DA 16-45 review

kitsios_spyros wrote:

flektogon wrote:

kitsios_spyros wrote:

Hi Peter,

I am afraid that if this is the case indeed, then your lens needs servicing. If it is a parfocal lens (more or less) as Dough suggests, then you need different AF-FA adjustment for 16mm vs 45mm. Zoom lens suffer from it all the time. Even if you buy it perfectly calibrated, this problem may develop with use.

If speed is not important, you can use it in live view. For PDAF shooting, calibrate focus for 45mm on your camera(s), focus at 45mm and then zoom-out and recompose while holding AF locked.

Hi Spyros,

Is it possible that the lens would be faulty if it focuses properly in the LV mode? I doubt. I still think that this is a problem of the PDAF in my camera. I have no idea how the back/front focus adjustment works in the Pentax bodies. Is this somehow done by mechanical movement (re-positioning) of the PDAF "unit"? Is it possible to make such adjustments independently for different lens angles? And I doubt that the K-x allows such adjustments at all.

Regards,

Peter

Hi Peter,

In LV-CDAF the focusing sensor is the same as the imaging sensor, so no adjustments - correlations between the two is needed.

In PD-AF the focusing sensor is another different sensor in the mirror box. What this sensor sees (actually the distance from the focus target that this sensor detects) must be correlated to the distance from the imaging sensor This brings the new for accurate calibration.

Now, in zoom lenses if all is correct, one must need one value in the AF-FA menu to properly set the lens on the camera. A few things can go wrong

1) Different values will be needed within the FL range of the lens. I can tell you that it is not always a linear trend. My Tamron 70-300 has a wavy trend of AF-FA values going from 70mm to 300mm and Tamron had told me that they may not be able to fix it as it is an old design (their words). I understand that it may not worth the effort for such a cheap lens.

2) The needed AF-FA value could be outside the +/-10 values the camera allows. This needs to move the global "0'" point of the camera. My SMC135mm/f2.5 is way outside the limits of my K5/K5ii camera, so no CIF. On my K10d it needed a +230 value in the debug mode (I believe that +100 on K10d must have been equal to around +8 on K5/K5ii cameras)

3) Both the above

4) The above may be different for different distances.

The only dSLRs I know that allowed adjustmens for different FL of zooms (2 values for wide and tele end) plus distance where a few from the last Olumpus dSLR (irony..), Canon 6D and we will see what D500/D5 wil bring. From the lens makers only Sigma allowed for combatibility for some of its newer lenses with the USB AF calibrating dock. Great tools, but why not for all Sigma lenses?

It surely worths the effort to callibrate your lenses on your cameras, especially for bright lenses.

But regarding this 16-45, I would not blame the camera for your problem, just the lens. Kx was free from the back-focusing errors under low intensity tungsten lights that plagued Kr/K5 series.

If you do like your 16-45, just send it in and enjoy your corrected lens. They should also probably fix possible missallignements of the glass elements from typical usage. This should improve sharpness at least at wide apertures too.

Thanks for the explanation Spyros. Well, I am not sure whether it is worth to have my lens re-calibrate. Very likely it would cost me as much as I paid for the lens, which was around $400 (CAD) and there is no guarantee that the Canadian "experts" would be able to achieve something. I have very, very bad experience with the warranty repair of one my Canon camcorder. My lens is decentered as well. Pentax definitely can't be proud on such a lousy product, regardless how great its optical performance could be (i.e. at least theoretically). But everything is only and only my fault, as I ought to be more diligent testing it while I could return it.

Regards,

Peter