[hjb981]

I bought a K-5 with DA 18 – 135 a couple of weeks ago. I have installed firmware 1.03. After taking some portraits of people where their ears were sharp but faces soft, I used the auto focus fine tuning to adjust it. To test the focus, I took pictures both of curbstones across the street (StigVidar's tip in an old thread I found) and a box of matches standing on a soft carpet (lots of fine detail). I suggested this fine tuning to someone else a few days ago on this forum, because after performing it I get very nice and sharp portraits.

Now I have taken some more pictures, many of them outdoors on far-away things. When using the lens’s shortest focal length (18 mm), I noticed that things looked slightly soft, and that things closer to the camera were sharper. Turning off the +4 focus fine tune I had set earlier made it better, so I tested it more thoroughly and ended up on the minus side of the scale for the sharpest pictures.

To summarize: with no adjustment made, there is a tendency towards back focus in portrait situations, and a tendency towards front focus in landscape situations. I can not compensate for both of these at the same time in the fine tune settings. This is all with natural light (btw, I get sharp portraits with the +4 setting in tungsten light also).

Is this something one could expect from a phase detection auto focus system with a zoom lens if something is not calibrated, or is it very strange (unusual)? Is it possible to get the camera and lens calibrated to maximize the focus accuracy at all focal lengths?

Thanks in advance!

/Jonas

---

Last edited by hjb981; 05-02-2011 at 06:39 AM.

[Croc]

you might want to check my thread here, sounds a lot like you could be having same problem https://www.pentaxforums.com/forums/pentax-k-5-forum/142227-pentax-auto-focus...ed-fixing.html

[hjb981]

I did read your thread, Croc, and I think that I am experiencing something different. As far as I can tell, I get good focus with all types of light (tungsten, natural, fluorescent etc), as long as I have the right auto focus fine adjustment for the circumstances dialed in (and this is the same indoors and outdoors). I have taken pictures of my matchbox on the carpet both in daylight through the windows and i tungsten light, and the plane of focus ends up in the same place in both those situations. I have also made portraits both in tungsten lighting and in bright sunshine that are sharp.

What happens for me is that I need different auto focus fine adjustment settings depending on the distance to the subject and possibly also the focal length used with my lens. This is only noticeable outside, since it only changes with subjects requiring more or less infinity focus compared to closer subjects.

I am going to contact Pentax tomorrow to see if it is possible to have the camera calibrated to compensate for this. Hopefully someone here can give me some insight - I always like to have some understanding of what is going on. If I understand what is going on, it is easier to discuss it with a service technician.

[sworks]

This sounds quite similar to the problems I've been having. I only have the 35mm 2.8 macro Limited lens but a +2 calibration works well in close up situations in all kinds of light, however, outside with subjects further away, even it great light, the camera still produces fuzzy images. I've been trying different settings today between +2 and +5 outside with no clue really as to what is going on! I've also been trying to see exactly where the camera has focused in relation to the red single point focus indication that I've been using. I get the impression that it is focusing much further to the right of the focus mark.

[hjb981]

To find out the area of the focus sensor, try focusing on the edge of a box that is close, and then point the camera slightly more to the side of the box, focus again and repeat this until it focuses on whatever is behind the box. I have focused on tilted things from different angles, and my center focus sensor (the one I have been using) seems to be in the middle, like it should.

I would greatly appreciate it if anyone can point me to a more technical understanding of what is going on, e.g. this is something that can happen because of X, and the solution is Y. If anyone would have such knowledge or a theory, of course.

Best regards,

Jonas

[rhodopsin]

That zoom changes aperture with focal length change -- f/3.5-f/5.6

Set auto focus fine tuning with lens zoomed to widest FL possible at f/3.5. Then setting will be correct at all focal lengths.

What happened -- fine tuning was set when lens was zoomed to long FL at f/5.6. That setting was not accurate, when lens was zoomed to shorter focal length where smaller focus adjustments make bigger differences, and when aperture is wider at f/3.5.

[30v iPentax]

Siksi minäkin ostin Pentax 50-135 F2.8. Ai jai kun tulee hyviä kuvia, ilman konstailuja.

[Ash]

Jonas, it will be difficult to be sure of a cause and solution of your problem unless it were seen on a real shoot. Posting examples of this phenomenon may help, but it still may not give an answer. Also give details of what lenses, shooting distances and focusing technique you used to get these anomalies. I'd suggest being consistent with lighting, focus point (without recomposing the image after focus lock) and aperture settings before testing again.

[5teve]

Hi HJB

i didnt quite understand if you are getting focus adjustment drift throughout the focal length range?
eg it is front focusing at 18mm and back focusing at 135

I have a lens that exibits this behaviour also, from 28-50 its fairly accurate at 0 adjustment, but from around 50-70 it requires anything upto +8. Makes it a pain to get decent shots in low light. My new tammy 70-200 shows no such problem tho... that i have noticed!

Is this what you are experiencing, or is it at a fixed focal length (say 18mm) the accuracy of the focus system varies from front to back focus dependant on the subject distance from the lens?

Just wanted to clarify....

Steve

[hjb981]

Technical details: Pentax K-5, DA 18 - 135 F3.5 - F5.6, Hoya HD UV filter (should be a good filter). Focus technique: Center focus point used. Far-away shots were taken at about 30 - 150 m (100 - 500 feet). Close-ups at about 1 - 5 m (3 - 17 feet). I have no pictures at the moment (I deleted the test shots, and I don't want to post other peoples portraits without asking them), but I will try to make some new ones and get back.

And to clarify: I think that the change has more to do with distance than focal length. I will try to investigate a little more and get back.

Thank you all for your answers!

[Ash]

Also try with a screwdrive lens (if you have one) and see if there is any difference in focus accuracy.

[simico]

DA 18 - 135 F3.5 - F5.6, Hoya HD UV filter (should be a good filter)

First drop the filter (no matter how good it should be in your opinion), then also try the lens on another/different camera to see if your lens or camera is at fault. Different focus error at wide and tele end is usually a lens problem, in that case the lens needs calibration in the repair center.

[rawr]

Agreed - dropping the filter off your camera should be your first step. Pentax didn't calibrate their cameras and lenses with random UV filters in mind.

[hjb981]

Thanks again for your suggestions. I think that I may have found the problem: my fine tuning procedure. It is embarrassing to admit, but I think that I may have misjudged where the focus actually was. I was trying to have approximately the same amount of sharp on my carpet when going forward and backwards from my matchbox in the picture, but with somewhat more sharpness when looking further away from the focal point, since the angle of light changes less when going further away, and thus the focus depth should be bigger in that direction. Then, of course, things (detail in the carpet) gets enlarged when they are closer to the camera, which further complicates proper judgment of how much of focus I have in front of my target as opposed to behind.

There is one thing that still puzzles me - the several portraits with sharp ears and less sharp eyes - but since that does not seem to reproduce, maybe there was some other reason for that. I find it unlikely that something would shift back and forth when it comes to how the auto focus works, so perhaps it can be attributed to user error. I will now continue taking normal pictures (not tests) and if I would see anything strange after having many pictures to judge (statistics is always good when judging these things), then I will see if there is a need for adjustments. In that case I will post some pictures and ask again, but for now, maybe I don't need adjustments after all. In that case, I will also try without the filter. My apologies for taking your time with something that is likely a non-issue.

[30v iPentax]

Minkä hinnassa säästää sen lopputuloksessa menettää.