Is DOF critical in evaluating corner sharpness of WA lens?

[vietnam photo]

Say I have a wide angle lens which I want to evaluate sharpness at

corners as well as at center. I shot a brick wall and per this URL

http://www.dofmaster.com/dofjs.html, I have plenty of DOF (f8, focal

length 15mm, distance from camera to wall, 3-4 feet. The lens is 15mm

3.5 Nikon).

 

So given that DOF is quite a bit, why is it critical that the camera

must be really parallel with the wall? In my first shot, the camera

was a bit tilted. The righ corners are more blurry than the left

corners. I adjusted the camera and the results are reversed, i.e.,

right corners are sharper than left corners. Thanks.

 

-- Cuong

[vietnam photo]

I did try to have the camera really parallel with the wall but I did not really succeed. Any trick? And why does a "little bit" of tilting make a difference in measuring the sharpness at corners for WA, especially when at F/8? I don't recall having this problem with lens at longer focal length like 85mm. Thanks.

[jbq]

Put a mirror in the center of the test frame.

[vietnam photo]

I am thick today. How does having a mirror in the center of the frame would help?

[User_502260]

This isn't really a good test with a 15mm lens. The depth of field is great but tilting the lens will produce other problems. In the first place you will get distortion. The lines will not be perfectly parallel. The second problem has to do with the angle at which light is going through the lens. Wide angle lenses are sensitive to whether the subject is lined up so that it is either on axis or off axis. Off axis images are rarely as sharp especially in the corners. In any case even a very high quality 15mm lens made for non-macro work in the 35mm format will show much more distortion and less sharpness at the corners than a 50mm or 100m lacro lens when you are shooting a subject like a brick wall.

[Ed_Ingold]

You must have the camera perfectly parallel to the target, and the target must be flat. DOF is, by definition, a compromise. There is theoretically only one plane of sharp focus. With curvature of field, even that statement is questionable. A 15mm lens is not a copy lens. 3-4 feet is too close for a good test, considering the myriad of compromises to design such a lens.

[jbq]

The mirror helps you make sure that the camera is perfectly aligned with the center of the target (you'll see the camera through the mirror) - it eliminates enough degrees of freedom (2) so that the rest of the aligning can be done through the viewfinder.