Like a few others with the K10D and *16-50 combination,I have had a few problems with the AF in the 16mm-20mm,and f2.8 sometimes f4 areas when focussing at infinity. I decided to determine where the points of sharpest focus lay for a particular object at ∞ (about 70/80m) for the range of focal lengths.I used manual focus and by trial and error discovered that at 16mm(and all f-stops ) sharpest focus lay with the cursor on the lens barrel just to the right of the lemniscate;at 20mm and 28 mm the cursor pointed to the middle;at 35 and 50 the cursor pointed to the middle of the second oval on the loop of the lemniscate.
On testing the AF with the different focal lengths and f-stops I found it to be accurate in that it adjusted the lens barrel to the same points I had discovered manually except in the cases of 16mm and 20mm and at f2.8 and f4,where the lens barrel seemed to have travelled too far.
This confirmed the problem that I'd already experienced so the solutions seemed to be:
a.return the camera for adjustment.
b.manually focus in the areas indicated as 'iffy'.
c.focus at 5.6 or above and then stop back up when focus is obtained
But this led me to a conundrum. If AF is carried out with the lens fully open,why should stopping down lead to accurate focus ?If my aperture is set to 2.8,there is some blurring,if its set to 5.6 or above ,it's sharp. It can't be a dof thing since the AF adjusted the barrel,time after time,to the same points of sharp focus that I had determined manually.So what have f-stops to do with autofocussing?
Do any of you who understand the mechanics of the AF system have any explanations as to what is going on ?
Incidentally ,I'm not a compulsive tester but now I feel more comfortable knowing when I can rely on the AF system and when I must work around it and can then concentrate on the photography.
|