[simico]

When subject is about 2.5 meters away (and under), AF Micro adjustment of +10 gives shaper result than 0.
When subject is around 6 meters away, AF adjustment off/ position 0 gives shaper result than +10.
When subject is closer to infinite focus, AF adjustment 0 gives clear results while AF adjustment of +10 is very very front focused.

This is usually a lens problem, your lens needs calibration at the repair centre. Assuming your method was correct and free of user errors.

[philbaum]

This is usually a lens problem, your lens needs calibration at the repair centre. Assuming your method was correct and free of user errors.

For modern AF lenses, does lens calibration refer to a mechanical adjustment or a electronic data table, or both??

I've had the same experience as Katsura, e.g. calibrating close up then finding errors farther away. I understand the theory about dof and distance from the subject, but i wonder if the size of the focus sensor throws a monkey wrench in the theory. In any case, for my theatre work, i shoot from 10 to 30 feet and i see no problem with double checking the focus check at those distances where i work.

[bossa]

I've been trying to get my lenses calibrated all week between two K5 bodies and it's been a nightmare. I finally did a few LiveView focused shots and even they're front focusing on the one particular lens I've been working on today. Do AF fine adjustments affect LiveView?

BTW> I am perplexed as to why the OP is making '+10' AF adjustments when his lens is already front focusing. Surely that will only make it front focus even more.

[yuareblessed]

I agree with all the above post you guys posted. I've used 50-135 f2.8 and FA50mm f1.4 lens with my K5 and it just frustrates me with the inconsistent AF issue. I've tried +10 AF adjustment and thought it worked the best. But once I change my shooting condition (ie other place and lighting) the AF is still not the sharpest and there are front focus & back focus issue.

Seriously, I'm considering of switching to other system such as Nikon or Canon. This is not the only body that i own that was plaque with this AF issue.

As for now, I use liveview to AF if my aperture is larger than 2.8. Just have to live with it for the moment.

[Turtle]

Sorry to hear it yuareblessed. Are you using the latest firmware? I found it helped.

[PhilippeG]

You're not crazy Katsura The (very) low light issue is still present, and the firmware updates didn't fix it. It improved a lot with the updates, but it's still not 100% fixed. Below about EV 2 or 3 in tungsten/yellow/red light you will get a front focus that will only get worse as the light dims.
I have sent my k-5s back several times because of multiple issues, so I've been able to shoot with three different ones, "old" and recent serial numbers, and I can tell you even the cameras manufactured in mid 2011 show the same issue.
It's been thoroughly tested and documented by Falconeye, and is known to be a hardware problem.

It won't be a problem for most people (and some will insist there is no issue) because they just don't shoot in such low light. But I do and it drives me nuts.

[hagla]

Why is no 1 shot at ISO 1600 and no 2 shot at ISO 6400..?

[yuareblessed]

Yup turtle, I'm using the latest firmware which is 1.03. I didn really helped. I almost bought an Sigma 30mm f1.4 but after considering this AF issue, i would probably cease buying pentax mount lens and slowly switching to other systems. Sigh.

Really like the K5 but reliability issue + the amount I've invested in Pentax makes me think it's not worth it any more. The biggest culprit is the AF issue. Other stuff like slow SDM is the minor one.

[Quicksand]

I'm using the latest firmware which is 1.03.

The latest firmware for K-5 is 1.11. There's nothing in the official changelog for 1.10 or 1.11 that says anything about focus, but you never know.

Latest K-5 Firmware Update : Software Downloads : PENTAX

[pentup]

I'm using the latest firmware which is 1.03.

That's not the latest. The latest is 1.11. Latest K-5 Firmware Update : Software Downloads : PENTAX

But it appears that the firmware which you have is the last one that addressed the AF issue. I don't know why I bothered to update to 1.11; it didn't add anything I needed, and for all I know, it might have introduced new bugs.

[filorp]

