http://www.pentaxforums.com/forums/pentax-k-5-forum/136324-k-5-1-03-versus-k20d-auto-focus-tests.html
The results are self-explanatory. Testing conditions including lens in used and the lighting level is stated clearly below each pair of the comparison test pictures.
It can be seen that the old K20D K.O. the K-5 with the latest "improved low light AF accuracy" firmware 1.03 hands down! :-o
It seems that Pentax has gone back to the old *ist D days for poor AF accuracy with their latest SAFOX IX/+ used in the K-r and K-5, although in speed the new 9th generation of the Pentax system seems to be faster! >:-(
Nevertheless, a few other K-5 users have carried out some random ad-hoc checks on the AF accuracy of their K-5 units at lower light. It seems that some improvements over the old firmwares are shown but I can still see some OOF samples:-
http://forums.dpreview.com/forums/readflat.asp?forum=1036&thread=37937710
So, the quick conclusion is that the new K-5 firmware could improve something but if it is compared to some old Pentax DSLRs like the K20D, the AF accuracy of the K-5 still, sucks.
My personal experience is that my K-m has the best AF accuracy, which has a mature SAFOX VIII. In fact, I found that the AF accuracy of the K-x had deteriorated but just have never imagined that the K-r could be even worse! Too bad!
Related:-
"Hysteria"!
Improved AF Accuracy of the K20D
AF Speed Difference of K-m Vs K20D
K-x' Time Longevity on AF Accuracy
K-r's AF Accuracy Affected Much by Colour Temparture of the Light Source
Magnified AF Errors of the SAFOX IX than the SAFOX VIII
Logging you in...
Corros · 730 weeks ago
RiceHigh 110p · 730 weeks ago
Michael A. · 730 weeks ago
RiceHigh 110p · 730 weeks ago
Mr Bigg · 730 weeks ago
He admitted it and others have pointed it out.
But that didn't stop you from carefully selecting and posting the information dishonestly did it?
And don't try and tell me you didn't follow the entire post either.
We both know you visit the Pentax Forums several times a day(through multiple accounts) and have been actively following this thread.
Now hurry-up and delete this post before others see how sick you are.
RiceHigh 110p · 730 weeks ago
josh05 · 730 weeks ago
K-x has the "mature" 11-point SAFOX VIII, K-m has an unnumbered 5-point SAFOX System acording to official data sheets.
RiceHigh 110p · 730 weeks ago
"8. High-precision 5-point AF system
The PENTAX K-m incorporates a five-point wide-area AF system (SAFOX VIII with five cross-type sensors)"
josh05 · 730 weeks ago
(I couldn't find any official pentax data sheet for the k-m/k-2000, where its AF as named as SAFOX VIII)
RiceHigh 110p · 730 weeks ago
Btw, the only Pentax SLRs that have the original SAFOX AF system are the SF series cameras.
Toomas Kadarpik · 730 weeks ago
In the test we see handshots and non controller environment, this test is from the testing point of view not accurate. To get anything out from testing I suggest metrology a discipline I hated in physics curriculum a lot but it has an important point to get things done. http://en.wikipedia.org/wiki/Metrology
Falk Lumo has science educations and he can cope with realistic labs, lets wait his results and then we can talk. Rice you are just making noise here.
Guest · 730 weeks ago
rvannatta · 719 weeks ago
I have no idea how to judge its 'accuracy' if you set it to a multi point focus. when you have a whole bunch
of points and different distances (beyond the depth of field). I have used the multi-point focus mode and for the record I don't see evidence that it is doing anything really stupid, but it can't fix depth of field that is less than the depth of the depth of the subject matter.