[saladin]

just been playing with my new K5 for two days. To start with, i was disappointed. I was getting better results from my k10, it looked like the 10 was out-resolving or "sharper" detailed than the 5, which seemed ridiculous unless i have a fully defective K5.

after much head-scratching and thoughts of sending it back, i finally twigged that as supplied the camera was on firmware 1.01. I loaded in 1.12 and that seemed to improve matters greatly. Since doing that, i have tried the AF fine tuning and found that both the DA 18-55 and the 150-500 sigma were still definitely back-focusing. I've dialed in +6 for the DA and +4 for the sigma. now these are both indoor paper results, so i need to see how it goes in the real world.

I think the camera is now giving results that i expected. It may also be that the DA (it's a series one) simply isn't good enough for the k5 and i need a much better lens to do the camera justice? possible that the camera is looking for more detail than the lens can provide?

either way, the AF fine-tuning seems to have helped. but i suspect the firmware update was the big improvement for IQ

[aleonx3]

just been playing with my new K5 for two days. To start with, i was disappointed. I was getting better results from my k10, it looked like the 10 was out-resolving or "sharper" detailed than the 5, which seemed ridiculous unless i have a fully defective K5.

after much head-scratching and thoughts of sending it back, i finally twigged that as supplied the camera was on firmware 1.01. I loaded in 1.12 and that seemed to improve matters greatly. Since doing that, i have tried the AF fine tuning and found that both the DA 18-55 and the 150-500 sigma were still definitely back-focusing. I've dialed in +6 for the DA and +4 for the sigma. now these are both indoor paper results, so i need to see how it goes in the real world.

I think the camera is now giving results that i expected. It may also be that the DA (it's a series one) simply isn't good enough for the k5 and i need a much better lens to do the camera justice? possible that the camera is looking for more detail than the lens can provide?

either way, the AF fine-tuning seems to have helped. but i suspect the firmware update was the big improvement for IQ

Did you try turning off that option in the menu? I found that turning AF adjustment to OFF puts every (almost in my collection) Pentax lens without any adjustment.

[fuf]

Did you try turning off that option in the menu? I found that turning AF adjustment to OFF puts every (almost in my collection) Pentax lens without any adjustment.

I have tested this idea with the only lens in my collection which needs correction (FA31) and noticed that turning the adjustment off yields worse results (consistent slight backfocus). From the technical point of view, I don't see any reason why turning off the adjustments would improve the AF in any way. If the body/lens contains various information on AF then in my opinion it makes sense to make the AF adjustment additive, i.e. added to that factory information and not 'overwrite' it.

[octavmandru]

I have tested this idea with the only lens in my collection which needs correction (FA31) and noticed that turning the adjustment off yields worse results (consistent slight backfocus). From the technical point of view, I don't see any reason why turning off the adjustments would improve the AF in any way. If the body/lens contains various information on AF then in my opinion it makes sense to make the AF adjustment additive, i.e. added to that factory information and not 'overwrite' it.

I had the same idea, but additive would make OFF = Zero adjustment.
But I found that microadjustment ON @ zero is very different from microadjustment OFF.

If lens microadjutstment value are not absolute but relative to a body, than the microadjustment values are not transmisibile.
Think is logic to work with absolute values (on both camera and lens) and not relative values.
In this way a manufacturer can calibrate a lens (via the lens chip) in the factory and that lens will be useable to multiple bodies.

[fuf]

If lens microadjutstment value are not absolute but relative to a body, than the microadjustment values are not transmisibile.
Think is logic to work with absolute values (on both camera and lens) and not relative values.
In this way a manufacturer can calibrate a lens (via the lens chip) in the factory and that lens will be useable to multiple bodies.

Why should the adjustment values be transmisible? Both the lens and body have their own tolerance levels and must be calibrated together. The same lens which focuses correctly with one body might misfocus horibly with another one. If you encounter this case you need to have calibrated the lens and all the bodies.
As for your second statement, as bodies have their certain tolerance limits so do the lens. Thus most lens should work with most bodies but in certain combinations misfocus can occur. The following article explains it quite well: Variation Facts and Fallacies: Digital Photography Review .

[adi3000]

I repost this, maybe you missed and is good to try it.

Quantitative Method for Micro Focus Adjustment for DSLRs - Enzo's Home

I tried and it works very well, unexpected. I have a DA* 16-50 and I been unhappy with it even if I adjusted the focus before. But using this method give me another setting better for every situation, even in low light now is perfect.

- in my situation the compiled class was not working. I use only the .java file and use "compile and run" (you need jdk for this).

[octavmandru]

Why should the adjustment values be transmisible? Both the lens and body have their own tolerance levels and must be calibrated together. The same lens which focuses correctly with one body might misfocus horibly with another one. If you encounter this case you need to have calibrated the lens and all the bodies.
As for your second statement, as bodies have their certain tolerance limits so do the lens. Thus most lens should work with most bodies but in certain combinations misfocus can occur. The following article explains it quite well: Variation Facts and Fallacies: Digital Photography Review .

