Over the years, I've accumulated quite a number of Pentax (and Samsung-branded Pentax) bodies and lenses, and with every new body or lens I've added to my collection, I've needed to carry out AF fine adjustments to ensure accurate auto-focusing. Whilst doing so, I've learned some obvious similarities and differences in adjustment for various bodies...

With my K-5, K-3 and K-3II, the minimum AF fine adjustment step is +/-1. A +1 adjustment on these cameras brings the focus distance ever so slightly closer to the camera - hence, it's used to correct for back focusing - whilst a -1 adjustment pushes the focus distance slightly further away, correcting for front focusing. With my K10D, GX-10, *ist DL and GX-1L, the minimum adjustment is +/-10, and the direction of adjustment is reversed - so, a +10 adjustment pushes the focus distance further away, whilst a -10 adjustment brings it closer.

As I carried out adjustments for each of my body and lens combinations over time, I gradually noticed a pattern or "relationship" emerging in the settings...

For my K-3, my DFA100/2.8 lens required zero AF fine adjustment, whilst my GX-10 (well, one of them ) required +110. This was my "baseline" for both bodies. For my FA50/1.4 and DA35/2.4, the K-3 required adjustments of +2 and -3 respectively, whilst my GX-10 needed +90 and +140. So, relative to its baseline adjustment of zero for the DFA100, the K-3 required 2 x +1 steps to correct for the FA50/1.4 and 3 x -1 steps for the DA35/2.4... and, relative to its baseline +110 adjustment for the DFA100, the GX-10 required 2 x -10 steps to correct for the FA50/1.4 and 3 x +10 steps for the DA35/2.4. I'd found that each body required the same number of front or back-focus adjustment steps to correct for each lens, relative to the baseline settings.

I analysed all of my preferred AF fine adjustment settings for both the K-3 and GX-10, and found that in the majority of cases, I'd selected exactly the same number of steps from the baseline to correct each body / lens combo. In a couple of instances my settings were out by +/-1 for the K-3, or +/-10 for the GX-10, but the differences really were wholly insignificant. I then analysed my preferred AF fine adjustment settings for my K-5 and confirmed that the same number of steps were required, relative to a baseline with the DFA100.

At this point, I concluded that, given a baseline adjustment for each body with one reference lens, and the necessary adjustments for a reference body on all lenses, the adjustments for every other body and lens combination could be easily calculated.

I rigged up a rather ugly spreadsheet for this purpose and have been using it successfully for some time. Then, a few days ago, I decided to tidy it up and make it available for other PentaxForums members' benefit.

You can download a copy of this tool from my Google drive, HERE.

The tool was created in LibreOffice Calc, but saved in Microsoft's XLS format. It's a very simple spreadsheet, using basic formatting and cell calculations, and no macros. As a result, it should work in just about any spreadsheet programme capable of handling basic XLS files. The workbook contains two spreadsheets in different tabs. The first is unpopulated and ready for use. The second is an identical copy populated with some of my own example data (camera serial numbers changed) to demonstrate how it's used. Both tabs are protected (without any password) to prevent over-writing of calculated cells, but can be unprotected for further alterations and development as required.

I provide this tool free and without copyright. I offer no formal support, though I'd be happy to address any queries here in this thread or via PM. You're welcome to modify the tool, but if you do so, please consider offering your improvements to the PentaxForums community (via a separate thread)... and if you do, please consider crediting me for my part in its development (it's not essential, but would be greatly appreciated).

Instructions for use:

Hopefully, it's fairly self-explanatory.

The yellow cells are for user input. Everything else is fixed.

Let's say you own three bodies and five lenses.

Choose the most recent of your bodies (since it has the best AF system) and enter the model and serial number (or unique ID) in cells C4 and C5 respectively. In cell C6, enter the minimum step required to correct for back focus on this particular body (on a K-3, this would be +1... on a K10D, it would be -10... etc.). This is your "reference" body.

