[mxlinn]

Hi, all

Just acquired another copy of D-AF 100mm f2.8 Macro WR. The previous owner seems to be a savvy shooter and claims no focus issue when he owned the lens. But I've experienced the following - if intended objective is between:

8'-infinity - complete back focus
3-8' - moderate back focus
3'-macro - acceptable

Not sure what the problem is but haven't tried AF fine tuning on my K5. To me, the AF is corky and completely unpredictable and believe repair or calibration is needed. The attached photos roughly represent the distances mentioned above. Please express your opinions and suggestions on this. Thank you in advance. The lens is overall pretty clean with no major visible marks but it's NOT like new, optics are flawless though.

Single AF point aiming at the girl with the dog - 20' away. I took a series of photos at this theme and focus failed completely

Single AF point aiming at tip of the front leaf - about 5' away

Single AF point aiming between the eyes - about 3' away

---

Last edited by mxlinn; 01-15-2013 at 08:18 PM.

[jatrax]

I'm sure you won't want to hear this but I don't think there is anything at all wrong with the lens. Looks sharp to me at the point it focused on. The problem is that the AF 'boxes' on the k-5 are MUCH larger than the little red square. The red square is only there to indicate which AF point is being used not how big it is. Calling it an AF point is probably the source of much misunderstanding, it's not a 'point' its a 'box' that is quite large. Anything in that 'box' that attracts the AF system will be what is locked on to.

The reason you are seeing acceptable AF at close range and poor performance at a distance is that at a distance the AF box covers much more area and thus has a better chance of locking onto something you did not intend. Take your first image for example, the center AF box has been demonstrated to cover just about everything between the focus screen brackets. So focusing on the girl allows the box to cover the girl as well as some of the more contrasty background that it actually locked onto.

There are only 11 AF boxes on the K-5 and they cover most of the image so based on that you can see how much area each one covers. Other brands have more 'points' and thus each point covers less area, but on Pentax this is what you get.

Back focus / front focus is clearly demonstrable using controlled tests and the error is generally inches or even millimeters. You have miss-focus not front/back focus, and only improved technique will correct that.

[Ole]

Did you use center point auto focus or did you let the camera decide which point to use?

[mxlinn]

I'm sure you won't want to hear this but I don't think there is anything at all wrong with the lens.

Not at all, I've long been suspecting my technique isn't up to par. So, what do you suggest - using auto-select mode on AF points? All photos posted were with single-center-point AF.

[mxlinn]

Did you use center point auto focus or did you let the camera decide which point to use?

I manually selected center-point AF on all photos.

[jatrax]

Auto-select is worse in my opinion, you should not trust the camera to select what you want to focus on.

And I didn't mean to disparage your technique though upon re-reading it might have come out that way. What I meant was with Pentax, since the boxes are so large, you have to understand that when the red box lights up something in that general area is in focus, but not necessarily what you wanted. So the technique you have to develop is making sure that what you want in focus is the only possible high contrast target within the box. That's not always possible and sometimes you have to use the quick shift to adjust a bit.

I always use center point as well. What I do is try to get the target within the box, lock the AF and then re-compose the shot. Never give the AF a choice of things to lock onto. I get good results doing that, though by no means 100%. One thing I have found that seems to help is to focus twice, lock onto your target, then let the button go and then re-focus. Since the lens should already be close, I get a higher percentage. Not sure why that works to be honest. But if you do that and the lens re-focuses then you have a situation where the lens has two or more possible targets. This should be a red flag that you have a good chance of missing focus.

I use the AF button on the rear, not the shutter button to focus with by the way.

Most important thing is just be aware that the red box is NOT the AF point. Possibly on some brands you can trust the little box to be the focus point, but you cannot do that on Pentax, at least with current models.

[mxlinn]

I use the AF button on the rear, not the shutter button to focus with by the way.

Very interest and perhaps valuable, any reason?

[jatrax]

Very interest and perhaps valuable, any reason?

Because I use focus and recompose a lot, the back button allows you to separate the focusing action from the shutter action. So using back button: focus on target and as long as the target (or you) do not move then you are still in focus no matter how many frames you take.

For something like street photography where you may have pre-focused you don't want the shutter button to try and focus every time you use it. You can set to manual focus but by using the back button you can focus using AF then just sit and wait for something to come into frame without risking losing the focus.

For macro or product work on a tripod, use back button for focus then you can mess with lighting etc all you want and the image stays in focus every time you trip the shutter. If you used shutter button to focus then every shutter press would re-focus and take a chance of locking onto something different.

It is not for everyone, and it takes a lot of re-learning if you have used shutter button focus for awhile, but I think it is worth the effort to learn. Try it and see what you think.

