RiceHigh's Pentax Blog: K-r Focus Varies with Different WB Settings! +_="

Wednesday, November 17, 2010

K-r Focus Varies with Different WB Settings! +_="

See this experiment by this Pentaxian whom has newly purchased a K-r:-

http://www.dchome.net/viewthread.php?tid=953899&extra=page%3D1
(Traditional Chinese, Google Translated English Page Here)

In this reply post, the OP writes the test conditions of his and and the shocking finding/conclusion of his experiment:-

http://www.dchome.net/viewthread.php?tid=953899&page=1#pid11231193

Translated, "On tripod, 2 sec self-timer, with each WB setting 5 shots were taken (3 from infinity to a mid-distance object and 2 from near distance to the same object for the lens), the focus for each WB group are close enough whilst the focuses between different WB groups are very different! So, it is concluded that a (different) WB setting actually changes the focus. If it is not a real problem, I have nothing more to say!"

He also stated that he has been a user of K100D, K200D, K-m and K-x also and has never ever encountered the same problem! He further said that he would go to the local Pentax Customer Service Centre for help when he has the time. But what can I say to him is just: Good Luck, Guy! May God bless You!!

I dare to bet they will "play" him to death till the end of the world without any practical solution nor any real and responsible action that could help the suffered users, as that have been happening and being reported from time to time by our Pentax users, almost daily! >:-[

Comments (25)

Logging you in...

Dashboard | Edit profile | Logout

Logged in as

0 Vote up Vote down

Erika · 747 weeks ago

On the related line:
Does the focus on lower end Pentax DSLR's depends on color temperature of the scene, focus sensor sensitivity to IR or both?

If it does depend on color temperature, then maybe Pentax firmware on K-r is trying to adjust focus (based on color temperature set) in order to avoid focusing errors?
Just a thought.

1 reply · active less than 1 minute ago

0 Vote up Vote down

RiceHigh 110p · 747 weeks ago

Even it is the case, these are all *guess* works afterall, but not any actual *measurement*!

+1 Vote up Vote down

Moritz Schwertner · 747 weeks ago

Those different WB Settings have NOTHING to do with the different Focus settings. It has very much to do with what light source he's using.
A light source which flickers a lot creates different white balance settings in a glimpse of a second, every single camera will struggle with those, I guess!
http://www.moritzschwertner.de/

0 Vote up Vote down

FieldofWheat · 747 weeks ago

I suspect that different WB settings are causing the camera to slightly adjust the AF to compensate for colour bias in the AF sensor.
Once upon a time RiceHigh lambasted Pentax for focus inaccuracy caused by shifts in the colour of the light, now he's at them for correcting the problem.
The solution is, of course, to do what any photographer would do and set a correct white balance for the conditions.
The moronic thing to do is to run some sort of "test" that the camera would not encounter during normal shooting conditions and then put the word problem on the camera rather than the "tester".
We saw the same thing with the DPReview comparison test where the K5 images were just plain out of focus, and then they said the camera churned out soft images.
Proper testing is a very rigorous thing. DPReview seems unable to do it, Rice's Chinese friend seems unable to do it (or at least come to the right conclusion), and I've yet to see a test from this blog that I would consider trustworthy.

0 Vote up Vote down

Zebooka · 747 weeks ago

That's not a bug, it's a FEATURE :)

10 replies · active 747 weeks ago

-1 Vote up Vote down

RiceHigh 110p · 747 weeks ago

This "feature" is just a non-sense. WB can be set randomly at the time of shooting (even the AWB could be wrong), how come the focus should be dependent of that?

If Pentax could not measure the error of AF with their SAFOX non-Plus, just let it be and don't "compensate" it by silly guess works via software trick!! >:-(

0 Vote up Vote down

Anu · 747 weeks ago

If it is a real issue, it is likely to be a feature.

The colour of light does influnce the operation of the AF of the phase detect sensors of all DSLRs. If setting WB (correctly) manually allows for more accurate AF operation, that is a very good feature indeed.

0 Vote up Vote down

RiceHigh 110p · 747 weeks ago

Not "all" the AFSLRs. Only Pentax DSLRs (except the K-7 and K-5) suffer most! It's just an unique problem to most of those Pentax DSLRs!!

My top-of the-line *Pentax* MZ-S and the low-end MZ-30 film AFSLR do not suffer, both neither! And neither do the Canon DSLRs that I have ever used, high or low end regardless, either! So?

0 Vote up Vote down

Anu · 747 weeks ago

No,. ALL DSLRs with phase detect AF. This is the very nature of the phase detect AF system. I seem to remember that Nikon uses information of the euto exposure system as well as wb-information to improve the accuracy.

