[stormtech]

I've read just about every thread about the new AF system I could find for the K-5IIs. My questions comes to those of you who have gone from the K-5 to the K-5IIs, and who may have had the same problem I've been experiencing with my K-5.

The short version - AF consistency is not so great with my K-5. I guess a proper term might me AF float?

With my first K-5, after about 6 months of use, I was getting soft images little by little. When I sat down and did a focus check for my lenses, I found half of them to not be correct even at a +10 adjustment. I sent that one off to CRIS under warranty. It came back with the service performed "Adjust AF" - that was about it. After that all my lenses were spot on with no AF adjustment or no more that +-1 - great!

I ended up selling that K-5 as I had two at the time, and kept the newer one. Well, guess what - it's been about 6 months and getting the same exact thing now. No problem sending this one off to CRIS, but while this is happening I get to wonder about the K-5IIs. With this new focus system I wonder if this AF float problem has been addressed. From my reading here on the forum, it looked like Ogl and Creampuff upgraded for exactly this reason. Can anyone else chime in about this issue?

Here is a graph showing the inconsistency with my sharpest lens - FA*300 4.5. Notice the green dots - those are each shot within the same AF adjustment (4 shots taken per AF adjustment).

---

Last edited by photolady95; 04-17-2016 at 02:47 AM.

[CB Knipp]

I use the K5IIs and can say that I have yet to have any focus problems with any lens I have used on it, in
any lighting.

[stormtech]

I guess I am the only one who has this problem with their K-5. For me, this is the second one that did the same thing.

After digging around a lot more on this subject, it seems I see a lot of people have needed to do lens adjustments with their K-5's. What really made this stick out for me is reading a thread about the K-5iis where a few folks said they had to do no AF adjustments at all with their new iis's but had to do quite a few adjustments with their K-5's. This tells me that this might be a regular problem with the K-5 and most people don't even realize it.

My second K-5 is on its way back from CRIS with the same work done as my first one - camera was back focusing - adjusted to optimal resolution. I also wonder if they physically adjust the AF sensor of do it via software like using debug mode. Also makes me wonder if or when the AF will "un-adjust" itself again.

But, since there were no replies here about any similar issues, I guess it's just me having 2 K-5's in a row like this.

[bjdavey]

I never had the need to use lens focus adjustment in either my K5 or IIs - both were spot on with all my lenses. The difference has to do with speed of focus, decisiveness, and low light, with the IIs outperforming the K5 in all areas.

[F-pix]

Nice graph!
I would like to try to do the same with my K-5 IIs, from where do you get the sharpness values?

[stormtech]

I never had the need to use lens focus adjustment in either my K5 or IIs - both were spot on with all my lenses. The difference has to do with speed of focus, decisiveness, and low light, with the IIs outperforming the K5 in all areas.

I appreciate your reply in that I had it in my mind that buying a iis might alleviate this problem. I was hoping that another AF improvement might be AF stability in the iis, but certainly can't prove that.

Nice graph!
I would like to try to do the same with my K-5 IIs, from where do you get the sharpness values?

The graph was made with software called Focus Tune. It has been out for a while, but only worked with Canon and Nikon. In the new beta of version 2.0, Pentax is now included. The formal public release should be very soon. Focus Tune is great in that I finally have a controlled way of testing my lenses. The graph output showed me exactly what was going on with my K-5 and also shows you exactly where your AF fine adjustment should be for each of your lenses.

Focus Tune

[One3rdEV]

Stormtech, This is an interesting and relevant topic (especially if the optimum AF fine adjust can truly drift).

In order to help our understanding or ability to further read between the lines on this topic, do you mind clarifying a few points?
- What targeting method were you using to perform AF fine adjust?
- As F-pix mentioned, what technique did you use to qualify the sharpness values in your chart?
- And can you add some more clarification to how you determined that the camera 'was back focusing - adjusted to optimal resolution'.

[One3rdEV]

LOL- looks like you posted while I was writing

[stormtech]

LOL- looks like you posted while I was writing

No problem - if you have any other questions about how Focus Tune works, just let me know. After 2 weeks of pulling my hair out trying different methods to check focus, I really needed a very consistent controlled method to do the testing. I needed to prove to myself that it was in fact the camera and not the lenses before I sent the camera off for evaluation and repair.

Here is a better shot of the ,pdf that Focus Tune generates when you run a set of images. This is the AF Adjustment Analysis - you can also run a test for AF consistency and aperture consistency which will show you at which aperture your lens is performing at its best.

[jamesm007]

I've read just about every thread about the new AF system I could find for the K-5IIs. My questions comes to those of you who have gone from the K-5 to the K-5IIs, and who may have had the same problem I've been experiencing with my K-5.

The short version - AF consistency is not so great with my K-5. I guess a proper term might me AF float?

With my first K-5, after about 6 months of use, I was getting soft images little by little. When I sat down and did a focus check for my lenses, I found half of them to not be correct even at a +10 adjustment. I sent that one off to CRIS under warranty. It came back with the service performed "Adjust AF" - that was about it. After that all my lenses were spot on with no AF adjustment or no more that +-1 - great!

