Background
Since K-r and K-5 have been marketed, many Pentax users have been complaining about the obvious Front Focus (FF) issue resulted when using these cameras, when they were shooting at lower light and more so under tungsten light source, actually. A recent report here.
In fact, this "Yellow Light FF Symptom" was with Pentax users for long since Pentax made their first DSLR, the *ist D. But I did find that the problem was not as severe as before since Pentax had made the K-m and with the two successive models of the K-7 and K-x, too. But rather unfortunately, with the K-r, it seems that this issue strikes again and more of those adverse reports arose, which was also confirmed shortly after a few days use of my new K-r! :-(
So, I decided to do yet another controlled experiment, to compare the old SAFOX VII of my MZ-S, the SAFOX VIII (latest version) of my K-x and the SAFOX IX (first version) of the K-r:-
Test Methodology
1. The focus target is as below. It comprises of a stair case of 3 flat surfaces. The centre one is the main target, which has a good contrasty pattern, and is used for the AF, with the central single focus point of the cameras, in AF-S mode.
2. Each of the camera body is then mounted on a sturdy tripod. The light source is changed between Yellow and White, photos are taken and then where the final focus arrives in the final image per picture is observed. The lens focus/distance scale for the final focus measure is also recorded, for each picture.
3. The shutter release button is re-pressed for a few times until there is no re-focusing/hunting and that the camera believes firmly that its decided AF focus is correct.
4. Other testing notes and conditions: FA 43/1.9 Limited in Av mode at f/2.0, ISO 200, SR off, AF Single, Central AF point, AWB, latest firmware in K-r and last latest firmware in K-x.
The Results
Observations and Conclusions
1. Both K-x and K-r (*without* focus adjust) are fairly accurate with white light source.
2. Under yellow light source, the K-x focus shifts forward and too with the K-r. But the span of the shift of the K-x is considerably less than that of the K-r. As a result, the K-x' picture is still somehow acceptable even it is front focused. On the other hand, the K-r picture under yellow light is totally out of focus for the target owing to the serious front focusing. This is verified from the lens' focus scale readings in the above the four respective pictures, too.
3. With focus adjust at -10 for the K-r, now it focuses more correctly under yellow light, but then it then causes obvious back focusing in white light. Again, the span between white and yellow light is large, although it is not as large as when the focus adjust is off.
4. The MZ-S / SAFOX VII is affected least by the yellow light with minimal effect. By looking at the focus scale, it matches well with the best focus pictures under white light for the K-x and K-r.
Afterall, how does that ten years old SAFOX VII still outperform both the last version of SAFOX VIII and now the latest SAFOX of IX (but which is now the worst!)? How come Pentax yet once again has sucked quality from the SAFOX IX, which they had already improved it to avoid the problem somehow with the latest revision of the SAFOX VIII?? :-o
And since I purchased the K-r to upgrade but not downgrade, its more serious yellow light FF symptom than the K-x really upset me. And this is for sure a big disappointment - I did not expect the K-r could be and should be worse than the K-x! >:-[
Remedials, then?
1. Use the LiveView to focus, the LV CDAF is not affected by the yellow light, last time I checked and verified. But with LV, of course it has its problem and limitations, such as no OVF could be used and the prolonged shutter lag (the shutter will be fired twice before a picture is taken).
2. Manually change the focus adjust setting each time for Yellow or White light source (do remember to set the memory function for the last accessed menu item to avoid more troubles if you do this! And do remember to check frequently to avoid "user error"!!).
But of course, the pre-requisite for you for doing this is that you must first have an accurate K-r in AF which requires no -ve focus adjust by default in the very beginning, just like mine. Otherwise, you will never be able to enter the required amount of (the -ve) correction if you have already used up some of the quota, just in case - which could happen with the "excellent" QC of Pentax nowadays.
Read Also:-
How AF Errors Can Decrease *Effective* Resolution?