Sharp pix?

[qingtian]

qingtian,
I agree with some guys comments on diffraction. Shooting at min aperture doesn't mean your pics will turn out sharp. The word "sharp" is argueable bcoz some ppl like it VERY sharp,some ppl just like it sharp so long as u can see the building. Some ppl are addicted to sharpen their pics so much that the human's face (after processing) is overkill. There's a prob call atmospheric haze which u can never get it off unless u shoot in very dry condition like in Europe,Northern Japan,Russia,etc. Btw 17-40 (in my opinion) is not really a sharp lens. Sharp lens is best to be fixed focal length eg. lens by Leica (ASPH) & Zeiss (Distagon series). Try shooting fr the same scene with 16-35mm L Mark II. All the best.

Agreed on the too much of sharping on human face...
Most of my cityscape and landscape shots which I took, I will do some basic sharping.
However I do not have any habit to sharpen human/portrait shots as it indeed look a bit un-natural to me.

Hmmm... only one question here leh, where got '2 questions'?:bsmilie:
Anyway, to my humble knowledge, sharpness has everything to do with the lens setting. Not with the camera body setting.
The only setting remotely linked to your camera is that built-in 'Sharpness' parameter in 'Picture Style' of our Canon camera. But as a general rule, don't touch that one first until you have a better grasp of the principles of lenses.

Lol, there is 2, the 2nd questions is I just want to be sure that problems lies with the human (me) on the camera setting instead of a lemon 7D body that everyone else is discussing about...

Understood on the picture style thingy!

 

[qingtian]

Also, we need to know how you resampled your image in the first place and where you uploaded it for hosting. Your resampling method may have thrown out quite a bit of data and your host might not be optimal for photo hosting (for example, what looks good in Flickr will look crappy in Facebook or Photobucket).

Thanks for you inputs, I do know bout this as I do notice the diff in IQ between flickr and facebook. Thanks for sharing!

 

[Kit]

Hi Guyz,

Need some help over here...

I was wondering are the 2 pics that I took below consider sharp:-
As the more I look at it, it dun look sharp to me?

Pics are taken using EOS 7D _ 17-40mm L on a tripod.
Appreciate for your attentions!

Noticed that the photo shot later in the day is significantly "sharper" than the other one? Something that many people failed to realise is that quality of the ambient lighting does affect our perception of sharpness.

The ambient lighting in the first photo is rather flat and uncontrasty and so youor buildings will appear that way, especially for the buildings in lighter shades e.g. UOB, OCBC, etc. The photo shot later has better ambient light which is further enhanced by the buildings' night lights.

If you are using the same equipment and same settings, sharpnness level should not differ. I can't emphasize enough of this. Watch that light.

 

[shaoken]

Hello, qingtian.
I checked the EXIF of your photos.
Photo number one is at f/22 at shutter speed of 2.0 seconds.
Photo number two is at f/18 and at shutter speed of 30 seconds.

In my opinion, there are two possibilities:

1. Diffraction. At narrow apertures of f/18 and f/22, diffraction would cause you to lose sharpness.
2. The second photo at shutter speed of 30 second would be more susceptible to shakes caused by wind blow and such, even on a tripod.
My opinion.

In daylight, why shoot at f/22? :dunno:

For the night shots, I do have experience with very long exposure.
Yes, when it is too slow shutter speed, shakes may occur.

 

[Kit]

In daylight, why shoot at f/22? :dunno:

For the night shots, I do have experience with very long exposure.
Yes, when it is too slow shutter speed, shakes may occur.

If the tripod is correctly set up and firmly planted to the ground, longer shutter speeds are not a problem.

 

[Kit]

I haven't really tried taking cityscapes at f/16 or beyond(don't see the need to) but I would think that diffraction isn't going to be that significant expecially for a scene.

 

[Jed]

Actually, while the negative effects of diffraction are correct, I'm not sure they are to blame here. We're looking at a very small version of the overall image and I'm not sure diffraction has that big an impact on image quality that you could still get a perfectly sharp 1000 pixel wide image. If you were talking about softness on a pixel level then that's a different story.

As Kit says, local contrast makes a big difference in "apparent sharpness", and if you apply intelligent USM to your images (which effectively is bumping your local edge contrast) then the shot starts to look a lot better.

And just for the record, diffraction starts to get annoying from f11 onwards on a crop sensor.

And also, your second shot is overexposed by my reckoning, just as an observation.

 

[Kit]

Actually, while the negative effects of diffraction are correct, I'm not sure they are to blame here. We're looking at a very small version of the overall image and I'm not sure diffraction has that big an impact on image quality that you could still get a perfectly sharp 1000 pixel wide image. If you were talking about softness on a pixel level then that's a different story.

As Kit says, local contrast makes a big difference in "apparent sharpness", and if you apply intelligent USM to your images (which effectively is bumping your local edge contrast) then the shot starts to look a lot better.

And just for the record, diffraction starts to get annoying from f11 onwards on a crop sensor.