As you bring film bodies into this, I seriously doubt, you've done controlled pixel peeping studies on them, comparable to the pixel peeping accuracy you may do on digital.

It is a contraint of the PD AF technology that does this.

-1 Vote up Vote down

RiceHigh 110p · 747 weeks ago

If you have a Pentax film AFSLR, mount the same Pentax AF lens on it, compare to a Pentax DSLR. Focus things under white and yellow light sources, read the distance scale for the discrepancy, you will find that the film camera always does better!

0 Vote up Vote down

Anu · 747 weeks ago

I do not need to make a test to see if the earth is round or flat. You've been teached about PD AF before in Pentaxforums (years ago), yet you insist on not thinking, or maybe you have given a thought, but just insist on following your agenda according which every Pentax is inderior to it's predecessors.

0 Vote up Vote down

RiceHigh 110p · 747 weeks ago

So, you're the one who have been telling the Earth is flat, but never admitted and could face any evidence showing that the Earth is round actually!

0 Vote up Vote down

skina · 747 weeks ago

What happen to your reasoning process..

In above article it is written "... he has been a user of K100D, K200D, K-m and K-x also and has never ever encountered the same problem!". But here you write "Only Pentax DSLRs suffer most".

K-r versus the whole other Pentax DSLR, isn't it?

0 Vote up Vote down

RiceHigh 110p · 746 weeks ago

Facts:-

1. Pentax DSLRs suffer from Yellow Light Front Focusing AF Symptom than any others including their own film AFSLRs;

2. Pentax tries to do software tricks to compensate the error of the Yellow Light Symptom but it is proven to be failed but just more inconvenience induced!

Clear enough this time?

0 Vote up Vote down

Anu · 746 weeks ago

Evidence please. Simply saying that you know is not evidence.

Fact: PD AF is problematic in different light conditions and with different lenses used. Maybe the lenses you used on film bodies had less sphrical aberrations (maybe they were less ambitious designs) than the one's you use on digital. Or maybe you just don't know how to properly analyze things as it would take all the fun out of Pentax-bashing :)

0 Vote up Vote down

j viviano · 747 weeks ago

It's not guess work, it's telling the camera how to interpret the light.

Lots of older cameras have trouble with auto-focus under tungsten. Here, pentax has implemented a solution to that problem, by recognizing when you are, in fact shooting under tungsten light.

6 replies · active 746 weeks ago

-1 Vote up Vote down

RiceHigh 110p · 747 weeks ago

How about if the user shoots RAW and adjust the WB later??

0 Vote up Vote down

Lurker · 747 weeks ago

Then it will blur the output to simulate focusing :D

I can't wait to get a K-r

0 Vote up Vote down

Anu · 747 weeks ago

If the camera takes the users WB setting as a hint for the AF system, it is the fault of the user if he chooses to use a wrong WB setting. If one uses autoWB, the camera should not do any special adjustments for the AF. Now, I don't know what and how and if Pentax does anything regarding this. I do know however that 99% of tests that are supposed to present a flaw or error are actually misconducted experiments.

Oh, btw, I think it is a bit silly how you always tend to put a "thumb up" to your own posts...

0 Vote up Vote down

RiceHigh 110p · 747 weeks ago

All the "arguments" that you have made are all meaningless, as only the final results are counted, i.e., the test pictures shown by both new K-r users are mostly Out of Focus, no matter how!

0 Vote up Vote down

Anu · 746 weeks ago

Maybe they should calibrate their lenses.

0 Vote up Vote down

Anu · 746 weeks ago

I finally checked the images and see nothing wrong. He is using a fast lens wide open - the fluctuations are perfectly normal. And not just for Pentax, but for all the cameras. This is f/1.8 after all and that is beyong the accuracy of today's PD AF systems.

0 Vote up Vote down

mike · 747 weeks ago

I too did a test under tungsten, although not seriously. i think the bf/ff under tungsten did exist!! Wish it can be fixed by firmware.
http://www.pentaxforums.com/forums/pentax-k-r-for...

0 Vote up Vote down

mike · 747 weeks ago

btw, I also tried with K-m, which in general did better result than K-r.
But setting K-r with tungsten wb and af adjust -5 looks more precisely.

0 Vote up Vote down

Chris · 682 weeks ago

I've suffering this colour/focus issue since I bought my K-r. I love wide open bokeh shots, which this tends to effect the most. Going to try to send it to Pentax for a second time to see if they can fix it.

Post a new comment

Comment as a Guest, or login:

Tweet this comment

Connected as (Logout)

Email (optional)

Not displayed publicly.

Submit Comment

Subscribe to

Comments by IntenseDebate

Subscribe to: Post Comments (Atom)