Sigh. Not another AF discussion. And as is customary, everyone is jumping straight to conclusions, based on a series of casual snapshots, and presumptions about Pentax vs Canon vs Nikon etc.
K-5 AF is rated EV-1 to EV 18 at ISO 100, so those scenes of OP are well within the illumination range for Pentax AF to work properly. (Especially when flash is involved, which it is in these sample images. So we aren't really talking about low-light AF here anyhow...)
Originally posted by sanjeevdas Sometimes I will change the focus point to where I want the focus to be. In most of these I pointed at the face and recomposed.
If Pentax had face detection in it's phase-matching AF (via the view finder, rather than via liveview, where there is the option of face-detect) that might work. Otherwise the AF points of any camera are not laser targetting dots, but just regions of differential contrast in the area indicated by the focus points.
I also note that in all of the sample images where you report problems, flash is involved, and low shutter speeds too. Use of flash means what you are reporting is not really a low-light AF issue at all, but a flash settings + technique + AF issue.
I suggest that you test any purely AF accuracy speculations by doing your tests properly against static subjects.
At a minimum for any tests:
- choose a flat subject/ test target with *one* strong area of vertical or horizontal contrast right in the centre,
- set AF to centre-spot,
- set ISO to 100
- set AF to AFS
- set lens aperture wide open
- set the target at a medium distance (ie more than 2 m),
- put camera on a tripod,
- turn off SR,
- take any UV or other filters off the lens
- provide good even illumination of test target for initial tests, then vary light strength and as required light temperature (eg tungsten, LED, daylight, flash, whatever) if that's what you want to test.
Then draw some conclusions.