Luminous Landscape Forum
Equipment & Techniques => Landscape & Nature Photography => Topic started by: Jim2 on September 01, 2012, 06:08:42 am
-
I don't know what's wrong with what I'm doing, so I hope someone more knowledgable can help me.
I took just took this photo today at sunset and while I'm happy with the image, I'm not happy with the sharpness.
(http://a2.sphotos.ak.fbcdn.net/hphotos-ak-snc6/257012_10151132864994826_530144150_o.jpg)
100% Crop attached.
My gear:
- Canon 1ds3
- Sturdy tripod (gitzo 3 series 4 section) + head (arca swiss cube)
- Singh Ray Grad ND (resin, not glass - and yeah it's not the cleanest - no scratches on it but it probably has some oily smudge but it is not noticeable - I did try to wipe it with a microfibre cloth the best I could - maybe I need a new cloth?)
- 24-105mm zoom
- Using Mirror lock up, cable release
- Shot at 24mm F/20 1.3s
It wasn't a windy day - from the image there is no 'shake'.
The exposure was underexposed by about 1 1/3 or 1 2/3 and then I adjusted exposure using Lightroom, and also Increased the "Fill Light".
So how can I achieve a nice clean / sharper image?
I want both the grass at the bottom (first 100% crop) and the mountain to be tack sharp.
What do I need to change?
-
F/20??? There you go!
P.S. Google two words: difraction and hyperfocal
P.P.S. Nice image, btw
-
Hello Slobodan:
Googled the words and found some very useful info regarding hyperfocal distance. I am sure it will make my pictures sharper. Thanks for the same.
Praki.
-
A couple of further observations:
1. There was absolutely no need to underexpose, especially not by that much. That results in noisy shadows, further robbing the image of cleanness/crispness. Check the Expose-To-The-Right (ETTR) concept by searching this site.
2. You should remove CA (chromatic aberration) in post-processing, quite visible in your corner crop. That would make the image a bit cleaner/crisper too
3. I would not say that sharpness (or alleged lack of) is a problem in your image. It looks sufficiently sharp to me. Yes, shooting at less diffraction-prone f/stops and using hyperfocal distance focusing would result in a slightly sharper image, but that would only be visible in a direct comparison and in huge enlargements and/or 100% pixel peeping.
4. Absolute sharpness, front to back, is overrated imho. It results in the lack of that three-dimensional feel many photographers are after. A slight decrease in sharpness across the field, i.e., from foreground to background, creates that 3D feel. Just ask Rembrandt. In that respect, your mountain crop appears sufficiently sharp.
5. Diffraction-induced loss of sharpness can be restored using deconvolution sharpening. Our forum member BartvanderWolf, is an expert in this field, so search for his post related to deconvolution. For instance, this thread: http://www.luminous-landscape.com/forum/index.php?topic=68089.0
-
+1
Erik
A couple of further observations:
1. There was absolutely no need to underexpose, especially not by that much. That results in noisy shadows, further robbing the image of cleanness/crispness. Check the Expose-To-The-Right (ETTR) concept by searching this site.
2. You should remove CA (chromatic aberration) in post-processing, quite visible in your corner crop. That would make the image a bit cleaner/crisper too
3. I would not say that sharpness (or alleged lack of) is a problem in your image. It looks sufficiently sharp to me. Yes, shooting at less diffraction-prone f/stops and using hyperfocal distance focusing would result in a slightly sharper image, but that would only be visible in a direct comparison and in huge enlargements and/or 100% pixel peeping.
4. Absolute sharpness, front to back, is overrated imho. It results in the lack of that three-dimensional feel many photographers are after. A slight decrease in sharpness across the field, i.e., from foreground to background, creates that 3D feel. Just ask Rembrandt. In that respect, your mountain crop appears sufficiently sharp.
5. Diffraction-induced loss of sharpness can be restored using deconvolution sharpening. Our forum member BartvanderWolf, is an expert in this field, so search for his post related to deconvolution. For instance, this thread: http://www.luminous-landscape.com/forum/index.php?topic=68089.0
-
Thanks for the tips!
I think the _main_ problem is that I under exposed this image too much. I just saw the histogram too :(
The next problem is diffraction but that's something I've been trying to understand better. I took some shots with different F stops, and when I took it with F16 compared with F22, the mountain in the far distance will appear less sharp on F16.
-
What should I do with the CA? While I am familiar with ETTR (my own fault for not doing it), diffraction (also my fault for doing otherwise), and hyperfocal distance, I haven't really looked into CA yet.
