Open Photography Forums

Open Photography Forums (http://www.openphotographyforums.com/forums/index.php)
-   Lenses: DSLR and Rangefinder, MF adaptions to 35mm such Zoerk (http://www.openphotographyforums.com/forums/forumdisplay.php?f=17)
-   -   f/22, soft edges and landscapes (http://www.openphotographyforums.com/forums/showthread.php?t=14317)

Nick Masson June 30th, 2011 09:11 PM

f/22, soft edges and landscapes
 
Hey all,
I have a question about technique when it comes to dealing with long depth-of-field and lenses that go soft in the edges when they get stopped down...

I am shooting with some older Nikkor lenses which I have chosen to work with because their small profile and weight allow me to bring them places I otherwise wouldn't or couldn't go (20mm Ai-S f/3.5, 24mm Ai-S f/2.8, 28mm Ai-S f/2.8, etc...).

Most of the time this setup is best suited for images with strong subjects and potent image value, but I do enjoy my share of tripod-mounted landscape work. These are not the best for lanscapes (given good wide lenses now, 17-35 etc...), but they are what I have.

To the point: When my subject is far off, everything is effectively at 'infinity' in as far as focusing. Sure nothing is truly at infinity, but there is no discerning between an object 1mile off and an object 2 miles off when focusing.

I have been told that it is still wise to use a small aperture in this case to maintain sharpness throughout the photograph, but is there really a difference between f/8 and f/22 when my subject is very far away?

I only ask because these small older lenses tend to get soft edges when stopped down, and are sharpest around f/8 (usually the case with most lenses), so I fear that whatever I may gain in shooting at f/22 instead of f/8 I would loose due to softening edges in the small-lens optics...

Any insight would be appreciated!
Thanks,
-Nick

Asher Kelman June 30th, 2011 09:28 PM

Quote:

Originally Posted by Nick Masson (Post 117311)
Hey all,
I have a question about technique when it comes to dealing with long depth-of-field and lenses that go soft in the edges when they get stopped down...

I am shooting with some older Nikkor lenses which I have chosen to work with because their small profile and weight allow me to bring them places I otherwise wouldn't or couldn't go (20mm Ai-S f/3.5, 24mm Ai-S f/2.8, 28mm Ai-S f/2.8, etc...).

Most of the time this setup is best suited for images with strong subjects and potent image value, but I do enjoy my share of tripod-mounted landscape work. These are not the best for lanscapes (given good wide lenses now, 17-35 etc...), but they are what I have.

To the point: When my subject is far off, everything is effectively at 'infinity' in as far as focusing. Sure nothing is truly at infinity, but there is no discerning between an object 1mile off and an object 2 miles off when focusing.

I have been told that it is still wise to use a small aperture in this case to maintain sharpness throughout the photograph, but is there really a difference between f/8 and f/22 when my subject is very far away?

I only ask because these small older lenses tend to get soft edges when stopped down, and are sharpest around f/8 (usually the case with most lenses), so I fear that whatever I may gain in shooting at f/22 instead of f/8 I would loose due to softening edges in the small-lens optics...

Any insight would be appreciated!
Thanks,
-Nick

Nick,

Your findings are not unique to "old lenses"!

Yes , your intuition that lenses "may not be sharp stopped down, especially at the periphery", is due to the fact the most lenses are not flat field, are better in the center and have higher resolution at worthwhile contrast when open at just a little off from the widest aperture. Some lenses are even designed to be totally sharp wide open. As one closes the aperture, the depth of field increases as you have sought with your f22 choice. However, there are costs! At smaller apertures, the resolution often gets less! Also, there's a second complication, namely diffraction. One eventually approaches practical degradation of the focus, due to the tiny aperture itself! In this respect, as a rule of thumb, with digital sensors and small format, one is O.K. up until 5.6, in almost all cases, and then by f11 there's some degradation. At f 16 or f 22, with tiny photosites, the degradation may actually become visible to you at the size you choose to print. Test it with your lens.

