The test was carried out with PMAF (Phase-Matching AF) and CDAF (Contrast-Detection AF) under LiveView mode for both the K-r and the K-x and the amount of Back-Fronting or Front Focusing, if any, is observed. The same test methodology and a similar setup was used, which has been briefed here before. The following illustrations are included for the current test setup and camera settings etc:-
The lens used is a DA 16-45 fixed at 35mm and a f/4 aperture.
The results are tabulated results as follows:-
| Camera / AF Method | Result (Click to Enlarge; EXIF preserved.) | Lens Distance Scale | Remark: |
| K-r / PMAF |  |  | Obvious Back Focusing |
| K-r / CDAF (LiveView) |  |  | Accurate Focus |
| K-x / PMAF |  |  | Slight Back Focusing |
| K-x / CDAF (LiveView) |  |  | Accurate Focus |
As it can be seen from the above, it is easy to note that back focusing is obvious when PMAF is used for the K-r but with CDAF in LV mode, the AF is dead accurate. The difference in the measured distances is quite significant, also. However, the most annoying thing is that whilst K-x do show similar back focusing tendency under the same testing conditions, the order of magnitude and the impact is comparatively less and the final result is considerably more acceptable, as the error is less.
Also, I recall last time when the 43 Limited was used, the K-r front focusing more than the K-x under "yellow" condition. Now the K-r just back focusing more than the K-x in this test with another lens. So?
The safe conclusion that can be drawn again is that the AF errors of the SAFOX IX system is clearly (far) larger than that of the SAFOX VIII, at least if it is to be compared to the one used in the K-x. It seems that any AF error, no matter what it is and what caused it, is magnified with the SAFOX IX! Damn! >:-[
Last Test:-
SAFOX IX Tungsten/Yellow Light Front Focus Issue Fully Investigated