[JamieP]

I am trying to adjust my lenses on my K-5. I am coming from a k10, so this is my first time going through the process and I expect that I'm the problem more so than the equipment. I'm getting inconsistent results at different focal lengths and am leery of setting a value that ends up making things worse.

A couple of notes on what I'm doing:

1. Using the focus chart from Yvon Bourque (45 degree chart)
2. Using the max aperture
3. Shooting in natural light. Wasn't super bright, but no problems locking
4. Shooting from the minimum focusing distance- 3 times, forcing the AF to re-lock each time
5. Reviewing the shots
6. Making a focus adjustment in the menu
7. Repeating as needed until the AF is on target

After competing this, I repeated the process at a distance beyond the minimum focusing distance, but not so far off that I wasn't confident of where I was locking.

My dilemma is that my both of the lenses I completed this for needed significant adjustments at the minimum focus distance (+5 for the 70 mm and +7 for the 15mm), but much smaller adjustments (+1) at the slightly further distance. My understanding is that process should be done and the minimum distance to minimize DOF, but I'm wondering if the AF is less consistent at its limits.

As a note, I'm expecting the DOF 50/50 front to back at the minimum distance and slightly shifted back from the second distance. This is based on DOF calculator for my lens, aperture, distance, etc.

Any help related to my process or comments about AF at the minimum distance would be appreciated.

[Parallax]

The actual focus point in the camera is not necessarily exactly where the indicator shows it, and is much larger than you would think. Mount your chart on the side of a box. Draw a vertical line up the box through the exact center of the chart. Make a mark where that line is on a piece of masking tape on the table top. Focus with the box/chart at 90 degrees, then turn the box to 45 degrees with the marks on the box and table lined up. That way you know that no matter how big the focus sensor, or what its exact location, it is focused at the distance of the exact center of the chart.

[JamieP]

Didn't think to mention it initially, but to save some initial questions I think I had the basics of shooting covered- tripod, remote, Av, center spot focus, etc.

Your help is appreciated.

[elho_cid]

With all respect to Yvon, his chart is capable of giving you some hint on BF/FF but it is not suitable for fine adjustments.
I suggest to make the adjustment in the field, during actual usage of the lens. Change the settings iteratively, after dialing +2 for BF give it some more try and only then go for further correction.
Also I see you own 2 Sigma lenses. These are difficult to compensate - and what is worse, my sigma lenses appear as one single lens to my K-5 so the adjustment made for one applies to the other. Really unfortunate as one of them is FF (180macro) and the other is BF (Bigma).

[dflorez]

Afraid my reference for focus adjusting is a k20, and not the 5.

But! as the fellow above here mentioned, it may be best without a chart. The first time adjusting I tried charts as well, didn't really help. I ended up just opening a book with lots of text, and chose a sentence as my point.

I eventually got it much closer (a tamron 28-75 2.8-0), but unfortunately never perfect. Also, what was weird, it would be great at one focal length, then off at another.


Interesting fact too, it took a full 10 adjustment on the k20 (forgot if it was minus or plus), but on the k7 it's been great from the start.

[MPrince]

I, too, struggled with autofocus adjustments.

Then I found this site:

AF microadjustment for the 1Ds mark III, 1D Mk3, 5D Mk2, 7D

And my struggles ended.

[dadipentak]

That definitely seems the way to do it, Matt, but--either way--this obviously requires patience and discipline way beyond anything I could muster (especially with multiple bodies and too many lenses.) I'd simply go nuts and start smashing things. I'm going to have to figure out how to get someone to do it for me.

[shang]

Jamie, I had the same problem as you before.

When you do focus adjustment, if you find that subject distance matters, then it means that your body has some problem. Just replace it or send it for repair...

[Class A]

A couple of notes on what I'm doing:...

You are doing a lot of things right and I'm afraid there is nothing you can do to improve the situation. Unlike shang, I don't think your camera body has a problem. Rather, it is a fact of life that AF adjustments are not constant regarding focal length and/or subject distance. This means that even if an AF adjust method (e.g.,like the one Matt suggested) works for one focal length / focus distance combination, it need not produce a setting that will work with another combination.

My understanding is that process should be done and the minimum distance to minimize DOF, but I'm wondering if the AF is less consistent at its limits.

Using the MFD helps to minimize DOF but then you are optimising for that particular distance. If, in practise, you'll shot other distances more often, you should use these distances to optimise your AF adjustment.

Have a look at my AF adjustment hints. Maybe you'll find another helpful hint or knowledge nugget.

When I knew I was going to do a portrait shoot a while ago, I took a doll and did a series of shots with varying AF adjustment. I settled on the AF adjustment value that gave me the best setting for getting the eyes into sharp focus. I didn't use a focus chart because the "AF points" are not points but areas and I wanted a setting that would work in practise. Maybe I dialled in some backfocus because the camera, in practice might focus on the eyebrows and my setting then achieves focus on the eyes (I don't think that actually is the case, I'm just using this possibility for the sake of the argument). In this case my setting would have been the wrong one for shooting flat walls but the latter wasn't my aim. Maybe I've been lucky but my approach worked like a charm.