In order to get the best of DOF and sharpness with small sensors, a procedure to consider is taking pictures, where possible, using 5.6 to 11.0. To do better, you might look into "focus stacking" to build images with greater DOF, sharp through the zones of your interest. Especially if you are in one fixed place and with a tripod, focus stacking and exposure bracketing is possible. There are even programed stages for this. We've just discussed this with macro images! The great advantage of a digital camera is that it obeys rules! As long as you collect the right set of images, you can do whatever you wish!

Asher

Doug Kerr June 30th, 2011 09:37 PM

Hi, Nick,

I will first think in terms of the "acceptable blurring" criterion commonly used for the full-frame 35-mm format size.

Let us think of a 28 mm focal length. Consider an aperture of f/8.

Then the hyperfocal distance would be about 3.2 m (10.4 ft). That means that if we focus at that distance, everything from 5 feet to infinity would be within our criterion of acceptable blurring.

Now, since your emphasis is in "essentially infinite" parts of the scene, we can consider focus at infinity. Now, everything from just about 10 feet to infinity would be within our criterion of acceptable blurring.

So there would be hardly any advantage of going smaller than f/8 with regard to depth of field.

And almost certainly, as you did go to a smaller aperture, you would see degradation due to diffraction.

Your best bet would probably be so shoot at the aperture at which your lens (focused near infinity) delivered the best resolution (as found in various reports). (That often turns out to be around f/8, but who knows.)

If we adopted a maximum acceptable blurring 1/5 that usually used (a much more demanding criterion), then (at 28 mm and f/8) we could focus at about 50 feet, and everything from about 25 feet to infinity would be exhibit "acceptable" blurring in that very stringent sense.

This bounty of course visits us because of the use of a small focal length. Were you speaking of shooting with a 400 mm lens, the story would be quite different.

Best regards,

Doug

Jerome Marot June 30th, 2011 11:35 PM

Quote:

Originally Posted by Nick Masson (Post 117311)
These are not the best for lanscapes (given good wide lenses now, 17-35 etc...), but they are what I have.

I don't agree with that idea. These primes, closed to around f/11, are probably the best you can use for landscape photography.


Quote:

I only ask because these small older lenses tend to get soft edges when stopped down
What happens exactly at f/22: does the whole image gets softer all over or does sharpness degrades more rapidly on the sides of the frame than at the middle?

Nick Masson July 2nd, 2011 08:49 AM

Thanks for the replies. It seems, then, that there is little use in stopping down to f/22 or so unless your composition includes a subject quite close to the lens (5 to 10ft).

Asher and Doug, would you suggest any literature about optics resolution, and particularly hyperfocal technique etc... It is something that has always interested me. And I am working with film right now, so unfortunately I don't have the digital luxury of multiple exposure techniques.

Jerome, I find that the images get noticeable softer in the edges when stopped down. I assume they're also less sharp in the center region, but I haven't done any empirical trials of side-by-side images.

Michael Nagel July 2nd, 2011 11:09 AM

Nick,

These articles could help you for a start:
Diffraction and Optimum Aperture - Bob Atkins

Digital Camera related:
Diffraction limits of resolution - Andre Gunther

Lens Diffraction & Photography at Cambridge in Colour

Best regards,
Michael

Bart_van_der_Wolf July 2nd, 2011 11:44 AM

Quote:

Originally Posted by Nick Masson (Post 117346)
Thanks for the replies. It seems, then, that there is little use in stopping down to f/22 or so unless your composition includes a subject quite close to the lens (5 to 10ft).

Hi Nick,

The unsharpness that's added as a veil over the entire image is directly related to the F-number. The F-number itself is a ratio of focal length and aperture diameter (therefore also written as f/#), and is the most important factor in determining diffraction blur. Small apertures, therefore large F-numbers, produce a lot of diffraction blur.

Distance is only important for the Depth of Field (DOF) blur at certain distances away from the plane of focus.

