[NorrisTudor]

I know this forum has loads of threads on the focusing issues. And I having had a K10d and now recently moved to a K20d am an expert in how to measure these issues using Focus Charts etc...

I sent back my k20d and just received a replacement as it was seriously front focusing. (using my 50mm @f1.4)
Now my replacement has the same problem, and I can't believe this is just bad luck. I am really fed up of the whole Pentax experience.!

Anyway...so I've checked my technique. 45 degree angle, good shutter speed, focus chart is flat, just focusing on the black line etc...but still the same problem.
So the other thing I thought about was the lighting. I was always doing my tests in my study in the evening after work in tungsten lighting.
After several googles on SAFOX and Tungsten lighting it appears that it does front focus under this colour temperature.

This weekend I will recheck doing the tests in daylight...however it leaves a dilemma. Does this mean you can never guarantee a good focus indoors under tungsten???

My question to the forum..is if I put a colour correction filter on the lens for indoor shooting..would that fix it?..or is it todo with the frequency of the light?

Any feedback on these thoughts would be great.

Paul
Flickr: NorrisTudor's Photostream

[axl]

I bought K10d recently, tested all my lenses with it and focusing chart, following all the instructions to dot. All tests were done under poor light way afther sun went down, in my living room on my coffe table. And guess what, all of them (F50 1.7, DA40, DA18-55, Sigma 105) came spot on, Sigma 10-20 backfocuses but I knew this long time ago and learnt to live with it. So to say, that safox VIII has problem with tungsten... I don't know, never googled it but....
BR

[axl]

PS: let me guess, you googled it and first thing that came up was RiceHigh's pentax blog?!?
(because that was the response I got from my google)

[NorrisTudor]

RiceHighs blog certainly came up...but his links pointed me to the dpreview forums, where there was specific examples.
Thats where the real information came from. Eitherway its a very obvious with the body I have, and its the 2nd one. Before trying another one, it would be interesting to establish this point as fact / fiction.

[RiceHigh]

Forget any "45 degree" testing, shoot real object, with real pattern and texture, a flat one, parallel to the "film" plane, i.e. perpendicular to the optical axis of the lens and see if your cam FF or BF (or maybe it will be okay!).

Good luck and report back!

[NorrisTudor]

Amazing... just did the test at work in daylight...and the focus much better. And with the AF adjustment at -7 it is spot on.

When I get home I will upload pics.

Its very clear to me that the light temperature makes a bigger difference. I'm not clear whether its colour or frequency. If its colour then this should be fixable with a filter yes?

[RiceHigh]

Amazing... just did the test at work in daylight...and the focus much better. And with the AF adjustment at -7 it is spot on.

When I get home I will upload pics.

I think after that you now have confirmed even by your ownself. Its time to temporarily forget my blog. See my this very old experiment and once again the yellow light front focusing symptom is straightly verified!

Pentax *ist DS AF Dependency on Kelvin & Exposure Value

Of course, this holds universally true with all my *ist D bodies (3 numbers) DS (3 numbers) and K100D which I have ever owned. And after that my blog articles point to numerous K10D and K20D users who reported the same. So, the conclusion is clearly out.

[janosh]

Not meaning to challenge the "Kelvin" explanation, I do want to point out that no fluorescent light has the color temperature it claims.... it can measure as daylight using the best meters, but it's doing a constant flicker. Slide film exposed with it may look and serve OK with casual observation...but not with demanding observation.

With K20D, I've found that daylight fluorescent creates deceptive sensor dust indications that vanish when I use tungsten or real daylight.

[RiceHigh]

Not meaning to challenge the "Kelvin" explanation, I do want to point out that no fluorescent light has the color temperature it claims.... it can measure as daylight using the best meters, but it's doing a constant flicker. Slide film exposed with it may look and serve OK with casual observation...but not with demanding observation.

I wish to point out that my above light sources are far more accurate and real than one could imagine! :-)

Tungsten lighting was using tungsten bulbs! Whilst the 3300K and 4800K light sources are Panasonic highest grade high frequency desklamps which are specified to have 10,000 Hz of refresh rate, per second. So, they will not flicker at 50Hz afterall (and neither the tungsten as they are just like continuous light owing to the slow deaddown time, even they are 50Hz).

[NorrisTudor]

Ok here is pics to show the case:






3rd one is under daylight



Last one is under daylight with -7 correction dialed in.


To me perfect. 1/3 front, 2/3 behind.

So still the question is will a colour correction filter change the situation. Not sure what to get to try this yet.

[Marc Sabatella]

Ok here is pics to show the case:
To me perfect. 1/3 front, 2/3 behind.

Actually not. That's an oft-repeated oversimplication. From what I've read, the actual correct ratio depends on focal length and subject distance, but is far closer to 50/50 than generally assumed. So I think your original picture probably shouldn't have been messed with.

BTW, I don't know if the first picture you posted was cropped, but you didn't have the focus target centered. The camera was likely focusing on the side scale, not the target. As posted, that picture didn't prove a thing. But it is true that under tungsten light, focus is sometimes a bit forward.

[NorrisTudor]

Didn't know about the 50/50 so thx for that.

As for the 1st pic...its one of about 100 taken all sorts of ways. The reason I chose this one is it shows the difference more visibly against the Ruler. But yes most of them were taken with the black line central - and still had the problem.

[dkittle]

Alas, I have the same problem with my K20D, it front focuses very badly with both AF and manual focus. I've currently got it out with Pentax for servicing, hoping that they can at least reshim the damn thing and fix the manual focus issue. For AF, I can dial in adjustment and get every lens working except for my FA50 1.4 - even at -10 it still front focuses quite badly. I've repeated the test 4 times now, always under tungstan.

I just bought a used K20D and it has some focus issues but not quite as bad. It front focuses with manual focus and is out a little either way with many AF lenses. The AF doesn't bother me that much as we can correct it (most times), its the manual focus being out that bothers me most.

I might go retry this test in daylight and see what results I get.

[NorrisTudor]

Definitely go test - I'm amazed at the difference. At least now I know, and can workaround it. Before this - I was really disappointed as I couldn't trust the AF.
Everything else about this camera is so superb.

[dkittle]

Well, many AF lenses later, here are my observations.

The lenses that I 'calibrated' under tungstan needed no further adjustment when I shot under daylight.

My lenses needed both positive and negative adjustments.

My Sigma 70mm macro and 30mm lenses are both seen as the same lens. Thankfully a -8 or -10 adjustment focuses them both fairly well. I remember my other body (currently in for repair) also needed a -8 adjustment to correct the Sigma 30mm lens.

My Sigma 100-300 still front focuses a bit even at -10. In real world shooting, though, it's plenty sharp even wide open. Guess my targets are always 10' or further away.

Even my 77mm needed some adjustment (I suspected this back in my K10D days), but my 31mm is spot on.

Measured distances with a tape measure so I was very close to always having a 45 degree angle to the chart.