<p>This is incorrect. Stopping down to a smaller f-number gives you worse diffraction, not equal. Perhaps you are confusing diffraction with depth of field?</p>

<p>When Ansel Adams shot his 8x10 camera at f64, he obtained the same resolution on film as one would get shooting a 35mm camera at f64. That is, pretty crap...around 12.5 lp/mm.</p>

<p>But the difference of course is that he had nearly 60 times the film area, all objects in the scene were recorded about 7x larger in mm size, and he could make a 16x20 print with only a 2x enlargement (6 lp/mm), while 35mm film would need a 15x enlargement (0.8 lp/mm). Huge film real estate is why his images still appear to be "sharp" in the plane of focus, and view camera movements allowed him to tilt that plane for maximum effect. And this answers Jean-Marie's question.</p>

 

[q.g._de_bakker]

<blockquote>

<p>so, detail/resolution/sharpness is one thing; focus/dof is another? but of course the two must relate somehow? surely both 'out of focus' and 'lack of detail' are 'unsharp? I still dont quite get it.</p>

</blockquote>

<p>It is quite simple, really.<br>

A lens is supposed to form a point image of any point in the subject it is presented with. Due to diffraction, and aberrations, it doesn't. Instead of bringing all rays together in a point, it spreads them to form a tiny circle.<br>

To keep things clearly understandable, that circle was dubbed the "circle of confusion".<br>

The diameter of the circle determines how fine the detail is that you can still see (detail smaller than that circle, and detail that is separated by less than the diameter of that circle, will be lost in overlapping circles of light.<br>

The smaller the aperture, the larger that circle, the more detail below the size of that circle is lost.<br>

(Diffraction, the spreading of light, occurs where rays of light are limited, i.e. at the edge of the aperture. The ratio of edge to diameter changes in favour of edge when the diameter gets smaller. Hence the greater effect of diffraction at smaller apertures)</p>

<p>Focussing is a quite similar thingy. You set the lens to film distance such that the rays coming from points on the the bit you focus on are all brought together in a single point.<br>

If focus distance is off, the lens does not project points, but discs, circles again (the film plane intersects the cone of light in front of or behind where the point image would have been formed).<br>

So again, fine detail is lost. The greater the focussing error, the greater the circle the coraser the detail you can still resolve.</p>

<p>DOF is a phenomenon that exists grace to the fact that our eyes have a resolution limit too. If these circles of unsharpness are small enough, we just can not see that they are not the ideal point images but blurry circles.<br>

The diameter of the cone of light that was supposed to be brought to a point focus, but isn't because it is in front of behind of what we are focussing on, depands on the diameter of the aperture. Smaller aperture, smaller circles. The smaller the circle, the harder it becomes for us to see that it is not a sharp point but a blurry circle.<br>

So the smaller the aperture/the larger the f-stop, the more appears (!) to be in focus.</p>

<p>If you stop to remember that DOF thus is unsharpness that appears (!) acceptably sharp, a curious thing emerges. How sharp somehing appears, relative to what is indeed sharp, depends (bear with me: i know it's a tautology) on the difference between the sharp and unsharp bits.<br>

Now if the unsharp bits appear to be sharper (while they still are not), the apparent difference is smaller, DOF greater. Right?<br>

But if the sharp bit gets less sharp (due to diffraction), the apparent difference gets smaller, DOF greater too.<br>

So stopping down does more to DOF than the ubiquitous (and silly) DOF formulae suggest.</p>

<p>So, to summerize: the effect of both diffraction and not being focussed is that you lose resolving power. Fine detail is lost.<br>

The difference is that (ignoring diffraction) the loss of detail in DOF is only in the bits that are not in focus, but are deemed to be still sharp enough (or rather not yet unsharp enough). The loss of detail due to diffraction is everywhere, in every bit of the image.</p>

 

<blockquote>

<p>Does loss of detail at the smallest aperture relate to the (reduced) amount of light that gets through the lens?</p>

</blockquote>

<p>Nope.<br>

If it did, you could leave the shutter open longer, and regain sharpness, detail, that way. Sadly, that doesn't work.<br>

;-)</p>

 

[kelly_flanigan1]

<p>Here is a simple example.<br>

<br /> I purposely use a square image.<br>

<br /> I purposely us a wide set of cameras to show what is happening.<br>

<br /> Let us say the final print is to be 20x20 inches for all cameras below:<br>

<br /> the print is to be viewed close; ie at 1 foot; thus we place a criteria of say 5 line pairs per mm on the print as being sharp for this example</p>

<p>Case (1) MF negative with a usable 2x2 inch image.<br /> Case (2) 35mm negative with a usable 1x1 inch image.<br /> Case (3) 4x5 negative with a usable 3.5x3.5 inch image.<br /> Case (4) 8x10 camera negative with a usable 7.5 by 7.5 inch image.<br /> Case (5) our 24x36" process camera</p>

<p>With case (1) a 20x20 print is a 10 X enlargement; for 5 line pairs on the print; we need 50 on the negative. If one uses a 1500 rule of thumb for resolution times fstop; thus points to F30 being the smallest fstop; mabe say f22 practially.<br>

<br /> With case (2) a 20x20 print is a 20 X enlargement; for 5 line pairs on the print; we need 100 on the negative. If one uses a 1500 rule of thumb for resolution times fstop; thus points to F15. Few folks can get 100 line pairs per mm on film best case; thus 50 to 60 is what folks get under super conditions.<br>

<br /> With case (3) a 20x20 print is a 5.7 X enlargement; for 5 line pairs on the print; we need 29 on the negative. If one uses a 1500 rule of thumb for resolution times fstop; thus points to F 51. On can probably shoot at F22 thru F45 and see the same 20x20 prints results<br>

<br /> With case (4) a 20x20 print is a 2.67 X enlargement; for 5 line pairs on the print; we need 13 on the negative. If one uses a 1500 rule of thumb for resolution times fstop; thus points to F 110. On can probably shoot at F22 thru F128 and see the same 20x20 prints results<br>

<br /> With case (5) a 20x20 print is a contact print ie NO enlargement; for 5 line pairs on the print; we need 5 or abit more say 7? on the negative. If one uses a 1500 rule of thumb for resolution times fstop; thus points to F 200 to F300. On can probably shoot at F22 thru F256 and see the same 20x20 prints results</p>