Should/Can I fix it with Lightroom, Photoshop, or Capture1? If so, how i.e. which tool? I guess I should google :)
-
Hi Jim, nice shot. Regarding the reason for the lack of sharpness you observed, I believe there are two main things which are adding up. First, you have been using a zoom lens which might not be of prime quality - that´s obvious from the red color fringe you see in the corner (already discussed as CA). A prime lens, in particular when stopped down that far, should not have any visible CA. As you know, while zoom lenses have become quite good recently, in many cases they still are not as good as a really good lens with fixed focal length.
Second, you stopped down far beyond the f-stop where the lens provides the best image quality, so diffraction is an additional problem. However, of course, shooting at a larger aperture reduces the depth of field, and I personally am a great fan of images that are tack sharp from the foreground to the horizon. While I´m using a large format camera which allows to virtually stretch the depth of field by tilting the lens, a possible workaround for you would be to do "focus stacking". Following this concept, you are shooting at a favorable f-stop (usually stopping down 2-3 f-stops) and make two or three different exposures, one focussed to the background, one to the foreground and possibly one more focussed to the middleground. Then, the in-focus parts of your exposures are merged in Photoshop. It´s not a lot of work, and this way you simply can squeeze higher image quality out of your equipment.
Best,
Frank
-
Hi Frank,
Thanks for your tips. Can you please show me / point to me this 'red colour fringe in the corner' you're referring to?
-
Actually it´s not in the corner, as I just realized - I´m referring to the first of your 100% crops, where you see it where the grass overlaps the water. Typically you see color fringes along edges with high contrast - that means, when they are there at all, which they shouldn´t.
-
Regarding focus stacking, here´s a link (http://www.cambridgeincolour.com/tutorials/focus-stacking.htm) with some examples. The merging of the different images can be made manually or by a software such as Helicon Focus.
-
For fixing CA, Lightroom has a specific tool for it in the Lens Corrections tab.
As for where it is, I put several arrows to point it out. Usually there is a red component and a green one.
A very good, brief explanation for CA, hyperfocal focusing and difraction can be found here:
http://www.cambridgeincolour.com/tutorials/diffraction-photography.htm
http://www.cambridgeincolour.com/tutorials/lens-corrections.htm
http://www.cambridgeincolour.com/tutorials/hyperfocal-distance.htm
-
... when I took it with F16 compared with F22, the mountain in the far distance will appear less sharp on F16.
Only because you did not utilize hyperfocal focusing.
Check this out (from: http://www.dofmaster.com/dofjs.html). You will see that, had you used f/11, a stop even better than f/16 (for diffraction), you would have had sharpness from 2.91 feet in front of you to infinity! Nothing in your image appears to be that close to you.
-
You've already received some great advice. I'll add to it a bit.
First and foremost, as has been mentioned, nail the hyperfocal point for maximum depth of field. There are charts you can download or you can use an app to do the calculations. Either way, make sure that you use a chart/calculator specific to your camera because the circle of confusion is slightly different for every camera and using the wrong one will give you a false hyperfocal point.
Secondly, most lenses have a sweet spot in the aperture range that runs from about f/8 to f/11 - give or take. The image you posted should only require f/11'ish for maximum DOF if you use the hyperfocal point. When possible, try to stay in the "sweet spot". Not always possible but when it is, you want to use it to avoid diffraction issues.
Third, the 24-105mm is a damn sharp lens. The whole "prime vs. zoom" was an argument back in the day but most pro level zooms these days are just as good as a prime of the same focal length. There are exceptions but the 24-105 isn't one of them. I use it for probably 60-70% of my images.
Lastly, you might consider some tilt/shift lenses. Not cheap, but they're tough to beat when your primary concern is depth of field - whether it be shallow or deep DOF. They're also the way to go for perspective control.
Hope this is helpful!
-
Thanks for the Hyperfocal distance info - I had another quick look again at it, especially this tool http://www.dofmaster.com/dofjs.html My questions:
- When using a zoom lens, e.g. 24-105mm, do I use "105mm" as the focal length, regardless of zoom level? IF I need to use the actual focal length- I can only "guesstimate" the actual focal length based on the marker on the lens ring?
- Then will I also have to guesstimate the distance on the focusing ring since it's not all that accurate / detailed either? It also seems to be on a logarithmic scale rather than linear.
-
I don't care about this hyperfocal distance idea at all for maximizing dof. When I'm not using Tilt-Shift lenses, I abide by Merklinger's thumb rule which, in most instances, works well.
http://jimdoty.com/learn/merkdof01.pdf
-
...I abide by Merklinger's thumb rule which, in most instances, works well...