Quote:

Asher and Doug, would you suggest any literature about optics resolution, and particularly hyperfocal technique etc... It is something that has always interested me.
One of the most accurate on-line descriptions can be found at:
http://toothwalker.org/optics/dof.html and
http://toothwalker.org/optics/dofderivation.html

And of course Doug has written a paper about the subject, but his webpage seems to be not responding right now.

Cheers,
Bart

Doug Kerr July 2nd, 2011 03:51 PM

Hi, Bart,

Quote:

Originally Posted by Bart_van_der_Wolf (Post 117356)
And of course Doug has written a paper about the subject, but his webpage seems to be not responding right now.

Works from here. Perhaps you have the old URL. The current URL is:

http://dougkerr.net/Pumpkin/

Best regards,

Doug

Bart_van_der_Wolf July 2nd, 2011 05:17 PM

Quote:

Originally Posted by Doug Kerr (Post 117362)
Hi, Bart,



Works from here. Perhaps you have the old URL. The current URL is:

http://dougkerr.net/Pumpkin/

Hi Doug,

Yes, Google has a few old ones stored. Let's generate some traffic on the newer ones for them to at least show up at the top of the list. Thanks for your collection of recommended papers/essays/gems.

Cheers,
Bart

Maris Rusis July 3rd, 2011 06:18 PM


Gelatin-silver photograph on Fomabrom Variant III exposed in contact contact with a Fomapan 200 8x10 negative developed in Xtol. The camera was a Tachihara 810HD triple extension field view camera with a Fujinon W 300/5.6 lens. Exposure was 1 second at f90.

This isn't a digital image but it illustrates the uncomfortable compromises needed to balance lens resolution, image sharpness, image detail, and diffraction. The lens, a Fujinon-W 300/5.6 is maximum sharp in the centre of the field half way between f5.6 and f8. When used as an astronomical objective it gives me good views of Saturn and its rings; it's that sharp. But for the deep focus Mountain Image all that is of no use.

The mountain-shaped rock is only two metres in front of the camera and the most distant mountain (Mt Townsend, Australia's second highest) is 10 kilometres away. How to bridge the focus gap?

Tilting the front and back of the camera puts the Scheimpflug plane half-way up the near rock ( lichen patch) and through the top of the distant mountain. The tip of the rock and the herbs along the bottom edge are way outside the depth of field and rather blurry. Under the dark focussing cloth and with a 8x magnifier on the ground-glass I stop the lens down slowly and watch what happens.

The lichen patch in the plane of focus gradually loses sharpness as I stop down. The bottom edge herbs and rock tip gradually get sharper as the increased depth of field swallows them. Eventually at f90 the tiny leaves are as sharp as they can get; f128 is definitely worse.

It is a dreadful compromise. A lens that can image the moons of Jupiter is delivering an image infected with diffraction at only about 5 lines to the millimetre. This still looks sharp to the eye, just. One more stop down, f128, and the whole picture would look soft. I've destroyed resolution, I've damaged sharpness, I've let diffraction run riot, but I have maximised the total amount of detail in the photograph.

I reckon out-of-focus is a bigger killer of image sharpness than diffraction and sometimes it's worth taking things to the limit.

Doug Kerr July 3rd, 2011 08:05 PM

Hi, Maris,

A very nice, and useful, essay. Thanks.

Best regards,

Doug

Bart_van_der_Wolf July 4th, 2011 03:20 AM

Quote:

Originally Posted by Maris Rusis (Post 117394)
[...]
It is a dreadful compromise. A lens that can image the moons of Jupiter is delivering an image infected with diffraction at only about 5 lines to the millimetre. This still looks sharp to the eye, just. One more stop down, f128, and the whole picture would look soft. I've destroyed resolution, I've damaged sharpness, I've let diffraction run riot, but I have maximised the total amount of detail in the photograph.

I reckon out-of-focus is a bigger killer of image sharpness than diffraction and sometimes it's worth taking things to the limit.

