Originally posted by pakinjapan Thank you for all the suggestion.
I found this
Spherical aberration article on
Wiki, and I just want to confirm if this is the problem you guys talking about on my *55?
I will go back to the store this Saturday for a replacement.
No, probably not spherical aberration. Your lens
appears to fail focus at infinity. Normally one would expect near focus to fade to the distance as the aperture is opened and far focus to stay pretty much the same. Instead, what your examples show is near focus staying about the same and far focus decreasing. In human terms, your lens is myopic, "short sighted". There are a number of possible causes, but since your lens is quite sharp over all, the most likely is that the infinity stop of the focus mechanism was set a few notches short of the appropriate distance for "infinity" when it was made.
Note that I added the word "appears" with emphasis, though in this case I believe the infinity focus issue is real. If your examples had been cropped, I would have had my doubts. What we are seeing in your last example is the far limits of depth of field for your lens at its infinity setting. Normally, resolution is poor enough at maximum aperture that the far limits at "infinity" are no more fuzzy than at "infinity".
I would expect some wrinkling of the nose at this point. After all, infinity is supposed to be amazingly far away and a far limit for DOF would not be defined. That would be true for an ideal lens, but you lens was made in a factory and infinity was set on an optical bench against a target at a virtual distance that approximates infinity as the point of apparent convergence of two parallel lines. By convention a distance of somewhere between 100x and 200x the focal length is used.* Assuming that that 200x is being used, infinity would be set to 110 meters from the camera. At f/1.4 on a K-3 one would expect everything from 55m to forever to appear sharp, assuming no pixel peeping and normal viewing distance. If 100x were used, everything from 34m to ~95m should be acceptably sharp. Given the expected performance of your lens, I would expect 200x.
Steve
* I am going on memory here, so the actual distances might be something else and might be further depending on system resolution, visual acuity of the final observer, size of final image, and so on.