2
\$\begingroup\$
  1. This image was shot with a Canon 5D (the older one) with a EF24-105mm f/4L lens.
  2. The lens was set at 45mm at f/11, and the shutter speed was 1/90 of a second which was +0.5 from the meter at ISO 200.
  3. The white balance is set at 5700K and it is shot as a raw file and processed in Photoshops Raw file converter.

This photograph has been taken by someone else.

Anyway, my aim to shoot landscapes is to get crystal clear pictures with ultra sharpness that too without spending "hours" on the GIMP. So, F16 (the least aperture) on a common DSLR will be resulting in diffraction.

In the below picture, the leaves, the grass, the bark of the tree etc. is NOT sharp, if fact it is blurry.

1. With a DSLR of min aperture F16, what level of sharpness should be expected? Can anyone post some sharp photographs of landscapes?
2. Is there some other kind of DSLR which can give more sharp results than the following one?

Autumn, by Michael Gustafson on 1x.com

EDIT 1:

Can you explain further what you perceive as "blurry"? – mattdm

Well, by sharp landscapes I mean, that the I should be able to "see" the shapes,and the texture of the "leaves", the "bark" of the tree, the "blades" of the grass etc.
In that picture, I cannot figure out even one texture there. That picture shows the beautiful colours, ONLY.

\$\endgroup\$
8
  • 1
    \$\begingroup\$ I love the colors in this picture--with such great tones, I wouldn't mind a little blurriness one bit. The Canon EF 24-105 f/4L is an excellent lens, capable of producing very sharp photos. The Canon 5D is also an excellent camera, if a little dated. The aperture and shutter speed settings also sound fine, regardless of whether the shot was taken handheld or on a tripod. Without knowing anything else about the shot, I'd guess either the focus is slightly off, or maybe the camera was mounted on a tripod and the IS failed to detect it. \$\endgroup\$
    – user2559
    Dec 19, 2011 at 4:56
  • \$\begingroup\$ @KimBurgaard He says that there was a little fog? Perhaps that's effecting the things? 1x.com/#!/forum/photo-critique/30142 \$\endgroup\$ Dec 19, 2011 at 5:01
  • 2
    \$\begingroup\$ @Anisha, I think we need to define terms a little bit here. That image looks sharp throughout to me, but it's really hard to tell, as it appears that only a 950×633 pixel version is available to us. Downscaled that much, just about anything will look sharp, regardless of focus error, lens defects, or anything else. Can you explain further what you perceive as "blurry"? \$\endgroup\$
    – mattdm
    Dec 19, 2011 at 6:26
  • 1
    \$\begingroup\$ So, to address your edit: you can't see that detail because the image is not high enough resolution. At this size, a single pixel covers a large area, so each leaf is represented by only a few pixels. That's simply not enough definition to show texture. The detail you want may or may not be there in the original (or in a nice print of the photo). \$\endgroup\$
    – mattdm
    Dec 19, 2011 at 12:11
  • 1
    \$\begingroup\$ Sharp enough is when the image is sufficiently sharp to convey the tone desired by the photographer. That may be as blurry as all get out as a result... \$\endgroup\$
    – Joanne C
    Dec 19, 2011 at 14:16

4 Answers 4

14
\$\begingroup\$

On Sharpness

Sharpness is a factor intensely related to viewing context. To use more laymans terms, how sharp a photograph looks greatly depends on how large it is viewed, how closely it is being viewed, how well it is lit (if not on a computer screen), and the visual acuity of the viewer. As such, it is very difficult to judge sharpness without a proper context.

Pixel Peeping

First and foremost, be keenly aware of the differences of viewing a photograph at 100% scaling on a computer screen (especially a nice, large, professional grade screen that is properly calibrated!) The act of examining the image quality of a photo at 100% scaling (usually referred to as 100% crop) is called pixel peeping. Pixel peeping, amongst other things like eye fatigue and neck pain, is one that will lead down the deep, dark, endless rabbit hole of IQ Worry! Humor aside, there is a very big difference between IQ (image quality) at 100% crop and IQ of a fully processed photograph that has been cropped, scaled, de-noised, and sharpened to your liking for your desired presentation medium.

Pixel peeping has its place, and when used properly, is an extremely valuable technique in the digital photographers post-processing toolbox. Its helpful to view a photograph at 100% when removing noise, as noise tends to be "absorbed" by downscaling (i.e. viewing a photograph at less than 100%). It is also helpful when analyzing a photograph for lens aberrations or other defects that might affect IQ, such as spots from dust on the sensor, etc.