And also, your second shot is overexposed by my reckoning, just as an observation.

Yeah, I didn't think that diffraction(if any) would be detectable from the samples.

 

[Kit]

qingtian,
I agree with some guys comments on diffraction. Shooting at min aperture doesn't mean your pics will turn out sharp. The word "sharp" is argueable bcoz some ppl like it VERY sharp,some ppl just like it sharp so long as u can see the building. Some ppl are addicted to sharpen their pics so much that the human's face (after processing) is overkill. There's a prob call atmospheric haze which u can never get it off unless u shoot in very dry condition like in Europe,Northern Japan,Russia,etc. Btw 17-40 (in my opinion) is not really a sharp lens. Sharp lens is best to be fixed focal length eg. lens by Leica (ASPH) & Zeiss (Distagon series). Try shooting fr the same scene with 16-35mm L Mark II. All the best.

The 17-40 is more than capable to handle this kind of photo and actually, for the price they are asking for, the 16-35 MKII is not that much better. Definitely better in curvilinear distortion control though.

 

[limwhow]

Yeah, I didn't think that diffraction(if any) would be detectable from the samples.

Actually, while the negative effects of diffraction are correct, I'm not sure they are to blame here. We're looking at a very small version of the overall image and I'm not sure diffraction has that big an impact on image quality that you could still get a perfectly sharp 1000 pixel wide image. If you were talking about softness on a pixel level then that's a different story.

As Kit says, local contrast makes a big difference in "apparent sharpness", and if you apply intelligent USM to your images (which effectively is bumping your local edge contrast) then the shot starts to look a lot better.

And just for the record, diffraction starts to get annoying from f11 onwards on a crop sensor.

And also, your second shot is overexposed by my reckoning, just as an observation.

You experienced photogs have brought forth a new angle to my thought process here, and have put into detail description a notion that I have all along, but only in a vague way.
Thank you very much.
So allow me to paraphrase it for my own understanding: The amount of ambient light (and the amount of exposure, I presume) would affect local contrast to such an extent that a deficiency in this light would result in a perceived loss of sharpness.

Therefore in the first photo, a slightly better exposed picture would improve the perceived sharpness.
What about the second photos? As it has been well (over) exposed, then why is there still a perceived softness about it? Or do both of you feel that the second photo is not soft?

Thanks for your advice.

 

[PrimePhotog]

Photo number one is at f/22 at shutter speed of 2.0 seconds.
Photo number two is at f/18 and at shutter speed of 30 seconds.

Both images should be pathetically soft since they were shot at apertures much smaller than f10. The 7D has a low DLA due to it's high pixel density.

 

[limwhow]

Both images should be pathetically soft since they were shot at apertures much smaller than f10. The 7D has a low DLA due to it's high pixel density.

Sorry, I tried checking up Google and the Search in CS but can't find the actual meaning of DLA. Can you explain what DLA means?

 

[qingtian]

Noticed that the photo shot later in the day is significantly "sharper" than the other one? Something that many people failed to realise is that quality of the ambient lighting does affect our perception of sharpness.

The ambient lighting in the first photo is rather flat and uncontrasty and so youor buildings will appear that way, especially for the buildings in lighter shades e.g. UOB, OCBC, etc. The photo shot later has better ambient light which is further enhanced by the buildings' night lights.

If you are using the same equipment and same settings, sharpnness level should not differ. I can't emphasize enough of this. Watch that light.

I will take note of this.

In daylight, why shoot at f/22? :dunno:

For the night shots, I do have experience with very long exposure.
Yes, when it is too slow shutter speed, shakes may occur.

I got a mis-concepts that I should took any shot at the higher f-numbers in any way that I could inorder to get the best in-focus in the images. I guess I'm wrong and I am glad that the bros (Not sure is there any sis in my posting) clarify with me on that. Now I know city/landscape shooting should be stick between f/8 to f/16. And I will need to take note that f/11 and above will result in diffraction on a crop sensor as per Jed advised.

Actually, while the negative effects of diffraction are correct, I'm not sure they are to blame here. We're looking at a very small version of the overall image and I'm not sure diffraction has that big an impact on image quality that you could still get a perfectly sharp 1000 pixel wide image. If you were talking about softness on a pixel level then that's a different story.

As Kit says, local contrast makes a big difference in "apparent sharpness", and if you apply intelligent USM to your images (which effectively is bumping your local edge contrast) then the shot starts to look a lot better.

And just for the record, diffraction starts to get annoying from f11 onwards on a crop sensor.

And also, your second shot is overexposed by my reckoning, just as an observation.

Per your statement I guess "effectively is bumping your local edge contrast" am I right to say that is part of PP? Could I also check with you should I shoot at a lower megapixels as per what 7D offer would I encoutered such kind of softness issues as I dun really need a large resolution images.

 

[limwhow]

Qingtian, personally I think they were not referring to PP on the local contrast issue.
Anyway, Kit and Jed have raised a very valid point and I am very happy that they have given their professional views on this.
At the same time, I am awaiting to see what DLA means.
You know about DLA, qingtian?

