[biz-engineer]

I'm talking out of focus soft. This is LV focused on the tree directly in front and center, sitting on the back of my car, with a 2 second timer.

Yes. it's out of focus. Typically a tree that is far about 10 meters from the camera, but with the AF ring not yet to infinity. And even I even have my Tamron 70-200 that off focus with live view, but in sharp focus when focusing with the PDAF via the view finder... So, LV AF, that the except that prove that LV AF isn't working every time with every lens.

P.S: I have the same problem with my Tamron 70-200 when used with K-3 LV AF, but the same lens is working fine with the K-5 LV AF.

[mee]

Well, the crux of your original statement was that the LV shot was softer than something shot using the OVF. I too was curious to see a comparison. Also, I didn't see you mention it but what lens and is this with AF or MF?

DA* 16-50mm at f/4 AF

Not sure that really matters though.. it is now obviously a defect of the lens and will have to go back to Precision.

[biz-engineer]

Not sure that really matters though.. it is now obviously a defect of the lens and will have to go back to Precision.

Before you send anything for repair, if you can, try another lens and the same lens with another camera body to see if what needs a fix is the lens or the camera or both. Here in Europe, the repair service is asking to send both camera body and the lens.

[mee]

Well this one just came back from repair yesterday. So it should be on their dime now. But good idea to confirm it to them with another body.

[colonel00]

Ah, this was the one you just got back from repair? What was the original issue it was sent in for again?

Edit: Posted at the same time. Just as backup, I would document some test shots of both LV and OVF and pass the info along to Pentax as well showing them the shoddy repair.

[mee]

Yep this is the one. The original issue was simply lack of focus at infinity.

But yes more test shots are planned.. especially if the weather behaves (forecasted rain). Doesn't look like Precision is getting back with me this week.. considering there is only 1 work day left and I'd need to get the mailing slip from them and rebox it.

The work order says as followed:

Explanation of Repairs

LENS AND ASSOCIATED ITEMS REPAIR
A COMPLETE REPAIR AND RETURN ALL FUNCTIONS TO FACTORY SPECS.

AUTO FOCUS TEST, CHECK, RECALIBRATION AND REALIGNMENT
FIRMWARE DOWNLOAD: CORRECTIONS AND ADJUSTMENTS; ADDRESS VALUES MEASURED AND ADJUSTED
COMPLETED CLEANING (INTERNAL-EXTERNAL) AND OPTICAL SYSTEM


Description of Parts Used

MOTOR
LENS PC BOARD A ASSY

I know they must have actually replaced the SDM as it now squeaks when it autofocuses.. not sure about the rest.. but considering how it behaves in LV I'm guessing they really must have altered the firmware and did something to it.

Was more hoping for some pointers like 'press the menu button the OK button and the play button with your foot 2 feet in the air to reset this or that to get it all to calibrate' hahaha

With the lack of answers, it is more looking like it is still broken.

[colonel00]

I'm just wondering if all they did was to calibrate it for accurate PDAF which threw off the CDAF and this wasn't ever tested. I don't even know if this is possible but I don't see why it wouldn't be.

[luftfluss]

It isn't a one time event.. this is just a representation of how LV works with this lens currently... it zooms in on the LCD (red square on the point it is focusing) and it isnt even sharp there.. it is like it thinks it is in focus but it isn't even close!

through the VF it is noticeably sharper.. still softer than it should be but improved.. through LV it isn't even close!

Looks like Precision has managed to find a new way to break things haha. I was hoping there was an easy fix for this.

Have you tested with a subject that is close, like 10 feet away? I'm wondering if it's an infinity-focus issue. I had a DA21 with that problem.

[UncleVanya]

Did you happen to backup your lens files before sending it in for repair?

[mee]

Have you tested with a subject that is close, like 10 feet away? I'm wondering if it's an infinity-focus issue. I had a DA21 with that problem.

Well that was the issue originally. I'll have to take the tripod out and try it at various focal lengths and apertures in bright light asap.

---------- Post added 12-11-15 at 04:12 PM ----------

Did you happen to backup your lens files before sending it in for repair?

haha you know it and I've pondered that thought too.. but that is another last ditch item as I don't want to tamper with anything while the repair situation is still ongoing.