The Reality of Image Quality

Image quality, of a landscape, macro shot, or anything else that depends on high detail definition and sharpness is a bit of a chimera, a highly subjective thing that has real-world, mathematically definable bounds. (Contrasted with say portraiture, wherein a little softness and thinner DOF is actually a very desirable thing.) Image quality is dependent upon intrinsic factors, such as sensor resolution, focus, camera shake, as well as extrinsic factors such as presentation medium, viewing distance, and the visual acuity of the viewer.

Intrinsic factors definitely play a role, however how much they affect IQ outside of the realm of pixel peeping can be affected by extrinsic factors. A lower resolution sensor will be less capable than a higher resolution sensor at capturing all the detail in an image that has fine detail...such as a landscape shot containing lots of trees and other flora. The resolution of the lens used at the aperture shot at will also play a role, imposing diffraction limits on the maximum amount of detail that can be physically resolved. The presence or lack thereof of image stabilization technology, or the steadieness of the photographers hands if handheld, can also play a role in sharpness. How well the photo is focused will most certainly play a role, however it is entirely in the hands of the photographer, and is thus a controllable factor. These all impose a physical limit on how much intrinsic sharpness exists in a photograph.

Extrinsic factors can determine whether the intrinsic sharpness of a photograph will affect the final presentation format or not. In the most common cases, intrinsic factors like resolution, camera shake or defocus rarely have an impact unless they are fairly severe. In less common cases, usually in the case of prints blown up more than two times the native size of the photograph, small factors can be enhanced and become more apparent.

Common Presentation Formats and Relative Sharpness

In the general case, most images are viewed at significantly reduced size relative to their native size (100% crop size). This is true for both print and digital viewing mediums.

When published to the internet, the most common presentation size for the largest image dimension ranges between 500 to 900 pixels. With the average image size for modern high megapixel cameras ranging from 4000 to nearly 6000 pixels in the largest dimension, that makes the common internet presentation format about 5-12 times smaller! Sharpness when scaling down so significantly is rarely a problem, and what may be observed as marked softness or blurrieness when pixel peeping will most likely be entirely absorbed when downscaling. Additionally, when downscaling by such a large degree, a significant amount of noise will usually be absorbed, entirely nullifying what may appear to be detrimental degradation of IQ at 100% crop.

When publishing to printed mediums, the more common presentation sizes tend to be smaller than how large the image may be at 100% crop on a computer screen. With the highest megapixel cameras today, native print size is as large as or slightly larger than 13x19" at 300ppi. On top of that, print resolution tends to be 3-4 times higher (more dense) than computer screens. This is where human visual acuity comes into play, which clocks in at around 300-360ppi for an average print (say 8x10 - 13x19) viewed at a proper viewing distance (say a foot and a half). At such resolution, what may appear to be intrusive noise, problematic optical aberrations, or detrimental diffraction softening will generally be beyond the ability of human sight to recognize.

Less common presentation sizes include enlargements. Unlike reductions or prints at native size, enlargements can exacerbate IQ problems visible when pixel peeping. Keeping in mind that the average printer resolution is 3-4 times higher than a computer screen, it usually takes about a 3-fold enlargement for a printed pixel to approach the size of a computer screen pixel. Enlargements up to around 2x or so are usually devoid of any noticeable issues with sharpness or problematic enhancements to optical aberrations or noise. Enlarging a photo beyond 2x, however, will most likely make all the defects visible at 100% crop visible in the print, and put more pressure on the photographer to capture a sharper photo on site.

What is Sharp Enough?

The enhancement of IQ detractors when enlarging doesn't spell the end of the world, however. Scaling for enlargements greater than 2x is usually done with a fair amount of care, with meticulous attention paid to noise and finely tuned sharpening adjustments added. Enlargements do put greater pressure on the photographer at the time the photograph is taken to maximize sharpness. If you do not intend to enlarge, then the diffraction softening from an f/16 aperture on an APS-C camera is most likely a non-issue. I've often shot landscapes at f/22 on a relatively lower resolution Canon 450D, where I know for a fact that diffraction is affecting IQ, and its never been a problem in prints up to 17x22.

Diffraction at f/16 for a modern DSLR (APS-C or FF) is probably of least concern, as achieving perfect focus and ensuring no camera shake are both likely to be the primary detractors to sharpness. Minor effects of chromatic aberration are also usually a non-issue for most presentation formats, however they do have the tendency to make post-process exposure tuning and sharpening adjustments produce funky double-edges to finer details like tree branches, mountain/sky borders, etc. Scaling for internet publication will usually absorb it, however medium sized prints may exhibit issues.