 

[Numnumball]

Clearly Diffraction, furthermore subjects are too far away to utlise hyperfocusing..
Talk of Diffraction limited aperture (DLA). Normally the "sweet spot" is between f5.6 to f11 reducing Seidel aberations (mostly Spherical abberation) and increasing diffraction. And setting at f16 and above won't give u the best results as it did.

That basic rule holds just as well for APS-C as it does for full frame since they aren't really all that different in size..But with smaller sensors (due to its higher pixel density) diffraction kicks in easier..

 

[Rashkae]

That basic rule holds just as well for APS-C as it does for full frame since they aren't really all that different in size..But with smaller sensors (due to its higher pixel density) diffraction kicks in easier..

Bingo.

So don't go above f/11 on a high-density sensor, and consider running a sharpening filter *after* you resize to a smaller size.

If you are using photoshop, you can try to resize with bicubic (sharper) and you'll notice a big difference.

 

[limwhow]

Clearly Diffraction, furthermore subjects are too far away to utlise hyperfocusing..
Talk of Diffraction limited aperture (DLA). Normally the "sweet spot" is between f5.6 to f11 reducing Seidel aberations (mostly Spherical abberation) and increasing diffraction. And setting at f16 and above won't give u the best results as it did.

That basic rule holds just as well for APS-C as it does for full frame since they aren't really all that different in size..But with smaller sensors (due to its higher pixel density) diffraction kicks in easier..

Thank you Numnumball for the explanation on Diffraction Limited Aperture. So you personally thinks diffraction plays a role in the softness of the photos here...

Bingo.

So don't go above f/11 on a high-density sensor, and consider running a sharpening filter *after* you resize to a smaller size.

If you are using photoshop, you can try to resize with bicubic (sharper) and you'll notice a big difference.

Thanks, Rashkae bro. Always can depend on you to give good pointers, especially here on sharpening and resizing.

 

[Numnumball]

Bingo.

So don't go above f/11 on a high-density sensor, and consider running a sharpening filter *after* you resize to a smaller size.

If you are using photoshop, you can try to resize with bicubic (sharper) and you'll notice a big difference.

Thank you Numnumball for the explanation on Diffraction Limited Aperture. So you personally thinks diffraction plays a role in the softness of the photos here...

Exactly.

If you want to keep your images sharp, don't stop down that much with an APS-C DSLR. The effects of diffraction are clearly visible at f22 and above and it significantly degrade the image. Actually optimal sharpness depends on the lens you are using. For a lens with significant aberrations may give optimum results. For a lens with less aberrations optimum performance is maybe around f11, though both f8 and f16 are very similar. For a really good lens like the AFS 17-35mm f2.8 (which is the one i am using now) with well corrected aberrations, performance may peak at f5.6 and f8 and great at f11. f16 is acceptable, but f22 and smaller apertures should be avoided like i what i normally do. (Understand ur lens and its optimal aperture.)

But imho, the smoothness of the photos be corrected easily in post processing if it is caused by diffraction. This is a small issue but if you want to get the maximum sharpness into your photos, diffraction is something to consider.

Just my 2 cents worth

 

[theveed]

Many folks think that stopping down to f/16 or smaller is required for DOF concerns, but discounting any calculations in terms of actual DOF requirement and computation for hyperfocal distances based on the focal length used and the distance of the camera to the subject(s) as well.

For example, using this http://www.dofmaster.com/dofjs.html calculator with a subject 30ft away.

With a 50mm lens at f/11 on a crop, you'll keep things in focus from 17ft to 137ft.
With an 18mm lens at f/4 on a crop, you'll keep things in focus from 9.5ft all the way to infinity. Switching to f/11 will only improve near-focus coverage...

With a scene like yours, there isn't anything near your camera's position anyway to need that small of an aperture.

 

[Jed]

So allow me to paraphrase it for my own understanding: The amount of ambient light (and the amount of exposure, I presume) would affect local contrast to such an extent that a deficiency in this light would result in a perceived loss of sharpness.

Not quite the "amount" of ambient light but the type and quality of it. So a sunny day for example always results in more contrasty images that could give the perception of greater sharpness. If you should a portrait in a studio for example then a very softly lit shot of a subject would appear "less sharp" than a harshly lit subject.

What about the second photos? As it has been well (over) exposed, then why is there still a perceived softness about it? Or do both of you feel that the second photo is not soft?

Personally I feel the second photo looks not dissimilar to the first, but the overexposure makes looking at it accurately quite difficult.

I've done a little bit of editing to hopefully illustrate what I meant originally.

I've sharpened the original image through two stages, a single low level general contrast enhancement, as well as the normal USM as you would do for publication. To me the second looks fine and a lot better already.

I still stand by the fact that while diffraction does impact your quality, it shouldn't stop you from being able to get a good image at 1/26th the size of the original. It's obviously however very easy to jump on the diffraction bandwagon.