I ended up selling that K-5 as I had two at the time, and kept the newer one. Well, guess what - it's been about 6 months and getting the same exact thing now. No problem sending this one off to CRIS, but while this is happening I get to wonder about the K-5IIs. With this new focus system I wonder if this AF float problem has been addressed. From my reading here on the forum, it looked like Ogl and Creampuff upgraded for exactly this reason. Can anyone else chime in about this issue?

Here is a graph showing the inconsistency with my sharpest lens - FA*300 4.5. Notice the green dots - those are each shot within the same AF adjustment (4 shots taken per AF adjustment).

Read Lens Rentals 'Roger Cicala' blog. He sees and tests thousands of lens. His data shows lens do vary as do bodies; when new and over time that is. The Pentax owners manual has always suggested a checkup at one or two year intervals for peak performance for the body and one year for the lens. Roger tells it better than anyone. Once you read all his blogs you might think twice about measuring your lens other than with your eyes.

The situation is made worse the more pixels we cram into a given area. And the more demands on the lens to hit those pixels with a tack sharp picture.

However I can't say I have ever seen any of my lens or bodies fall so out of whack it was noticeable even pixel-peeping. Pentax did adjust my K20D and that surprised me because it was taking tack sharp pics. I do have a K5 as well its too new to worry about.

Take a read - http://www.lensrentals.com/blog/2011/10/notes-on-lens-and-camera-variation/c...2#comment-2131

He has lots of blogs on his IMO of the test data. Interesting or not, it is valid industry standard test data. Moreover its done on thousands of lens and bodies.

[manishved]

I downloaded the focustune beta but its asking for a key to run it.

[Kerusker]

Stan,
thanks for your efforts. I was/am hesitant to get involved in AF as there are soooo many aspects which may come into play.
Did you find Enzo's 'Quantitative Method' from Jan. 2012?
https://sites.google.com/site/vincenzomiceli/quantitative-micro-focus-adjustment-for-dslrs
Please give more information on your test set up (as already requested by others).
Also, did you find a lens which needs more than a +10 or -10 adjustment?

I just had a K-5 II for one week and did some kind of focus adjust comparison to my K-5.
My finding (with my method) was that both need very similar adjustment values (which may be shifted along the adjustment scale).
The Best

[stormtech]

I downloaded the focustune beta but its asking for a key to run it.

You have to purchase the current public version first. Once you have that paid for and installed, then the beta version will work.

There is a "demo" version you can download, but it only contains some images that you can try the software out with - you can't run your own images through the demo version.

[stormtech]

Stan,
thanks for your efforts. I was/am hesitant to get involved in AF as there are soooo many aspects which may come into play.
Did you find Enzo's 'Quantitative Method' from Jan. 2012?
https://sites.google.com/site/vincenzomiceli/quantitative-micro-focus-adjustment-for-dslrs
Please give more information on your test set up (as already requested by others).
Also, did you find a lens which needs more than a +10 or -10 adjustment?

I just had a K-5 II for one week and did some kind of focus adjust comparison to my K-5.
My finding (with my method) was that both need very similar adjustment values (which may be shifted along the adjustment scale).
The Best

I haven't yet tried the Enzo's method - I was looking for something a bit more simple to try.

And yes, trying to get the AF right with a bunch of lenses can be exhausting, but I have now become kind of obsessed with it. I've become kind of disappointed with the K-5 in this respect as this is my second one which seems to have the AF system in the camera go out of alignment on its own. My K-5 is on its way back from CRIS - the work order shows that the camera was back focusing which they adjusted. Being this is my second time with this, I will have a hard time trusting it to stay put now. I guess I just don't understand why or how this can happen especially since I treat all my gear with kid gloves.

Anyway, here is a brief summary of the setup for taking the sample shots:

Print out the target provided by Focus Tune on 8.5x11 paper and hang on a wall. I put two of my continuous lights pointing at the target in hopes of a more accurate test.
Setup with a tripod - I was working mostly with my 300mm lenses, so the distance from the camera to the target was ~24'. There is a calculator on the website for distance between target and camera for various focal lengths.
Set the camera to jpeg, highest quality, single center point AF, fixed WB, SR off. Shoot the lens wide open.
Use the 2 second timer.

Then you start with taking 4 images at each AF adjustment value. You can start out using every other one, then to really fine tune you can select a range of the best from the graph and do a few at each AF adjustment value. So, I started out taking 4 shots each at AF adjustment 0,+2,+4,+6,+8,+10,-2,-4,-6,-8,-10. Between every shot you de-focus the lens slightly in the same direction each time. That is why I like this test as it makes the AF work for each image taken.

You will end up with 44 images. Then it is as simple as importing those images into Focus Tune, and hit one button, and the software does the rest. In about 30 seconds you end up with the .pdf graph like I displayed above.

[ve2vfd]

Thanks for the heads up on that software Stormtech! I was already using their lens-align scale and had some pretty decent results, but also found I had to readjust my K5d periodically as my focus point seems to drift. This will take a lot of guesswork out of the adjustments.

Pat