Originally posted by Shanti Is K3 AF more accurate than K5II
AF is accurate if the image content falling into the AF center point sensor area has enough horizontal or vertical detail/contrast to focus on.
K-3 has a smaller center AF area (relative to K-5) so it may help or not, depending on what you focus on.
With K-5 and K-3, I shot birds in flight together with my friend having a 7D (7D is famous for its excellent AF...) but 7D also misses some shots simply because it can't AF on flat focus target either... so, phase detect AF always have limitations. It is simply not true that AF can 100% succeed, regardless of the camera used. With the DA300, while you're focusing, the image covered by the center AF sensor is moving (unless on tripod mounted), which make it harder for the AF for lock. I'd say that the advantage of the K-3 is that you can set the amount of AF hold in AF-C mode, this feature does not exist in the K-5. I don't know is AF-C hold can be customized on the K-5ii . Also, with the K-3, I've found out that 9 center points AF works best since it can be pretty hard to have the AF center point remain on target while tracking birds.
Practically, the difference that I experience [with my copies of DA300, K-5, K-3] is that K-5 won't lock even if the image in 99% in focus, while the K-3 is able to lock. With the K-5 that I have , I consider that the accuracy constraint is too restrictive (because you can't take the shot if not 100% in focus, even when it is good enough to be usable). The K-3 will lock and take the shot if AF is "good enough". Also, at 200mm (Tamron 70-200), I get more keepers than with the DA300. And with DA300 + x1.4TC it's even harder...any slight movement of the camera+lens @ 630mm equ. displace the AF point while AF process is running.
In short, I don't think the K-3 AF is more accurate , I think that K-3 AF is better designed for dynamic shots simply because it has more AF points.