[Rupert]

Rupert,
As far as i know, there is no published standard way of testing K5's for low light focusing. I don't think any one would suggest that a 1000 members of Pentax Forums would all test cameras in the exact same way when there isn't a published standard.

For example, see Falk's recent statement above:



The value of Pentaxeros report is that ONE individual tested 100 K5's, presumeably all in the same way.

If either Pentaxeros or some photographer more experienced than i, would write up a test procedure for testing K5 for this issue, and then talk Pentax Forums into making a sticky out of it, that might help standardize discussion of this issue. In too many threads, we get these discussions about how exactly this person or that person tested their camera - a lot of wasted posts in my mind.

best wishes,

You are so correct that we lack any standard test or even a real sense of how many have a problem vs how many don't? How many think the recent Firmware solved their problem/how many don't? I would love to see a thread dedicated to just a few simple questions and responses to a standard test result without all the speculation, rambling and other wasted space information that is of little value to those of us waiting to buy the K5 for.......lack of any real information we can make some sense from.
A thread like that would benefit everyone and could show if this is as serious as we assume, if it is being addressed in Firmware already, and if it is likely to be a barrier in future sales of the K5 if not resolved.
Best Regards

[pcarfan]

There are a few reports of good tests done by users in DPR showing up today with no FF problems and some even saying their issue got better after sending it in for service...yeah! it could be a firmware issue, but having seen and heard of these, it could be a hardware issue, YIKES!

[Smeggypants]

OK, just to be sure, I've ran some tests again...
I have one desk lamp, one white/red cushion and the FA50/1.4... AF assist lamp is disabled.

My K5 seems to focus quite accurately on the red/white boundary as low as 1/80, f/1.4, iso 20000... That -0.3Ev if I'm not mistaken.
At iso 25600 (Ev -0.7?) focusing fails to lock and the green hexagon blinks quickly.

When I first had the problem two months ago, it was while trying to shoot my son, at 1/30, f/1.4, 6400 iso... Ev -0.1, so nearly the same light level...
Only differences are the firmware, and the AF assist lamp is now disabled.

I know I still had the FF problem two weeks ago, with the v1.02.21.07 (I can dig up the Ev value, I still have the shots somewhere).

But since the v1.02.21.10, it seems my K5 was cured...

Intriguing, no?

That firmware version has made no difference to my K-5's FF fault.

[jolepp]

Given what Pentaxeros is telling us and that even problems in hardware can often be worked around in software I would expect a FW update to be the solution. Since it is about a month into the problem (?) it is kind of early yet; 1-2 months from discovering a problem to a solution would be pretty agile from something of the size of Pentax (Hoya), especially if new firmware versions go through formal QA (which apparently has been less than stellar here ).

[dlacouture]

That firmware version has made no difference to my K-5's FF fault.

Yep, you're not the only one to say this. That's why I find my experience so strange, a placebo effect for my AF module, maybe?

Damn, it's almost as if the hardware engineers at Pentax decided to "improve" the AF module, without telling it to the software guys. This would barely surprise me, I've seen it plenty of times in my field of work... Even seen our hardware dept coming to see us Software guys saying "hey guys, we just bought this 500K$ machine, it will be in production next week, could you interface it ASAP with our factory?", when what they are asking is in fact a 6-month job...

[falconeye]

OK, just to be sure, I've ran some tests again...
I have one desk lamp, one white/red cushion and the FA50/1.4... AF assist lamp is disabled.

And who tells you that white/red contrast causes the same issue as normal subjects?


Really, there are too many ways to fail in an AF test. Another I've seen here in this thread is too little empty space around the focus point as the AF sensor areas are pretty large.

The problem with a standardized setup is the problem about standardizing a light source. Maybe, direct illumination from a 40W tungsten bulb at 30 degrees and a given (large=outdoor at night) distance would be an option. But then, you can't buy them in EU anymore... The target could be white paper with two black squares (my standard blur test target) at 1m distance (or 32x focal length or whatever). The blur could then be measured numerically, so no need to post images then. And from posted images, I could derive the numeric blur value.

[30v iPentax]

Olisi hyvä tietää onko näitä kaikkia vikoja K-5 ollut enemmän Nikon tai Canon käyttäjillä. Vai onko jokin muu merkki viellä lyöttäytynyt tähän joukkoon. Itselläni ei ole ilmentynyt viellä mitään näistä vaivoista. Itsellänikin ollut monia merkkejä senhetkisiä hittejä, mutta kyllä Pentax on ollut aina kärjessä.