If you do wish to enlarge your photos in print, the camera might already be outresolving the lenses. More important to improving image sharpness in the camera is high quality glass that is capable of resolving fine detail throughout its focal range. A higher resolution sensor will also help, and finding one in a larger format will help even more. A staple of many landscape photographers is the Canon 5D Mark II with either the 17-40mm L or 16-35mm L lens. That combination ensures that the lens is resolving enough detail for the sensor, and that the sensor is large enough to support a high resolution (in this case, 21.1mp). The issue is that the 5D II + L series lens is about a US $3200-$4200 investment.

A top of the line 5DII with an L-series lens is not necessary to get sharp landscape photographs. A canon 550D, 600D, 60D, or 7D with the EF-S 10-22mm lens will also do the job quite well. You won't have the ability to create huge enlargements that can be viewed within a few feet like you might with the 5DII and a 16-35mm lens, but you should be able to produce prints up to 17x22, or any size smaller than that, as well as any format for presentation on the web, without any need for concern about sharpness beyond the factors you can control in-camera.

UPDATE: What Details Matter?

In response to the updated question about whether small details matter. That is really a matter of composition. When it comes to landscapes, whether you see bark on a tree trunk or not, or the veins in the leaves, is entirely about how the scene is composed. Most landscapes are about capturing, in breadth and depth, an amazing landscape. Landscapes, given their expanse, usually do not contain much room for finer detail like bark or leaf veins. In some cases, particularly with ultra wide angle lenses, you can compose such that you get very close to and capture a lot of detail in the very near foreground (rocks, the shore of a lake, a tree stump or fallen tree, etc), while still capturing the greater expanse of a landscape in the background. In such a composition, it would be important that foreground detail is sharp and clear. That is all about focus, and in many cases you can focus such that the foreground is crisp and clear, and the backdrop of the landscape is just on the edge of sharp, but still within DOF enough that the average observer won't notice the slight softness of details way off in the distance.

There is no simple rule you can follow regarding what details matter in a landscape. Generally speaking, fine details are not as important as the whole scene, however when photographing with wide angle, near objects and their detail may be important...its ultimately a matter of choice on the part of the photographer.

\$\endgroup\$
6
  • 3
    \$\begingroup\$ ++ for not being worried about how things look at 100% magnification on a computer screen \$\endgroup\$
    – jwenting
    Dec 19, 2011 at 8:29
  • \$\begingroup\$ @jwenting: Aye!! I've been poking around other photography forums lately...and WOW! People are SO hung up on how things look at 100%!! I recommended the 7D to some guy, and a couple days later he was "earthshatteringly disappointed with 7D!" because, at 100%, in Lightroom, noise appeared really bad. Nothing to say about how its 48% more pixel dense than his previous 450D, or that despite a VERY SLIGHTLY higher noise at ISO100, it was capturing far more detail. People are pretty insane about their 100% quality, when in the vast majority of cases, it doesn't matter a wit. \$\endgroup\$
    – jrista
    Dec 19, 2011 at 20:33
  • 3
    \$\begingroup\$ +1 for pointing out the rabbit hole and then offering a guided tour around it. :) \$\endgroup\$
    – mattdm
    Dec 19, 2011 at 21:45
  • \$\begingroup\$ @jrista in my experience Canon users tend to be far worse for that than Nikon users. Many of them are obsessed with technical minutiae like that, and obsessed with always having the very latest equipment as well (I've known guys who'd preorder every new model Canon announced even if they'd not even received their previous preorder yet, and would instantly start slagging that 3 months older model they'd never used yet as utter rubbish when a few days earlier it was (unused) the best thing since before sliced bread). \$\endgroup\$
    – jwenting
    Dec 20, 2011 at 6:14
  • 1
    \$\begingroup\$ @jwenting: Thats about as naive a comment as I've seen you make on this site. EVERY brand has its insane, excessively hard-core, uber-diehard fanboys. Ever spent time on NikonRumors.com? Ever heard of Ken Rockwell? Its not just Canon, nor is it just Nikon or just anyone else. Create anything that obsessive people can obsess over, and they will. Nikon users are no different than Canon users in that respect. \$\endgroup\$
    – jrista
    Dec 21, 2011 at 18:30
3
\$\begingroup\$

This is a big question with a lot of variables!

Firstly diffraction: I have an 18-200 zoom which suffers a bit at f/22 and f/32. I have a 17-50 zoom which is sharp through f/16, and f/22 is only noticeably softer if you are looking at a test pattern. Real world shots you really can't tell the difference. My 50mm prime lens is sharp from f/2.8 to f/32. I could not notice any falloff due to diffraction. Although the web is full of articles on the subject, with photographic evidence, I'm not convince from personal experience that it's as big a problem as it's made out to be.