Choose a "reference" lens from your collection. I recommend a reasonably fast model that's able to produce quite shallow depth-of-field when photographing an AF test chart. Wider angle lenses tend not to be good choices here... I'd pick something like a 50mm f/1.4 or 100mm f/2.8, but whatever you have that gives shallow depth of field is fine (if you're choosing a zoom lens, set it to the longest focal length). Enter a description for the lens in cell B8.

Next, carry out your AF fine adjustment for this body and lens combination. When you're satisfied with the result, enter the resulting adjustment figure in cell C8.

Now, do the same for each additional lens, entering the relevant description and adjustment setting in columns B and C of subsequent rows.

At this point, you are ready to enter details of your additional bodies - camera, serial number and adjustment step - in rows 4,5 and 6 of columns D onwards.

Finally, carry out AF fine adjustment for your reference lens on each additional body and input the resulting figures in row 8. As you enter each one, the calculated adjustment figures for each lens will appear in the relevant camera's column.

Notes:

• At the time of writing, I own a K-3II, K-3, K-5, K10D and no less than three Samsung GX-10s (K10D clones). I am confident that the approach and calculation method I've used works well for me on all of these bodies with each of my lenses. I've also done limited testing in the past with my GX-1L on a small subset of lenses and believe it works here too, but can't be certain (I now have my GX-1L set to an average baseline that works well enough with all lenses due to the limited resolution, and don't have a record of my previous settings).
  
• I am certain that for all bodies from the K-5 through K-3II (and every body produced between these releases), AF fine adjustment is in increments of +/-1, positive to bring the focus distance closer, and negative to push it further away. I am similarly certain that for the K10D / GX-10 adjustment is in increments of +/-10, negative to bring focus closer and positive to push it back. I am fairly (though not 100%) confident that the K20D and earlier bodies all follow the same +/-10 convention as the K10D. For other bodies not specifically mentioned here, an element of investigation and/or experimentation will be necessary to determine the size of the adjustment step and the direction of adjustment.
  
• There would seem to be differing opinions on whether the linear relationships I've described for AF fine adjustment work as described. I can only say they work for me across the bodies I own - well enough that I now depend on the calculated figures. However, if you try the tool and find it doesn't work for you, please disregard it - though do let me know, just in case I can help.
  
• This isn't a tutorial on the process of AF fine adjustment, so please read up on that if you're not familiar with it. I will say that a suitable target, sensible target distance and good, natural lighting (not fluorescent) are key to obtaining good results. Furthermore, don't be surprised if, in real world shooting, AF accuracy is slightly "off" even after careful adjustment. Real subjects are typically more challenging for the camera than a flat AF chart, and lighting isn't usually so well-controlled. Auto-focus systems do a pretty good job in the right circumstances, but they're not miracle-workers - so don't expect perfect AF accuracy in every shot.

Finally, if you have any comments or notice any problems with the file, please let me know and I'll do my best to address them. I'll update this main post and the downloadable xls file accordingly.

Once again, the tool can be downloaded from my Google Drive here:

BigMackCam's Multi-Camera AF Fine Adjust Calculator

ENJOY!

[stevebrot]

What I was trying to convey is that AF adjustments on the K10D are in +/10 increments, with negative increments bringing the focus closer and positive pushing it further back, which is the opposite way round to the K-3's +/-1 increments. I've updated the wording to make this a little clearer.



Yes, AF fine adjustment for the K10D / GX-10 is a global setting performed through DEBUG (same with the *ist DL and GX-1L that I own). I leave the debug file on the SD card of my main GX-10 so I can start DEBUG mode in the field, and carry a printout of the adjustments in my bag in case I should want to change lenses.



Are you referring to "Apply One" vs "Apply All", or have I misunderstood?

Thanks for the clarification. Makes perfect sense now.

Yes, I was referring to "Apply One" vs. "Apply All". A full explanation would be seriously off-topic, so I will let it rest.


