1
\$\begingroup\$

I'm using a Sony A7R IV camera with a 12mm f/2.8 ED AS NCS Fisheye Lens. From the rescharts and MTF information that I've seen of this lens, the center sharpness is much better than near the edges. However, when I performed a similar sharpness analysis using the Imatest software my results showed that the MTF50 values near the image edges were higher than near the center. See attached photos; original image is found here.

This doesn't make sense to me, since I would intuitively expect that the center of the image, where the barrel distortion is the smallest, would be sharper than the edges of the image. In trying to work through this problem, I've narrowed it to some possible reasons:

  1. My interpretation of the MTF50 measure is wrong and so I have misinterpreted the plot

  2. The way that I focused the camera did not correctly focus on the center. For some context, I used the Focus Assistant of the Sony camera set to High to assist with manual focusing (maximizing the number of red lines that appeared)

  3. Uneven lighting effects are causing the edges to appear sharper

Any help or suggestions would be greatly appreciated. I'm really new to photography so I'm not sure if this was the right place to be posting my question and if how I maybe set the camera settings was correct. Thanks!

As a side note, I don't have the camera settings I used on hand, but I can try to find that information if needed.

3D contour plot of MTF50 values in Imatest software

EDIT: I retook some photos with a test chart here and here, keeping in mind the suggestions about tilt. Visually, I think that the top and bottom seem to be improved, but the right and left sides seem a bit off. When I ran my Imatest module on some slanted edges on the left and right, they both still had a slight improvement over an edge analyzed in the center. Any suggestions for how I could reduce or eliminate this sort of problem?

\$\endgroup\$
5
  • \$\begingroup\$ How are you aligning the camera with the test chart? \$\endgroup\$
    – Michael C
    Commented Jun 2, 2022 at 17:16
  • \$\begingroup\$ I don't do MS Office so I cant access you updated images. \$\endgroup\$
    – Michael C
    Commented Jun 2, 2022 at 17:16
  • \$\begingroup\$ I've updated the image links to be on google drive; hopefully that works better. In terms of alignment, I have the camera sitting on a tripod and I'm trying to align it with a straight rod that is perpendicular to the board. It's not the most rigorous approach \$\endgroup\$
    – Jason
    Commented Jun 2, 2022 at 17:43
  • \$\begingroup\$ Please see this answer to a related question. You may want to also check out this article from lensrentals.com's chief lens guru Roger Cicala. \$\endgroup\$
    – Michael C
    Commented Jun 2, 2022 at 19:18
  • \$\begingroup\$ Thank you for your help and resources! \$\endgroup\$
    – Jason
    Commented Jun 2, 2022 at 20:20

1 Answer 1

3
\$\begingroup\$

It looks like you're seeing the effect of tilt.

There are two main possibilities that can cause tilt:

  • Your camera and test chart are not aligned properly. If the camera's sensor plane is not perfectly parallel to the perfectly flat test chart, then one side or the other or, as in your case, the top or the bottom of the test chart will be closer/further from the camera than the opposite side. Whichever side is closest to the lens' focus distance will be most in focus.
  • Your lens is not aligned properly. This could be due to a misaligned lens element, to the flange ring on the lens not perfectly perpendicular to the lens' optical axis, or the camera's mount ring not perfectly parallel with the sensor. Any flange/mount misalignment will be more noticeable with very wide angle lenses than with longer focal length lenses.

Just based on the difference in curvature between the top and bottom of your test chart, as well as the difference in curvature between the left and right edges of the chart, I'd say that your main problem is one of camera/chart misalignment. It appears that your camera is closer to the upper left corner of the test chart and further from the lower right corner. It also appears that the lens was focused at the longer distance of the lower right corner.

\$\endgroup\$
3
  • \$\begingroup\$ Agreed... I'm not seeing "sharper edges," I'm seeing tilt/DOF. \$\endgroup\$ Commented Jun 1, 2022 at 12:58
  • \$\begingroup\$ Thanks for your reply! It was really helpful and definitely made sense. If you don't mind clarifying, how significantly can tilt impact the quality of an image? And does shooting at a longer distance away help to reduce the effects of tilt? \$\endgroup\$
    – Jason
    Commented Jun 2, 2022 at 5:26
  • \$\begingroup\$ Significance? That all depends upon what is causing it. If it is due to improper camera/test chart alignment it will have zero effect on real world image quality (but similar failures in shooting technique could). Distance? If it's due to a lens alignment issue it could be worse or less severe, depending on the exact nature of the misalignment and in which direction the misalignment is. But until you nail down proper camera/test chart alignment your results are telling you nothing about the lens' condition and everything about your (lack of) proper testing setup. \$\endgroup\$
    – Michael C
    Commented Jun 2, 2022 at 17:15

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.