How sharp is 'sharp'?

[Melee]

Hello CS gurus out there,

Sorry if this topic has been discussed to death, but I hope you'll indulge me... =P Already did a search but couldn't find what I was looking for...

I'm trying to evaluate the sharpness of my new Sigma 10-20mm f/4-5.6 EX DC HSM. I tried it at the shop and it seemed OK, but testing in the shop was sub-optimal: poor lighting + small camera LCD. I've now shot some test pics, but I'm still not sure how sharp is 'sharp'. My concern is that the edges (esp the top and bottom) always seem to be sharper than the centre of the image! Please help me to evaluate... would like to know if the lens is 'ok', or if there might be a QC issue, or even a back-focusing problem.

Thanks in advance!

All pics shot on a Nikon D80 at f/8, ISO100. In-camera sharpening: Medium High, tone compensation: Normal.

10mm, full FOV, scaled down

Centre 100% crop:

Left 100% crop:

Right 100% crop:

Bottom 100% crop:

 

[Melee]

More pics, this time at 20mm

Full FOV, resized

Centre 100% crop:

Left 100% crop:

Right 100% crop:

Bottom 100% crop:

Thanks in advance for any comments/advice!! =)

 

[Andy Ang]

WHere did you focus in the first place?

 

[Melee]

Oops, sorry, focus point was at the centre of the image, just below the word "PRIMARY". And the 100% crops were from the 10mp images.

 

[scenar]

crops look perfectly fine.
Go out and enjoy your cam!

Ref your sig: your 10% is settled!

 

[Kit]

How many times do you print 100% crops of your photos at A3 size? None? Why don't you try printing the actual image on A3 and see for yourself..... clears your doubts once and for all.

 

[morningstar]

Pixel peeping!

 

[Umeiko]

I think your image looks sharp.

I also experiencing some uncertainty on the sharpness of my 12-24. Maybe I can go to the same place to take and compare.

Was the picture taken at Simei?

 

[wong1979]

Looks fine to me. As you are using a super-wide angle, things will look less sharp at 100% crops when compared to say a 50mm crop at 20%. Similarly, a 50mm 100% crop will look less sharp than a 300mm 20% crop. Makes sense?

BTW, u stay at 101?

 

[Clockunder]

In your first picture and its various crops, the centre is indeed not as sharp as the sides and this is unusual because the centre of the lens is supposed to be sharper and has higher resolution power than the edges/sides. However, in this case, it could be due to the "Primary" building being further distance than the buildings by the side and road in front. So the lower sharpness of the centre could be more because of the lower resolution of a more distance object. When an extremely short focal length(10mm) is used, the distances become a lot more times the focal length and the (de)magnification of the actual size into image size is very huge and they appear very small in the frame. (demagnification depends on the distance divided by the focal length). So the centre which is much further away (in terms of times of focal length) than the other comparison areas appears not as sharp due to much lower resolution of details.

In the 2nd picture and its crops, the areas selected for comparision are now different from those in the 1st picture and appear to be more or less equally sharp. This could be because, firstly, now the difference in distances of the buildings used for comparison are lesser than in the first picture. 2ndly, a longer focal length (20mm) is used and therefore the (de)magnification factor is lower than the 1st picture and they are bigger within the frame. Taken together, these 2 factors means the resolution of details is better and so, the areas selected for comparison appear to be sharper and more or less equally sharp.

To test whether the lens have a problem with centre sharpness, you probably need to shoot a picture in which the frame is filled with various objects about equal/(not so much differences in) distances from you (That's what we usually see in studio test shots by review websites).

 

[zoossh]

could it be pincushion effect?

anyone i think you need not worry too much. other photographic components will outweigh that problem, giving you satisfactory pictures.

 

[Cap_Dingo]

If I'm not wrong, your subject (school) is 60-70 metres away. Quite a distance if you wanted to see if something was tack sharp? Night shots and contrast there may also define the subject better?

(also newbie living in Simei)

 

[m00nman26]

Bro,
To test an ultra wide lens for focusing, it would be much more helpful if you do close ups to see the difference. at 10-20mm, at F/4, even if you focus wrongly at long distances, the dof is so big that you can't tell the difference in back focusing.

courtesy of online dof calculator
at 50m away, 10mm F/4,

Depth of field
Near limit 1.22 m
Far limit Infinity
Total Infinite

 

[milamber]

Hmm... Shouldn't you turn off your in-camera sharpening if you want to test the sharpness of your lens? Best to shoot in RAW and i assume you used a tripod rite?

 

[Melee]

Thank you EVERYONE -- I've learned plenty plenty from all your replies! :vhappy:

I had a Canon G3 previously, and only recently switched to a dSLR. So I'm still getting used to the new camera and the pictures it takes (4mp to 10mp is quite a big jump...) The main reason I started this thread was so that I could get a better idea of what to look for when evaluating sharpness -- i.e., how sharp is 'sharp'. Thanks to all of you for being extra eyes and brains for this newbie!

Thanks all also for your very helpful thoughts on how to evaluate sharpness, issues with object distance, dof, etc. I think I have a better idea of how to do it next time... Wah, so much to learn...

ClubSnap ROCKS! :cheergal:

p.s. wahrao, alot of people here quite familiar with Simei siah... :sweatsm:

 

[xunjas]

Sharp images - http://forums.clubsnap.org/showthread.php?t=232914&page=3

sharp like those images in that link izit?