And that "thumb rule" would be...?
-
You would use the actual focal length.
You are right about the focusing scale. I'm not really sure it's logarithmic but it is not proportional. This is due to physics and no malice of the lens designers.
Best regards
Erik
Thanks for the Hyperfocal distance info - I had another quick look again at it, especially this tool http://www.dofmaster.com/dofjs.html My questions:
- When using a zoom lens, e.g. 24-105mm, do I use "105mm" as the focal length, regardless of zoom level? IF I need to use the actual focal length- I can only "guesstimate" the actual focal length based on the marker on the lens ring?
- Then will I also have to guesstimate the distance on the focusing ring since it's not all that accurate / detailed either? It also seems to be on a logarithmic scale rather than linear.
-
Hi,
Good suggestion from Slobodan, but you need to choose the circle of confusion according to your viewing distance and print size. The 0.03 mm is an old standard based on smallish print viewed at 25 cm. A large print viewed at short distance will be most demanding.
Diffraction can be compensated with proper sharpening, at least to some extent.
I have some write ups on the issue:
http://echophoto.dnsalias.net/ekr/index.php/photoarticles/29-handling-the-dof-trap
http://echophoto.dnsalias.net/ekr/index.php/photoarticles/29-handling-the-dof-trap
http://echophoto.dnsalias.net/ekr/index.php/photoarticles/29-handling-the-dof-trap
Best regards
Erik
Only because you did not utilize hyperfocal focusing.
Check this out (from: http://www.dofmaster.com/dofjs.html). You will see that, had you used f/11, a stop even better than f/16 (for diffraction), you would have had sharpness from 2.91 feet in front of you to infinity! Nothing in your image appears to be that close to you.
-
I went to the same spot again today to do some tests. Luckily the weather really sucked so I didn't have to worry about taking a "photo" but rather just experiment with focusing.
Here's what I did:
- F11 with lens set to 24mm
- According to the calculators, the hyperfocal distance is about 1.9xx metres
I took a series of images without changing the zoom / framing
1. F11 with the focusing ring set to "somewhat" 1.9m - the scale had 1.5m and 3.5m so I put it roughly in the middle (since the scale isn't linear anyway
Result: clearly nothing is in focus - front is very soft and so is the mountain
2. F11 with focusing set to almost infinity
3. F11 with focusing set to beyond infinity (basically all the way out)
4. F11 with my normal way of focusing (just focus on slightly further away to the closest object - which I noticed that the focusing ring is on near infinity - note the grass at the bottom of the image is roughly 10 metres from the camera)
The three methods above produced a similar result - everything is in focus except that it's not razor sharp, but reasonably sharp - this is before applying any sharpening but I think Lightroom applied a 'default' sharpening value of "25" which is applied to all the images including sample (1) above.
So my conclusion is that my way of focusing works for me, however maybe my lens sucks? Maybe it's a bad copy because it has fallen on to the concrete once.
How sharp is "razor sharp" for an image produced by 1ds3 or nikon d3x / d800e etc? Hard to compare different photos / scenes / lighting situation unfortunately. Is it the case that the only way to find out is to get a rental and try and shoot the same stuff with different cameras / lenses?
And here's the dirty question: will MFDB + technical camera (with tilt) totally solve the 'sharpness' issue?
The second problem as was pointed out above, was that my photo above was grossly underexposed so there is a lot of noise in the foreground grass since I had to add "Fill Light" to it.
The third problem as was also pointed out above, F20 does produce a softer image than even F16 or in this case where F11 would be just fine too. I usually take several shots at varying F stops just to cover all bases but in this instance I was too fixated over the clouds to remember doing it :(
-
Good suggestion from Slobodan, but you need to choose the circle of confusion according to your viewing distance and print size. The 0.03 mm is an old standard based on smallish print viewed at 25 cm. A large print viewed at short distance will be most demanding.
Hi,
I agree with Erik, the COC diameter used in that web tool is much (approx. 3x) too large for critical work where larger output may be required. Personally I prefer a tool like VWDOF (http://toothwalker.org/optics/vwdof.html) (a Windows only application), which allows to set one's own COC criterion, and output a number of different scenarios. It also allows simple text output which is easy to use in a spreadsheet or on a smartphone. That's what these DOF tools are most useful for, scenario testing before being confronted with a siuation in the field.
Diffraction can be compensated with proper sharpening, at least to some extent.