Note that I do not advocate using a book, or similar, as a target, unless you want to photograph books. You just don't know what the camera will focus on.

Summarising, you may or may not find focusing charts useful (my "AF hints" contain some criticism) but I'd strongly suggest to double check your settings under realistic shooting conditions. The proof of the pudding is in the eating, isn't it?

[rawr]

I am trying to adjust my lenses on my K-5. ... I'm getting inconsistent results at different focal lengths and am leery of setting a value that ends up making things worse.

My general caution would be that if you aren't very careful with your testing technique and AF adjustments, you could indeed make things worse, not better.

If you find yourself doing AF adjustments a lot, or have some doubts about your own test rig, this tool might be worth buying. It seems a handy bit of kit:

SpyderLensCal - Datacolor - Global Leader in Color Management Solutions

The Datacolor tool resembles several other test kits out there on the web, including the AF test chart used by Pentax in some of their camera service manuals.



chart here:
pentax af confirmation chart - rawr's Album: Useful - PentaxForums.com

[JamieP]

Thanks for all the responses. From what you all have posted and some additional reading I've done on the topic, I feel like my testing processes are sound. I'm going to test again to see if I get more consistent results.

I am also very intrigued by the interference pattern / live view testing method (thanks MPrince) and want to try that out. I've never taken to the time to understand how live view focusing works from a technical perspective, but it is interesting that this isn't affected the lens adjustments. I guess I'll read up on that next.

[chicomarks]

i had the same issue of focal length drastically affecting auto focus when using my k10d. id use the debug mode to calibrate to min focus diatance, then id try focusing on a more distant object and the focus would be way off. i did this with the 18-55 kit and the tamron 17-50 with similar results. i then tried with my 16-45 f4 and it seems to focus pretty consistantly through the range. you could put thia down to a higher f stop affording greater dof, but with my kit lens experience i would question this.

all this said, i still dont know the cause of the issue... id really appreciate some answers! is my camera body to blame? is it the lenses? im thinking of buying an af prime for eg the new cheapo 35mm, but these focus issues are making me wonder if its a good idea...

[JamieP]

I finally got to spend a decent amount of time try to adjust the focus of the DA 70 and now I'm really frustrated...

I was using the interference pattern image. My approach was to use LV to set the focus, switch to Centerpoint AF and see if it moved the focus ring. I would look at the images only as verification that nothing was out of whack and I was seeing a clear interference pattern. I really liked this approach and it was much less subjective than the focus chart. I landed on a +6 adjustment to lock in AF with LV. Once I had that adjustment set, I tried multiple distances from the target and had no issues. I was feeling really good at this point.

Then I decided to go into my basement and repeat to see if the amount of light affected things. This is where it went downhill. I was shooting in all natural light in both spots, but in the basement my shutter was 1/60 as opposed to 1/180 or 1/250 upstairs. All other settings the same- F2.4, ISO 400, AWB, etc and at no time did the camera seem to be struggling to lock the focus. However, in the basement I had to change the adjustment to -2. I again tried different distances and it had no impact.

I'm really surprised by this. 8 focus adjustment steps is a significant difference. I knew that artificial lighting was known to affect focus accuracy, but I wouldn't have expected a couple of stops of difference in natural light to have that impact. I plan to repeat this with my other lenses, but I might not be home again during daylight for several days.

Has anyone else seen similar behavior? Would you think this is an issue with the body or the lens?

[Class A]

is my camera body to blame? is it the lenses? im thinking of buying an af prime for eg the new cheapo 35mm, but these focus issues are making me wonder if its a good idea...

Unfortunately, I do not have the answers to your questions, but if I were you the problems wouldn't detract me from getting another lens.

Larger distances mean larger DOF, so I'd calibrate the lenses for distances roughly equalling 20-25 x the focal length (or closer if you use the lenses for close focusing often). Such a calibration should make sure that the subject is within the DOF most of the time.

Changing the camera might help, having the lenses calibrated by the manufacture might also help, but I cannot make any promises.

My approach was to use LV to set the focus, switch to Centerpoint AF and see if it moved the focus ring.

I think it would be better to set the lens to infinity and/or MFD and then see if engaging AF will produce the interference pattern again. It won't be the case a 100% of the attempts but you should use a compensation that gets you there most often.

I knew that artificial lighting was known to affect focus accuracy, but I wouldn't have expected a couple of stops of difference in natural light to have that impact.

On a K-5, which has the colour temperature sensor, artificial lighting should not impact on focus accuracy.

Has anyone else seen similar behavior? Would you think this is an issue with the body or the lens?

A lens issue seems unlikely.

Maybe your K-5's has a hardware problem but it may be worth waiting for the next firmware update. It is known that the K-5 has trouble to reach optimal focus in low Tungsten light. Maybe your issue is related. If the next firmware update addresses the known problem, maybe your problem will be gone as well.