[tempelorg]

My K-5 has a serious Back Focus issue with ALL my auto-focus lenses, but not with my manual focus lenses.

It doesn't make sense: With both types of lenses, the same phase-detection system is used.

* When I use a MF lens (A* 2.8 300mm or A 1.7 50mm), the focus detection (using the center AF point) is spot-on with zero AF correction.
* When I use a AF lens (DA 18-135 WR, DA* 16-50, DA 1.8 50 AL, DA 2.4 35 AL, FA 2.0 35), the focus consistently needs about +160-170µm correction (I use the DEBUG mode for this, because the normal "C.26 AF Fine Adjustment" isn't going far enough).

I check the focus both using a Katzeye split prism focusing screen and by manifying the recorded images. I've taken 100s of test images, both on flat high- and low-contrast targets (such as a printed test sheet and printed colored images). Always at max aperture. The result is consistent.

Even comparing the DA 50/1.8 to the A 50/1.7, both having nearly the same wide aperture, and same focal length, their phase-detection AF is 160µm apart. How can this be? And can this be fixed?

Has anyone else seen something like this?

(I also posted this first in the K-5 sub forum but then realized that this forum might be the better place to ask this)

[aleonx3]

My K-5 has a serious Back Focus issue with ALL my auto-focus lenses, but not with my manual focus lenses.

It doesn't make sense: With both types of lenses, the same phase-detection system is used.

* When I use a MF lens (A* 2.8 300mm or A 1.7 50mm), the focus detection (using the center AF point) is spot-on with zero AF correction.
* When I use a AF lens (DA 18-135 WR, DA* 16-50, DA 1.8 50 AL, DA 2.4 35 AL, FA 2.0 35), the focus consistently needs about +160-170µm correction (I use the DEBUG mode for this, because the normal "C.26 AF Fine Adjustment" isn't going far enough).

I check the focus both using a Katzeye split prism focusing screen and by manifying the recorded images. I've taken 100s of test images, both on flat high- and low-contrast targets (such as a printed test sheet and printed colored images). Always at max aperture. The result is consistent.

Even comparing the DA 50/1.8 to the A 50/1.7, both having nearly the same wide aperture, and same focal length, their phase-detection AF is 160µm apart. How can this be? And can this be fixed?

Has anyone else seen something like this?

(I also posted this first in the K-5 sub forum but then realized that this forum might be the better place to ask this)

Interesting... I have it the other way round, all my AF lenses seem to be spot on while I need -9 or -7 for all MF lenses. Both my k-5 and k-5IIs exhibiting the same behaviour. My AF lenses are DA* 16-50, 50-135, FA limiteds, DA 10-17, 18-135. My MF lenses are mainly primes including Pentax A, K and other brands. I use only center point mainly for focusing with wide-open aperture or 1-stop down from wide-open. I use catch-in-focus to test MF lenses, there is also a slight difference in focus accuracy when starting to focus from near point (more accurate) versus start from infinity (less accurate).

Thanks for the tip to use DEBUG if more adjustment is needed beyond +-10 on the AF adjustment scale.

[tempelorg]

I am aware of the manual focus accuracy issue depending on the direction, so I've used both to reach the phase-detected focus. Doesn't make a significant difference related to this issue.

Also, I forgot to point out that I see this issue only with my K-5 but not with my K-3 - there, both types of lenses are spot-on.

I thought the K-5II was supposed to have improved focusing. But you say they're showing the same issue? Too bad you cannot cross-verify with the K-01, as it doesn't do phase-detection focusing.

Anyhow. I always had problems with focusing on the K-5, and it took me a long while to understand what's going on. Even had it serviced in Germany, and it didn't do any good (the technician seemed clueless).

[Schraubstock]

Hi

I wonder, why wouldn't manual focus be more accurate. Unless your eyes have a backfocus (or fontfocus) issue.

[Wild Mark]

My K-5 has a serious Back Focus issue with ALL my auto-focus lenses, but not with my manual focus lenses.

It doesn't make sense: With both types of lenses, the same phase-detection system is used.

* When I use a MF lens (A* 2.8 300mm or A 1.7 50mm), the focus detection (using the center AF point) is spot-on with zero AF correction.
* When I use a AF lens (DA 18-135 WR, DA* 16-50, DA 1.8 50 AL, DA 2.4 35 AL, FA 2.0 35), the focus consistently needs about +160-170µm correction (I use the DEBUG mode for this, because the normal "C.26 AF Fine Adjustment" isn't going far enough).

I check the focus both using a Katzeye split prism focusing screen and by manifying the recorded images. I've taken 100s of test images, both on flat high- and low-contrast targets (such as a printed test sheet and printed colored images). Always at max aperture. The result is consistent.

Even comparing the DA 50/1.8 to the A 50/1.7, both having nearly the same wide aperture, and same focal length, their phase-detection AF is 160µm apart. How can this be? And can this be fixed?

Has anyone else seen something like this?

(I also posted this first in the K-5 sub forum but then realized that this forum might be the better place to ask this)

Yes I have now recently realised this too with my K5. I recently got a 31mm and immediately noticed the problem (I initially thought I had a bad copy - eeek). So I tested with my 80-200 f2.8 and again an AF problem. Hmmm. Manual focusing is fine.

AF adjustment is barely enough.

Interestingly I have the reverse problem with my K20 (which has a Katzeye optibright split prism) - time for some shims (but that is no solution for the K5).

[tempelorg]

My K5II back focuses, generally needing a +10 correction on all the lenses I have.

The fact that many Pentax users have a general BF or FF issue is not point of this discussion thread. The point here is that only the AF lenses have this issue while MF lenses do not. Please only comment your experience if you can relate to that effect, or this thread will get lost in off-topic talk. Thanks.

Yes I have now recently realised this too with my K5. I recently got a 31mm and immediately noticed the problem (I initially thought I had a bad copy - eeek). So I tested with my 80-200 f2.8 and again an AF problem. Hmmm. Manual focusing is fine.

I don't see the relationship to my particular issue here. Have you tested MF against AF lenses?


To rephrase this topic's main question: Is anyone else experiencing the effect that only their AF lenses need BF/FF adjustment while their MF lenses are spot-on?

And does someone know whether service technicians can make firmware-based adjustments to the focusing that are not accessible via the DEBUG menu?

I ask because I had my K-5 serviced for focusing issues and wonder if they managed to change the AF focusing in a way that doesn't affect the MF focusing. Makes little sense to me since both use the same technology, but you never know what the firmware's hidden features are...