If you need to shoot at f/11, f/16 or f/22 to get the necessary depth of field from your foreground to background, you are probably better off shooting at f/22 then shooting at an "optimal" f/8 and losing sharpness due to shallower DOF.

Sharpness depends on the quality of the lens, the aperture, the precision of your focusing, lack of camera/mirror shake, and the resolution of your sensor, among other things.

If shooting in RAW, the image will not be sharpened until you post process it, and the amount of sharpening is up to you.

As to the above picture, I don't know the quality of that lens. I don't know where the focus point was, whether tree leaves were being blown by the wind, or how much sharpening was applied. At 45mm and f/11 there should have been ample DOF though, as far as I can tell.

If your lens is sharp enough (and I doubt a 24-105mm zoom will be), then you can get more resolution out of a larger sensor. The above picture was taken with a full frame camera, but could be improved by using a sharp prime lens, and possibly better technique (tripod, mirror up, remote shutter release, good focus technique, post processing sharpening, etc - we of course don't know what technique was used in the above photo)

Sharpness is all about local contrast. The above image has a lot of dark areas. It's going to be hard to perceive sharpness in the tree trunks for instance. HDR techniques might help to improve the perceived sharpness.

\$\endgroup\$
1
  • \$\begingroup\$ Good answer. I've also found the diffraction to be a little overblown. Diffraction is real and it will start to eat away at resolution, but focus and exposure play a much larger role. In short, if you need to stop down to f/16 or f/22 to get everything is focus, just do it. \$\endgroup\$ Dec 19, 2011 at 5:35
2
\$\begingroup\$

First, a clarification. The minimum and maximum aperture is a function of the lens, not the camera. The maximum and minimum f-stop values are determined as the ratio between the focal length and the size of respectively the minimum and maximum aperture opening of the lens.

The maximum sharpness a lens is capable of producing depends on the lens construction, lens coatings, the aperture, as well as any filters put in front or behind the lens.

The sharpness of a lens is often expressed as resolving power or spatial resolution, i.e. how fine detail is the lens capable of transmitting to the sensor or film. Lens manufactures will often provide MTF charts that visualize the resolving power across the field of view. In and of it self, the MTF means little to me, but generally, the better the lines are able to stay flat at the top of the chart, the better performance you can expect in terms of consistent sharpness throughout the entire picture.

That being said, even the consumer grade lenses from the major brands are generally capable of producing very good results when used with a tripod and stopped down to f/8 or slightly smaller apertures. In your particular example, I know from experience that a properly working copy of the EF 24-105mm f/4L lens is capable of producing very sharp photos.

I'd recommend you try out equipment within your budget range to see if it satisfies your needs. Borrow or rent if you can, or buy from a shop that allows you to return the equipment if it doesn't hold up to a field test. In the end, seeing actual results beats any amount of MTF charts and theoretical talks about aperture, resolving power, diffraction limits, and...

\$\endgroup\$
1
\$\begingroup\$

First I want to talk about what is sharp enough:

  1. Normal people (that is, non photographers) don't view photos at 100%, they don't print photos at huge sizes and even if they do they view them from a long distance.

  2. Normal people don't look at how sharp the photo is, they see colors, contrast and how interesting the subject but they couldn't care less about the technical qualities of the photo.

As a result "sharp enough" is pretty soft in most cases.

Now that I said that, if you like to see texture in the grass and the trees you should try to take much-sharper-then-enough photos because this is art - and the point is to realize your vision as an artist (even if normal people will never be able to tell the difference).

So, what you need to do in order know how to take sharp pictures and at what point diffraction kicks in is to go to the nearest place to your home that has trees and grass with YOUR camera and YOUR lens (and your tripod and your taste) and the time of day you usually shoot landscape at.

Put you camera on the best tripod you have, switch IS off and mirror lockup on and use a cable release or the 2 seconds timer to trigger your camera.

And take lots of test pictures at different settings, take a picture at every aperture value from wide open to smallest, use different lenses, try different zooms if you use a zoom lens, etc.

And remember motion blur - things move, the grass and trees are moving in the wind (even if the wind is not strong enough for you to notice it) and even the most massive tripod is not completely stable - at the level of details you are looking for and at the shutter speed needed for f/16 aperture at classic landscape light I'd be surprised if diffraction has more effect than motion blur.

\$\endgroup\$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.