That's correct, but how well that works also depends on microdetail contrast (and thus subject matter). It also helps to shoot at low ISO settings to reduce the noise, which will interfere with restoration sharpening. Low contrast detail will already get lost at apertures like f/5.6 - f/7.1 (for sensors with a 4.88 - 6.4 micron sensel pitch), and all microdetail (even high contrast) will be lost at f/22 on most sensors. And when I say lost, I mean without any possibility for recovery, lost. Depending on the lens quality and sensor, f/16 or f/18 are at the limit of being able to resolve fine branches against a bright sky, but lower contrast detail on the ground (soil/sand/twigs, grasses, leaf detail) will be compromised.
So the recommendation is to use an aperture as wide as possible, and as narrow as needed. As for what is needed, there is also a possibility to use focus stacking which requires multiple exposures at different focus distances, and not too much motion (branches swaying in the wind may cause 'ghosting' problems).
Cheers,
Bart
-
The three methods above produced a similar result - everything is in focus except that it's not razor sharp, but reasonably sharp - this is before applying any sharpening but I think Lightroom applied a 'default' sharpening value of "25" which is applied to all the images including sample (1) above.
So my conclusion is that my way of focusing works for me, however maybe my lens sucks? Maybe it's a bad copy because it has fallen on to the concrete once.
How sharp is "razor sharp" for an image produced by 1ds3 or nikon d3x / d800e etc? Hard to compare different photos / scenes / lighting situation unfortunately. Is it the case that the only way to find out is to get a rental and try and shoot the same stuff with different cameras / lenses?
For those who don't shy away from doing some fundamental work, it is possible to find the optimal Capture sharpening settings (http://www.luminous-landscape.com/forum/index.php?topic=68089.msg538932#msg538932) for any lens/aperture/sensor combination. The best results can be had with a proper Radius setting to start with, then the Detail slider setting (try 50 or more if diffraction plays a larger role), and finally the Amount, to push things as far as the quality allows. Masking will reduce the effects of noise in smooth (e.g. sky) regions. Noise reduction also plays an integral part in the sharpening effort.
And here's the dirty question: will MFDB + technical camera (with tilt) totally solve the 'sharpness' issue?
Yes, tilt will solve some of the issues, but not all scenes can be shot equally well. There are still DOF limitations but they become more vertically oriented than depth/distance oriented. See the article from Anders Torger for some more info (http://www.ludd.ltu.se/~torger/photography/focus-landscape.html) on Tilt related issues in landscape photography.
Cheers,
Bart
-
Essentially your technical issues are easy to solve...hyperfocal distance focusing will get you acceptably sharp throughout the image and do a net search for your lens sweet spot (very basically, f8 and stay away from extremes of focal length is generally accepted as sweetest. ive rarely found a need to stray beyond f16 for landscape scenes.
Back to the image....like that, cracking colours and sky....defo worth a trip back there armed with all the tips posted here. If you find yourself not in possession HFD charts/apps etc try focus a third of the way into the frame and review the image at x10, adjust as needed.
Shaun
-
I applied "Lens Correction" on Lightroom which seems to remove most of the reddish CA. The green spots I suspect are actually the real thing captured. It's just the beginning of spring so most of the grass is brown but some have grown green leaves in parts - not sure.
I've also played a bit more with Sharpening and setting "Detail" to 50+ and increased radius a bit, and readjusting sharpening level as much as possible but without causing a defining line between contrasting objects. I've also readjusted the Noise Reduction + increased color noise reduction too which resulted in removal of a lot of reddish blotches in the trees in the foreground.
Result of 100% crop attached, before and after.
-
I just noticed the red dot in the picture - this is a result of the cmos sensor exposing for too long? Clone tool will fix it!
-
Back to the image....like that, cracking colours and sky....defo worth a trip back there armed with all the tips posted here.
Shaun
That was a once in a lifetime sky (aren't they all? :)) - I went back there today - clouds were too low didn't even see any sun light. All dark grey.
-
That's a shame Jim...it's not just me that happens to then....lol. I think there is a lot to be said for having a well established familiar technique when it comes to landscapes...I just spent a week in the mountains my techniques, which I considered second nature needed some concentration after a fitful cold night in a bivybag and shooting at dawn. Find a setup frame expose and shot routine that works for you and work it until you can do it half asleep....as you often will be lol. I look forward to seeing more of your work.
Shaun
-
since you mentioned the lens has been dropped - Roger Cicala at Lensrental.com has a couple interesting articles.
http://www.lensrentals.com/blog/2012/08/a-tilted-element-demonstration
http://www.lensrentals.com/blog/2012/05/testing-for-a-decentered-lens-an-old-technique-gets-a-makeover