[ramseybuckeye]

It looks severe, I don't know if AF adjustment could correct that. Can you return it?

[Canada_Rockies]

Because I use focus and recompose a lot, the back button allows you to separate the focusing action from the shutter action. So using back button: focus on target and as long as the target (or you) do not move then you are still in focus no matter how many frames you take.

For something like street photography where you may have pre-focused you don't want the shutter button to try and focus every time you use it. You can set to manual focus but by using the back button you can focus using AF then just sit and wait for something to come into frame without risking losing the focus.

For macro or product work on a tripod, use back button for focus then you can mess with lighting etc all you want and the image stays in focus every time you trip the shutter. If you used shutter button to focus then every shutter press would re-focus and take a chance of locking onto something different.

It is not for everyone, and it takes a lot of re-learning if you have used shutter button focus for awhile, but I think it is worth the effort to learn. Try it and see what you think.

There is a second option available, just for information sake.

On my K-10 I have just set my AF button to cancel AF, and will try that for a while. As long as I hold the AF button down, the camera will act as if it is in manual focus mode, but when I release it, the shutter half press will refocus. I've tried cancelling the AF on half press and using only the AF button as well. I'm not decided yet which I prefer.

[stormtech]

I agree that I find the first image totally unacceptable. jtrax has more than likely found an issue that is the culprit, I would try to do the following tests to really find out if it is the lens, the K-5 focusing system, or some kind of operator fault.

I noticed in the exif of you images that your K-5 is running firmware 1.11. The first thing I would do is update your firmware to the current version which is 1.13:

Downloads & Literature - PENTAX Imaging USA

Next, if you have a tripod, set your yourself up and choose a static object outside in good light - like a street sign, the trunk of a tree etc that is ~10' away. Take a couple images using the 2 second time which will kill any possible vibration. Then try a couple images using live view since that uses a different focusing system. Also try both at different aperture settings.

By doing this and closely reviewing your images, you might be able to tell if it is in fact a problem with the lens. You might want to try the same thing with a different lens at the same time also for comparison sake.

I do something like this every time I get a new lens - used or new - to see how it should perform. If nothing else, if the lens performs properly in a fairly controlled condition like above, if I have problems later I know I can't blame it on the lens and look at myself for operator error.

[langkata]

I've been meaning to weigh in here, but haven't had a chance until now. I purchased your previous copy of this lens last month and experienced similar issues on my k5 and k10d.

Based on my experience with the copy I have, I very much doubt this is due to technique, or if so only in part. I've taken this lens head to head with my 50-135 zoom to see if the macro can take its place (my original intention when buying the macro). I rarely have any problems with center-point focus on the zoom, so I suspected the lens from the start. This is the first lens I've had with an issue like this and it was very frustrating at first.

It's kind of hard to find information on this particular issue with this particular lens, but I did find a few things of interest. I read a suggestion that a macro lens such as this is optimized for the closer focus ranges and can therefore be more prone to being off for longer ranges. Seems pretty logical, I guess. Maybe it isn't recognized as much because it's not a lens that is used as often at the longer ranges?

I seem to have had success with a +3 micro adjustment after some pretty simple testing. I sat in a chair and chose a door knob about 9 feet away (just short of infinity focus) for my center-point focus subject, taking 3 shots: manual focus, auto-focus, and live view for each micro adjustment I tried. I don't suppose the manual focus is necessary here, but did it out of curiosity. Live view is going to be spot on, unaffected by the focusing issues, allowing you to directly compare each adjustment setting between the two (on the camera lcd then on monitor to be sure). My first guess was a +2 and +3 happened to get it right. After the adjustment I was getting pretty consistent results at all distances.

I may not have it exactly right yet, but will fiddle some more. Let me know if you have any questions and I'll pass along anything else I find.


Here's an example from a couple of days ago, post adjustment:

[mxlinn]

I consider this to be one of best lenses Pentax has in its current line up in terms IQ, and I really missed it so went out got another copy from the auction site. The newly acquired lens has been heavily used and nowhere near the condition of my first copy and the focus issue is so severe didn't even think I couldn't adjust it. It's now on +9.

[stormtech]

I consider this to be one of best lenses Pentax has in its current line up in terms IQ, and I really missed it so went out got another copy from the auction site. The newly acquired lens has been heavily used and nowhere near the condition of my first copy and the focus issue is so severe didn't even think I couldn't adjust it. It's now on max at +10 and still off slightly.

And what do you expect.....ebay....heavily used.....

[mxlinn]

Well, guess not everyone kept their gear as pristine as yours. Even the copy you sold me was back focused, still loved it!