[RKKS08]

through the VF it is noticeably sharper.. still softer than it should be but improved.. through LV it isn't even close!

And this may hint to the cause of the problem.
PDAF may get it better than contrast measurement from the sensor, if the lens itself has issues to deliver a sharp image.

[UncleVanya]

Well that was the issue originally. I'll have to take the tripod out and try it at various focal lengths and apertures in bright light asap.

---------- Post added 12-11-15 at 04:12 PM ----------



haha you know it and I've pondered that thought too.. but that is another last ditch item as I don't want to tamper with anything while the repair situation is still ongoing.

My point is that you could back it up again and do a file compare. If they differ you could restore yours from before and test it then put theirs back and test it again and compare.

[mee]

My point is that you could back it up again and do a file compare. If they differ you could restore yours from before and test it then put theirs back and test it again and compare.

Yes, that is precisely what I thought you were getting at.

And, it is for a last ditch effort. I'm not playing with the lens in that manner and risking failure unless I have no other options to prove it is still problematic.


Hopefully tomorrow will be bright enough that I can take the lens and a tripod out (along with another body and my 18-135mm) and test each at comparable focal lengths and distances.

The strange part to all of this is my lens claims it is in focus -- it gives the focus confirmation beep -- even when it is most obviously not (such as the LV shot above).

However, just playing with it today hand holding briefly before the sun went down.. it is still quite sharp close up through the OVF. Couldn't accurately try handholding while using LV -- I'm not that talented I'm afraid.

[mee]

Okay I found a non-rainy window of free time to try this out.

I brought along:

tripod
K-5II
K-r
16-50mm
50-135mm
18-135mm

I set up by the pool aimed at one of the brick pylons along the fence line aimed directly at it roughly 95 feet away. Continually set the distance ring to closest focus and then let the camera take care of focusing. Outside of a few tests where I let it focus then half pressed the shutter button repeatedly to see if it would dance to a different focus distance.

I shot at f/2.8 for both of the DA* then at f/4.5 again to test between the 18-135 and the 16-50 (since that is min ap for the 18-135).

The results were confusing to me.

The 16-50mm with the K-5II is not consistent. Sometimes it locks on correctly, other times it thinks it has locked on yet is very out of focus. And the distance scale/focus would indeed dance if I keep pressing the shutter button halfway using normal OVF. Interestingly, I did not notice this behavior with the other lenses using the K-5II; The 18-135mm and 50-135mm on the K-5II locked on each and every time. No dancing, no misfocusing, no nothing but locked on 'sharp' images.

Looks like the 16-50mm is the issue right?

But here is where it gets confusing to me...

On the K-r, all 3 locked on each time. ?!? I need to test that some more to double confirm. But I do not think the 16-50mm is doing the same bad things with the K-r

Issue with the K-5II? Issue with the 16-50? Issue with both of them? Gremlins?

---------- Post added 12-12-15 at 01:03 PM ----------

Btw it was around 12pm locally (noon) when I did this on an overcast day if that matters at all.

[mee]

Hold the phone, joan! There was a flaw in my test.

The K-5II was set to Contrast AF for Liveview while the K-r (I just discovered after going back again) was set to Phase Difference AF!

After I set the K-r to Contrast AF and tried again (same spot) with the 16-50, it is all over the place with what it thinks is correct focus too (at 95feet away from the flat, brick pylon).

To cross check, I configured the K-5II to Phase Difference AF and it no longer bounced around infinity with this 16-50.

So what does this tell us?

---------- Post added 12-12-15 at 02:40 PM ----------

Full sized example shots on the K-r:






Both shot on a solid tripod in Contrast AF, single focus point (aimed straight at the brick pylon front and center), 16-50mm @ 50mm f/2.8

Push the shutter half way, until I hear the camera beep, then press the shutter button and wait for the 2 second timer to click off a shot. Then redo back again. And mega blurry. The closest lounge chair seems to be almost there.

This is what it does on both cameras.. hit or very miss. And even when it does hit, does that look really sharp for infinity at 2.8 with this lens? Still seems rather soft to me..