[Ingemar Basth]

I have used Pentax cameras for many years and i think they are a god companion on the field.
I have taken hundreds and hundreds of pictures with K 5 and i am convinced that there is a serious problem with the AF. I have lost many good pictures because they are out of fokus and useless.
For me who bought this camera and very often working in low light the K 5 is more or less useless. The LV-mode is not a solution for me.

I am so sad, because the camera is a really winner ,,but without a working AF-system i have to say godbye and change system. So dear Pentax give me some explanation to hangup on or i am going to sell out my equipment.. i don´t want that but i need a system thats work, nothing else.

[simico]

Really, there are too many ways to fail in an AF test.

Falk, what do you think about my quick "test" that I wrote on page 2? Both that and a few hundred real life images (e.g. portraits and candids of my daughter, all shot under the same conditions as the "test") lead me to believe my K-5 works fine - whereas my previous K-5 failed miserably.

To be honest, proper testing or not, I'm satisfied with this K-5 as long as my real life images are ok (read: I can't notice FF on the images) under EV2-EV4 tungsten light.

[Wheatfield]

I have used Pentax cameras for many years and i think they are a god companion on the field.
I have taken hundreds and hundreds of pictures with K 5 and i am convinced that there is a serious problem with the AF. I have lost many good pictures because they are out of fokus and useless.
For me who bought this camera and very often working in low light the K 5 is more or less useless. The LV-mode is not a solution for me.

I am so sad, because the camera is a really winner ,,but without a working AF-system i have to say godbye and change system. So dear Pentax give me some explanation to hangup on or i am going to sell out my equipment.. i don´t want that but i need a system thats work, nothing else.

This is exactly the sort of junk post that is becoming increasingly irritating because it is hysteria driven drivel.
All of a sudden, a low EV glitch in the AF has become "without a working AF".
Does your AF not work at all under any conditions? For your statement to be true, you would need to have a completely inoperative AF.

[Ingemar Basth]

As refererence when i tested the K 5 i used my K 10D. In the same light conditions and same testing procedure by all means. The K 10D was not out of fokus. The K 5 was... So the K 5 has a various FF depending on light conditions.
I thought the K 5 with the high ISO-settings was a killer. Its something wrong with the AF-system. In low light it cannot focus ok. Why? The K 7 i borrowed for deeper testing as a second reference is also ok in low light conditions.
So the AF i K 5 is working ok as long as the light conditions is god. If not, out of fokus...

[Eric Seavey]

I take photos almost every week at a dance club with about 2 to 3 EV brightness. I set the AF fine adjustment to -10 and got a usable number of in focus shots. The k10d takes much longer to focus and often misses, the k7 is also slower and misses less often. The k5 locks in but consistently FF. I use a sigma 30mm f1.4. For now this will be sufficient I guess.

[dlacouture]

In the end, the problem is more that with high-iso prowess, the K5 made us try shots that we never would have considered, even with the K7...
And its high-iso usability is simply far greater than its low-light focusing ability.

At this Ev, K10 and K20 AF is simply useless in real-life subjects (meaning, not static targets with some high-contrast edge), but at least it says so by failing to lock.
Even with the K7, at such light levels, the focus simply either bounced about, or the AF assist lamp kicked in. But again it usually said it failed.
And in both case, without a flash or a tripod, the pic would be so noisy as to be hardly of any use except for 4x6 print where any focusing error would be lessened.

Problem here is that the k5 has a gap between its low-light AF limit and the AF assist lamp triggering, and its 12800 iso is too darn good.

[falconeye]

Falk, what do you think about my quick "test" that I wrote on page 2?

You mean post #18? (Sorry, but the thread only has 4 pages the way I configured it)

What you write looks a lot like your second K-5 is fine. But the reports of that kind are rather rare and they may point to a bug in the bug, so to speak

Moreover, without a more controlled test it really isn't possible to draw any conclusions. For others, it must remain hearsay and that's not good enough. Nevertheless, if I find a practical way to do a controlled test delivering numbers which can be compared, I'll ask you to participate

[bjan]

Dear Falk,

Did you have direct contact with any Pentax People regarding the issue concerned ?

Have a nice day