Of course the adjustment values MUST be transmisible! If lenses are thrown on the market without calibration, and cameras are sold without the calibration done, we would have a good match in a million!
I have read that article previously, but the explanation doesn't satisfy me enough.

-It is much more cheaper to make equpiment with a wider tollerance and then write on it's chip ... +5, instead of aiming for perfection.
Camera will take the +5, add it to it's calibration, say -2, and adjust the focus to +3. (this happens when the microadj function is OFF)
-I'm not arguing that, in time, the equipment won't fall off it's initial calibration value. Then, the Microdajustment becomes usefull.
-Don't forget that the microadj. function comes from the factory tuned OFF.
-I don't pretend that ALL my problems with lenses are solved since turning the microadj. to OFF. Just the accuracy of the autofocus.
-I'm not going to say ALL new cameras owners should turn the function OFF. For me it worked, since most of my equipment is new, with little use.

Finally, how to explain that not only me but some other forum members got better results with their own lenses by turning the microadjustment OFF?

---

Last edited by octavmandru; 02-20-2012 at 05:42 AM.

[peterh337]

I am sorry to ask a dumb question but how can you get "bad focus" with a DSLR?

Surely the AF system works by looking at the spectrum of the image and iteratively adjusting the focus so as to maximise the high frequency components, which corresponds to the sharpest image.

If e.g. one was using ultrasonics to measure the distance to the target and then adjusting the lens according to some lookup table, I can see that a lens which is a bit "bent" might give wrong focus, but surely that is not how DSLRs work?

Anyway, the K5 has the live view mode which *has* to work using the spectral method - I cannot see any other way.

I can understand the camera doing this poorly with the "wrong kind" of light, however.

I have tried testing the focus on my K5 and cannot make it go wrong, but admittedly I have been doing it in reasonable light levels.

The firmware is about 1 year old - the camera is a 2nd one after the 1st one went back due to the "string of pearls" contamination, and the two adjustment wheels not working properly.

[fuf]

Of course the adjustment values MUST be transmisible! If lenses are thrown on the market without calibration, and cameras are sold without the calibration done, we would have a good match in a million!
I have read that article prevoiusly, but it the explanation doesn't satisfy me enough.

Adjustment +/- X on a lens doesn't automatically mean that the lens will work on another body as the adjustment either might not be needed at all or be entirely different (even off limits). If the solution to disable the adjustment worked for you, then it is good. But to create a sound of it working for almost everyone is way too much, sorry.
Similarly, you will always find people find certain unreported issues both fixed and introduced with new firmware. I would expect the same in case of their findings with AF accuracy .

[fuf]

I am sorry to ask a dumb question but how can you get "bad focus" with a DSLR?

Please see Understanding Camera Autofocus or Canon EOS Autofocus - BFCV and micro-adjust (MA) - Open Photography Forums . Simply put there are many variables which come into play and if they don't work in unison, problems occur.

[octavmandru]

Adjustment +/- X on a lens doesn't automatically mean that the lens will work on another body as the adjustment either might not be needed at all or be entirely different (even off limits). If the solution to disable the adjustment worked for you, then it is good. But to create a sound of it working for almost everyone is way too much, sorry.
Similarly, you will always find people find certain unreported issues both fixed and introduced with new firmware. I would expect the same in case of their findings with AF accuracy .

Sorry, I think you may have misunderstood
I mean :
-the factory correction value is there to bring the lens as close to perfection as possible. This is written in the lens and taken into account by the camera IF the microadjustment is OFF.
-If ON, then the camera will ignore the factory correction value, and use the user value.

In this way a regular user does not have to fiddle with the settings (leaving the feature OFF), while another one with issues can turn it ON, and adjust.

Remember this: Default is OFF.

Sounds reasonable, no?

[fuf]

-the factory correction value is there to bring the lens as close to perfection as possible. This is written in the lens and taken into account by the camera IF the microadjustment is OFF.
-If ON, then the camera will ignore the factory correction value, and use the user value.

In this way a regular user does not have to fiddle with the settings (leaving the feature OFF), while another one with issues can turn it ON, and adjust.

Sorry, but this wouldn't make any sense if implemented as you suggest. The factory correction value corrects the lens in such a way so as to fall within the factory tolerance limits. Microadjustment helps negate the factory tolerance limits of *both* the lens and body combination.

[jmg257]

Huge disappointment from my FA 77. +20 didn't seemed enough. At this point I started thinking I must be doing something wrong.
Anyway, the adjustments didn't seem to make enough difference

Wondering how you set "+20" - or are you inferring such a setting wouldn't be enough based on tests with a +10 setting?

[octavmandru]

Wondering how you set "+20" - or are you inferring such a setting wouldn't be enough based on tests with a +10 setting?

+10 on the lens + 10 on the adjust all...

[octavmandru]

... Microadjustment helps negate the factory tolerance limits of *both* the lens and body combination.

What do you mean by "negate"?