Steve

[eminker]

Thank you Mike.

[Mikesul]

Thanks for all the work you put into this, Mike.

I do appreciate the effort in preparing that data. However, I suspect that there's been an assumption made, namely that any K-3 will require the same degree of adjustment for a particular lens. Since both are manufactured products, I see the problem as a matter of matching an individual lens to an individual body, and those are going to have a gaussian curve type distribution of variation.

[BigMackCam]

I do appreciate the effort in preparing that data. However, I suspect that there's been an assumption made, namely that any K-3 will require the same degree of adjustment for a particular lens. Since both are manufactured products, I see the problem as a matter of matching an individual lens to an individual body, and those are going to have a gaussian curve type distribution of variation.

Thanks for the feedback, and please allow me to clear up a small misunderstanding.

I've made no assumption that any body, K-3 or otherwise, will require the same absolute AF fine adjustment for a specific lens as any other body, K-3 or otherwise. My assertion is merely that the adjustment increments (or "steps") for all the bodies I've tested - and, I'm fairly certain, those I haven't - result in the same shift in focus distance; i.e. a +1 adjustment increment on a K-3 results in the same shift in focus distance as a +1 adjustment increment on a K-5, or a -10 adjustment increment on a K10D.

Hence, once you've determined the correct AF fine adjustment for your first camera (the "reference body") with each of your lenses, and the correct adjustment for any additional bodies against just one of those lenses (the "reference lens"), it's possible to calculate the adjustments required for all remaining combinations of bodies and lenses in your possession.

Does that make sense?

...
Hence, once you've determined the correct AF fine adjustment for your first camera (the "reference body") with each of your lenses, and the correct adjustment for any additional bodies against just one of those lenses (the "reference lens"), it's possible to calculate the adjustments required for all remaining combinations of bodies and lenses in your possession.

Does that make sense?

Yes (to me, anyway), thanks.

[swanlefitte]

Yes (to me, anyway), thanks.

If you look at the spread sheet, it has 3 gx10 cameras all with different adjustments. It is also the reason for camera serial numbers, to make sure the right camera numbers for right camera.

[BigMackCam]

If you look at the spread sheet, it has 3 gx10 cameras all with different adjustments. It is also the reason for camera serial numbers, to make sure the right camera numbers for right camera.

Bingo

I think perhaps the confusion might have stemmed from the fact that my K-3 required zero adjustment for the DFA100 reference lens, and my K-3II - by coincidence and rather good fortune - also required zero adjustment for that lens. I'm lucky that both my K-3 and K-3II seem to have been identically and accurately aligned at the production stage. As such, by pure chance, both bodies require the same adjustments. In hindsight, perhaps I should have made up the example data rather than using my own real numbers, to avoid confusion.

The reality is, most folks with two or more bodies - whether they're the same or different models - will observe a small or possibly considerable difference in the necessary adjustments for their chosen reference lens. Such is the case with my three GX-10 bodies, as you rightly noted. But, having determined the baseline adjustments for those bodies with the reference lens, the number of adjustment increments for all other lenses - relative to the baseline - is the same.

[DavidDwight]

Great work and thank you for sharing.

Begs the question why every body and lens delivered shouldn’t be delivered with a calibration “number” / certificate .... this value could be measured as the final QA step before boxing?

When the user receives a new lens all they need to do is cross reference a table/calculation where they input the body’s value and the lens value and it tells them what value to input into the fine adjustment? This should work should it not?

[Serkevan]

I probably should get around to doing this for, at least, the Tokina 70-210/4.5 and Tamron 90/2.5. The DFA 28-105 seems to be in the clear.

[dcpropilot]

Very interesting....thank you for creating this. I'll give it a shot!

[Pentaxians]

Thanks for sharing! Great stuffs!

[aslyfox]

this is one of the greatest benefits of these forums

members willing to share knowledge

thank you very much for doing do