i have exemined ff issue with my k5 very thoroghly.... the conclusion is as fallows:
with the wide angle lenses when the subject is very small comparing to AF point the ff is obvious - if i use zoom lens da16-50 or 50-135 i set the zoom to the narrower angle press the af button and rezoom to desired angle, with the fixed lenses (like da14) i use liv view....
i think that the reason of ff is the size of the subject (or contrasty part of the subject like eyes or mouth) to small for af sensor - even in good light with the wide angle lenses am having problem with focus accuracy - hence i sold da12-24 (optically brilliant lens) as it was thoroughly inconsistent lens with k20 and with k5....
the af point must cover entire contrasty target if its not, it misses, even if it cover entire subject like face but the contrasty points like eyes or nose or mouth are to small it misses anyway in low light i think there is something wrong with af compensation in such situation and small af computer sets incorrect formula then ff appears...
the problem is apparent especially in low light, for lens f2.8 below 4EV, for lenses f1.9 f1.8 and f2.0 below 3EV i can't get consistent results - solution is as above rezoom or live view
i cant understand why pentax doesnt allow to use AF auxiliary light in such light condition, i dont have to tell that my d700 focus perfectly in the same light condition with or without assist beam.... i got so frustrated with pentax af system that i'v started slowly swapping my equipment into nikon brand, very slowly as i dont like ergonomy of different brand cameras, i gave Pentax two years to improving af system up to pro standard if thay dont i will have to live with weird nikons ergonomy

[bossa]

Thanks filorp. That all makes rather good sense to me. I need to try that zooming technique of yours and see if that improves things a bit. Is it possible that the single point AF target is just too small? Maybe better success would be had with AF if the 5 zone mode was selected. I know that isn't as refined as one would like because it doesn't really allow picking one small part of the main zone to focus on but it might just get more successful results generally.

[jolepp]

The latest firmware for K-5 is 1.11. There's nothing in the official changelog for 1.10 or 1.11 that says anything about focus, but you never know.

Latest K-5 Firmware Update : Software Downloads : PENTAX

For me the upgrade to 1.11 seems to have made a difference for the better as far as AF goes (see: https://www.pentaxforums.com/forums/pentax-k-5-forum/165532-k5-firmware-upgra...ml#post1714104).

[Katsura]

Here's my latest update after spending nearly a whole day, under different lighting conditions, against my collection of 10 or so lens:

1. I am not sure why, but updating firmware *may* affect your old lens calibration data. I can not 100% confirm this (I can only upgrade, not downgrade the firmware) however, calibrated micro AF adjustment on an old firmware *could* cause FF/ BF issue on a newer firmware. We need more data to support/ confirm this, but, I suggest if you do update firmware, then reset and recalibrate your lens. It's okay for ppl with just 2, 3 lens, but a pain if you have more.

2. Use a variety of techniques to check AF accuracy, this should include AF adjustment charts, compares between LiveView focus and standard AF.
Someone in this topic, mentioned that you should always adjust AF against a close subject at large aperture. While I agree you should ALWAYS use the largest aperture for adjusting AF. I do not agree to ONLY USE CLOSE subjects for AF adjustments. See below.

3. Out of my collection of 10 or so lens, more than half of them, both prime and zooms, exhibits slight to moderate BF symptoms (under normal lighting) when tested against a focus chart. HOWEVER. once compensated for the BF issue by dial in positive values of AF Micro Adjustment. These lens all exhibit moderate to severe FF faults on distant subets. The most extreme example is my FA77 Ltd. On the latest firmware, it exhibits a moderate BF while focusing on a chart that is about 1 meter away. An AF adjustment of +5 seem to give the best result compensating for this. However, if I change the test subject to a wall close to infinity, the lens at AF adjustment +5 has severe FF. This is verified by live view AF and live view assisted zoom MF - I take a mark of where the actual focus distance on the lens barre while using assisted zoomed MF in LV, then switch to standard AF - every time the AF shows a very noticeable FF (the focus distance scale comes up short of the marked actual focus distance.

4. So, unless my technique of using the focus chart is very wrong (in real life situation, I do notice slight BF when AF adjustment is 0.) then the lens are exhibiting weird behaviour that makes them BF at close distance and FF at far subjects. In the end I decided I can live with the ever so slight BF because the FF is mroe significant (ie, the adjustment result gives worse result at far than the benefit they give at close), and not bother with them.

5. In conclusion, the out of focus photos of the original post were cased, in this case, by
- not resetting AF adjustment data when upgrading the firmware
- using close distance subjects only for adjusting AF, gives bad results when the subject is further away
So, in this particular case at least, low light AF inaccuracy *may* not be the cause of FF, still, I'm not saying low light FF doesn't happen, just that in this case it is probably not the *main* contributing factor.

I recommend that
- use BOTH close and far subjects when adjusting AF, I can't stress this enough! Do not assume correct adjustment at close range will be accurate for far subjects.
- reset your adjustment data after a firmware update.

[audiobomber]

I don't see anything in your tests about wavelength. When you tested your FA 77 on a near wall and far wall, were they both illuminated only by the sun? My K-5 had a problem in tungsten light, not sunlight, and increasing the intensity didn't resolve the FF. AF just didn't work properly in yellow light.