Test Conditions and Procedures:-
1. Mounted an AF lens onto the camera body (of which the battery was fully charged*), forced it to turn to the minimum distance. In this case it is at 0.3m for the FA28 lens in use. (At first I wanted to do this test with the DA35/2.4 but which lacks a focus scale which is not good for this test and thus I used the FA28, which I believe is still a standard lens on APS-C. Nonetheless, I checked and verified that the AF and hunting behaviours of the two lenses were very similar on my K-5.);
*MZ-S was using two LiFePO4 batteries with two hollow dummy conductors in AA.
2. Put the camera in AF-Single mode with only the Central AF point was selected and activated, focused it onto a target which was illuminated by 4000K and 2700K light sources roughly equally. The lighting level was measured to be at about 7.5EV (ISO 100 equivalent). The target object was at about 3 metres away;
3. The same test was repeated for five times (The camera was hand held). The K-5 was tested twice with the new firmware;
4. Recorded the sound and imported the sound file into a sound wave editor and measured precisely for the AF time required, from the AF started to work and then right before AF PCV confirmation tone was heard. All timings are recorded for all the AF achievements. See the following illustrations:-
5. For the last time of the AF, the photo was taken to verify the AF accuracy. (Not applicable for the MZ-S, which is just a film camera!)
The Results:-
Camera | AF Time (1st) | (2nd) | (3rd) | (4th) | (5th) | Average | Sound File |
K-5 FW 1.12 | 0.326 s | 0.400 s | 0.331 s | 0.333 s | 0.607 s | 0.399 s | Download |
K-5 FW 1.13 (1st Test) | 0.548 s | 0.324 s | 0.790 s | 0.407 s | 0.321 s | 0.478 s | Download |
K-5 FW 1.13 (2nd Test) | 0.327 s | 0.333 s | 0.340 s | 0.329 s | 0.406 s | 0.347 s | Download |
MZ-S | 0.316 s | 0.419 s | 0.505 s | 0.632 s | 0.692 s | 0.513 s | Download |
So, my safe conclusion is that the new firmware should make no difference. I have quickly checked the three last photos taken in 100% viewing size that the AF accuracy is not much different, neither. There are some randomness (inconsistency) of the AF performance as you can see from the above results which is not negligible and is actually significant. I guess at this point some would say that I should do the test on a tripod, but I think hand-holding the camera is a real-world case for what I wanted to check in this test. At least, I had tried to hold the camera steadily and pointed the AF sensor onto roughly the same part of the target as far as possible. That's it.
Of course, as usual, I Blog, You Decide! After all, different people will still have different conclusions and opinions even for the same tests, as always! :-)
Read Also:-
K-r AF Accuracy Comparison: FW V1.10 Vs 1.01, under Yellow and White Light Sources
My Pentax K100D Full Review