Hi Maris,

Indeed, we photographers are faced with compromises imposed by physics. We use our knowledge of these compromises to optimise the solution.

Your example works fine, because the large negative you shoot on, needs relatively low magnification to produce a certain size output. A smaller film size would require more output magnification, e.g. a 36x24mm film frame needs 7-8x more magnification for the same output size. That means that it must be 7-8x sharper to give the same output resolution. Therefore f/64 must become f/8 and because we can use a shorter focal length for the same field of view, the DOF will be similar.

That's why output magnification is such an important factor in judging whether a certain level of compromise is tolerable.

Of course there is sometimes also the possibility to use focus-stacking, when the subject is stationary. For certain subjects it offers the best of both worlds, shooting at the lens' optimal aperture, and DOF where we want it.

Cheers,
Bart

Doug Kerr July 4th, 2011 04:46 AM

Hi, Bart,

Quote:

Originally Posted by Bart_van_der_Wolf (Post 117404)
Indeed, we photographers are faced with compromises imposed by physics. We use our knowledge of these compromises to optimise the solution.

Well said.

Quote:

Your example works fine, because the large negative you shoot on, needs relatively low magnification to produce a certain size output. A smaller film size would require more output magnification, e.g. a 36x24mm film frame needs 7-8x more magnification for the same output size. That means that it must be 7-8x sharper to give the same output resolution. Therefore f/64 must become f/8 and because we can use a shorter focal length for the same field of view, the DOF will be similar.
An important point, again well said.

Best regards,

Doug

Jerome Marot July 4th, 2011 10:11 AM

Actually, on small formats, the effects of diffraction are limiting the photographer's choices quite a lot nowadays. I'll give some examples:

-on 35mm film, or on a 12 mpix full frame DSLR (which is about the same resolution), the effects of diffraction are measurable from f/8 and start to be visible over about f/11. Considering that typical 35mm primes were open around f/2, f/2.8, and because aberrations tend to diminish with aperture, we have that idea that lenses have a sweet spot between f/5.6 and f/11...

-on a 12-16 mpix crop format DSLR (or on a 24 Mpix full frame DSLR), because we have higher linear resolution, diffraction starts to be measurable from f/5.6 and starts to be visible over about f/8. In addition, the typical zoom lens usually has a maximum aperture of f/4 or f/5.6 so, basically, all what we can choose is full aperture or closed one stop to improve corners a bit. Anything above that and we will start to degrade center sharpness.

-on the typical digital P&S, diffraction usually start to be a problem as soon as you close the diaphragm. Luckily, these cameras usually have no diaphragm... ;) Some use a neutral density filter instead. This is also known from professional video cameras (which also have a small sensor, but lower resolution -HD is 2 Mpix- and usually faster lenses): don't close the iris (the video name for diaphragm) above f/5.6, use a ND filter instead. Many of these cameras have a set of ND filters built-in for that reason.


Of course, there is more to a picture than absolute sharpness. Sometimes we need more depth of field and need to compromise resolution. Sometimes we need less depth of field and don't care about optical aberrations in the corners. But for a reasonably flat subject (landscape far away), these are the constraints we have. In other words: don't go above f/11 (f/16 for film) unless you must.

Mike Shimwell July 15th, 2011 04:00 PM

Just an aside. A few years ago I bought my father a copy of Mountain Light by Galen Rowell (deceased). Actually it's more than a few years ago because it's the first edition dating from 1986:)

Throughout the book Rowell talks about using small and very small apertures to maximise depth of field and, it's certainly the case, that a lot of them are not really very sharp. However, they remain a set of powerful photographs and quite an interesting document of their time as well.

Mike


All times are GMT -7. The time now is 10:34 PM.


Posting images or text grants license to OPF, yet of such remain with its creator. Still, all assembled discussion 2006-2017 Asher Kelman (all rights reserved) Posts with new theme or unusual image might be moved/copied to a new thread!