[zosxavius]

I did some testing and discovered that my k-5 was backfocusing significantly in low light. I set it to -10 and things started looking more accurate. Outside I see not a whole lot of difference between -10 and 0. I use mostly primes, so AF is just helps confirm usually for me anyways so I haven't noticed any real issues until now.....

So long story short, my 12-24 is in need of some repair, but beforehand I started noticing lots of shots had elements that should be in focus given my fstop, but weren't. On my k-7, I rarely had issues with focus other than times where my dof wasn't long enough and I locked on something in the foreground as sometimes happens when you are moving fast. I would shoot a building @ f11 and 12mm and the top wouldn't be in focus. I thought my lens might be decentered, but shots on a tripod always looked fine across the frame. So then I dropped my camera and 12-24 off a tripod. The mount on the lens sheared the screw holes. There is a good bit of plastic left as thankfully they only broke halfway down. I think I can fix it. If anyone knows what threads and dimensions the original screws are let me know. I at least want to get it back together and mounted on a tripod to test it. The elements might still be ok. Its a pretty tightly built lens. The hood and mount took a most of the damage. The glass is all still in one piece thankfully.

So I started using the kit lenses I have since my M28/3.5 also needs sent out to be cleaned and is just too dirty and hazy to use anymore. Especially at night. Now I have long suspected that my crappy mk1 copy of the 18-55 is decentered, and after using it again for a while it certainly doesn't seem right. I cleaned it up and it looks a little better optically (the rear element had a smudge I never noticed), but I'm still not very happy with it. Where it is sharp, it is indeed pretty sharp, but at 55mm and 18mm, the sharpness shifts from good on one side to awful on the other. It moves around too, so something internally might be loose. I got the same result with my k-7, so its just the lens. Oh well.

I've been using an FA 28-80 and found it to be pretty sharp stopped down. I think this lens might not be perfectly aligned either because sharpness shifts to slightly worse on one side. My real problem is this though (and I apologize for rambling):

In using this lens at 28mm, I find that the k-5 wants to focus beyond infinity. It has a hard stop at infinity, which may not be perfectly accurate, but the k-5 keeps wanting to turn beyond that. My 18-55 goes past infinity slightly so I can't compare. When I put the 28-80 on the k-7 it fires perfectly find at infinity, but on the k-5 it refuses to go. Moving AF to +10 makes it fire more consistently. As soon as I flip it to AF.C it fires every single time. On AF.S it doesn't want to fire. It seemed to have more problems locking focus with the 12-24 I can remember as well. I never thought much of it at the time, but in hindsight, the k-7 was much more reliable. I would send it in, but I'm afraid they are going to tell me that the AF issues are my fault because the camera obviously has damage to the bottom plate on the corner and the lcd screen cover has been popped off, scratched, and stuck back on. I don't really trust the AF on this camera though. Too many times has it refused to confirm focus in semi-decent light. I just don't know if its the lens combo or what. The same lens works fine on my k-7, so that tells me something else might be amiss. The fact that I have to make such huge micro adjustments when I'm getting consistent misfocusing is troubling.

Also the meter seems to overexpose. I'm finding that I use -1/3 to -1 EV constantly. This could just be the k-5 though. In my comparison I was getting similar meter readings with the k-7 on the given scene, so it could just be the lens or whatever. I know certain lenses will consistently over or underexpose. At least its consistent. I apologize for the long message. I'm not the most experienced with DSLRS. I bought the k-7 a little over a year ago initially. I've learned a lot in a year or so. I just don't like this nagging feeling something might actually be wrong. I'm sure someone has been down this road.

BTW, I tried testing my 18-55 with the zeiss circle target on my computer screen and the donut was looking just fine. No massive decentering according to that test. The image circle looks relatively even on full frame too, so I have no idea why it is so unsharp across a random 1/3 of the frame even on a tripod. Its a cheap lens. I'll probably just eventually get a WR version. I do like the sharpness of it where it is sharp. Wish my copy performed more consistently...

[rvannatta]

the thing that I notice about the K-5 on both the focus and the metering is that the results are very different depending on whether you set the control options
for 'spot metering' vs 'area metering' and the focus for spot focus or area average. I give some thought to which setting for each I'm going to use on virtually
every scene. the fact that you never mention either of these controls suggests that perhaps they aren't high on your radar.

[zosxavius]

I usually use multimetering and chimp to watch for highlights. As for focus, I'm generally in center point, but changing to using all the points isn't helping my k-5 lock and fire with this lens. Hmm. Its very possible my focus is just out of calibration. Close objects seem to focus ok, but its hard to test with this lens being so slow. A lot is in focus at 5.6 on the long end and it doesn't focus particularly close. In good light it seems to hit ok close up. Its hard to judge infinity focus too because I'm usually stopped down to F11-F13 on this particular lens where it seems to be sharpest.

edit: I would also like to add that live view fires fine at infinity on this lens. Its only PDAF that wants to keep focusing past infinity. I definitely did not have focusing issues like this with the k-7. The k-5 focus is faster, but it misses quite often in comparison and only gets a good lock on a 2nd attempt where it fine adjusts. I did try blowing out my mirror box pretty well thinking that maybe the af sensor has a bit of dust or something. Hmmm. I'd hate to send it back and end up with a huge bill for a problem that existed since day 1.

[zosxavius]

the thing that I notice about the K-5 on both the focus and the metering is that the results are very different depending on whether you set the control options
for 'spot metering' vs 'area metering' and the focus for spot focus or area average. I give some thought to which setting for each I'm going to use on virtually
every scene. the fact that you never mention either of these controls suggests that perhaps they aren't high on your radar.

Also I don't have exposure locked to the focus point. I know you can use focus points individually, but don't know what you mean by spot or area when it comes to focus. I rarely use spot metering. Sometimes I will flip to centerweighted average, but rarely spot unless I need something really critically metered. I can always make EV adjustments. Most of my subject is usually the whole scene anyways. When doing portraits its a little different.

[zosxavius]

Houston we have a problem!

I realized in full magnification in live view that the fa 28-80 did not look focused at all at infinity. On the k-7 it looked good. Look at these two crops and tell me if you notice a difference. My k-5 is not at the right flange distance on the mount. How the hell does that happen??? The mount is solid and not broken. No wonder I can't focus to infinity on this particular lens!

EDIT: I realized that the differences in apparent magnification are due to different file sizes. The k-5 was only set to 10MP JPEG, so therefore looks smaller. Both frames look about the same either way. Looks like my mount is screwed up somehow or that the sensor isn't the right distance.

[zosxavius]

So I took the mount from my k-7 and swapped it. It was pretty easy. Sure enough my k-5's mount was warped. The problem is that it won't focus to infinity even now. The threads weren't stripped or anything. I really don't know what to do. I can't find any more damage and if I send it in, they are going to blame it all on me, which it may be my fault, it may not be. Pretty bummed now. Here is a comparison shot. The k-5 is on the right. It does show improvement, but still won't get to infinity.

[zosxavius]

It finally dawned on me that it is unfair to compare an unsharpened raw to a jpeg. After the misery of feeling defeated set in, I set my camera to jpeg and took one to compare. I think my problem is actually resolved with the mount ring from the k-7. Now I just need to order a new one from pentax to make my k-7 good again.

Here was my last test in jpeg. Comparing it to the k-7 shot it looks more or less equal and better in some ways.

Attached Images