[Fer]

My findings don't quite match what Pentaxeros says about the problem being confined to fast glass (such as some Limiteds) at wide apertures. As shown in my samples 3 or 4 pages back, the issue is very visible with the DA 18-55 at f5.6. I've repeated the test a few times and it's very clear.
Other than that, I confirm from "knowing" him from the Spanish forum where I regularly post, that Pentaxeros is a most knowledgable and reliable source. Still, I'm relieved that Pentax tech support confirmed to me in writing that they are on top of things in this issue.

[pcarfan]

Hola, como he dicho anteriormente, a TODAS las cámaras les ocurre.

Yo administro en España una tienda especializada solo en Pentax (La Tienda Pentaxeros.com), y todas las cámaras que entran y vendo, son probadas por mi una a una y actualizadas el firmware, etc.

Y todas las unidades probadas, incluso de las ultimas con numero de serie: 39XXXX, todas, bajo las misma situación lumínica controlada que tengo, y que se reproducir, actúan igual todas las K-5.

El porque muchos no lo detectan, es debido a varios puntos:
- Solo es con objetivos luminosos, con enfoque de tornillo. (No todo el mundo tiene limited, o lentes luminosas)
- Solo es con ciertas tonalidades de tungsteno.
- Solo ocurre en situaciones con muy baja luminosidad, y en el umbral en el que no se activa la luz de ayuda. Si se activa la luz de ayuda al enfoque, el enfoque es preciso.
- Además en cuanto el diafragma se cierra desde f/5.6, la profundidad de campo absorbe el front focus, siendo inapreciable.

Las situaciones en las que se da es muy bajas.

Como dije, realice un estudio que reporte a Pentax, con toda la información, y se de mano directa de Pentax España y Europa, que están trabajando en ello, y que la solución vendrá por una actualización de firmware.

Saludos cordiales...

Thank you very much....added reputation.

[Class A]

Thank you very much....added reputation.

+1

Thank you very much as well for doing the test, reporting the results to Pentax, and keeping us informed.

[bjan]

Pentax Benelux also made a report to PENTAX Japan several weeks ago like the Spanish did but soo far Japan did not even respond to their reporting . ("very bad communication in terms of response to our comments and questions")
What makes you believe that the same kind of report from Spain will cause an other reaction ? The problem is not even confirmed by PENTAX JAPAN . Leaving their dealers around the world struggle to keep people loyal and not running away . I can understand the efforts from PENTAX SPAIN and PENTAX BENELUX to calm down the situation by saying that they are working on a soution. However the importers cannot do anything other than reporting things to PENTAX JAPAN and wait for the mighty emperor to respond. Poor situation for the importers that they have not faced any support yet that they can give proof of.
The problem is very obvious for many weeks now and PENTAX has not moved towards many customers around the globe that are not certain about an eventual solution. Again I can understand the Spanish poster as he has big interest in keeping the case calm for his business and Falk responding positiv to these "softening" messages because of his obvious good relationship with PENTAX . BUT... for us the user who has paid 1200 Euro for the K5 and having a finite livespan (Falk..) this is a quit dissatisfying way of dealing with your customers and worth a loud cry for fast action rather than a sweet talk about a possible firmware that might come up one day when PENTAX might even consider our problem also their problem.

[JamieP]

I didn't fully repeat my previous tests today with the new body as there wasn't enough daylight coming into the house for me to feel confident in adjusting the lens. However, from some quick tests (interference pattern method) using the DA 70, I needed a +10 adjustment.

Later this evening I started doing some informal shots under tungsten lighting to see if the adjustment was still in the right ballpark. I found that the adjustment was way off in the artificial light (expected) but was shocked by the difference in focus performance based on what I shot in the room. I initially shot at a diploma hanging on wall. It was about 5 feet away and was receiving as much light as was available in the room. From shooting this I worked down to a 0 adjustment for the lens. Then I tried shooting a box with a barcode on the shelf. It was a about 3 feet away and was not receiving as much light as the diploma. Every shot was out of focus. I tried using flash to get a faster shutter, but still blurry. I then shot it with LV and it was much sharper. Finally, I turned the lights out in the room to force the focus assist light to come on and again got a much better focused image. I took multiple shots of each and this was very consistent and repeatable. 100% crops from the series below. Keep in mind this isn't a sharpness test. Just compare the images relative to each other to get an understanding of how lighting is affecting focus.

1- Diploma on the wall (decent tungsten light, viewfinder focus)
2- Shoebox (reduced tungsten light, viewfinder focus)
3- Shoebox (reduced tungsten light, viewfinder focus + flash)
4- Shoebox (reduced tungsten light, liveview focus)
5- Shoebox (nearly dark room, viewfinder focus + focus assist lamp)

[bjan]

What is your conclusion Jamie ?

[JamieP]

What is your conclusion Jamie ?

The shots I took tonight just illustrate the issue as defined by Pentaxeros. The only surprising aspect was that the effect was triggered just by focusing on different subjects in the same room. To the eye, the lighting on the diploma and box wasn't very different so that was not expected.

In the back of my mind is the quick tests earlier in the day that I was shooting at +10 for accurate focus. This coupled with the shots posted (with 0 adjustment) would indicate there isn't just 2 scenarios in play, but possibly many based on the amount or type of light. I'm not drawing any conclusions about that because the quality of my tests don't warrant it.

I hate doing this stuff, but really want to understand it and figure out how to best use the camera for now. Hopefully a fix will be available soon and I can just forget about it.

---

Last edited by JamieP; 12-22-2010 at 11:01 PM.

[Tommot1965]

just too add a little perplexing question....what if pentax cant fix this problem...?..and its hardware based rather than software..{firmware}

i dunno enough about it...but im about to pull the trigger on a trade in deal with my K7..this issue has me wondering if I should ditch the K7 / K5 deal..and grab a D7000...as Ive only two Pentax lenses..im hardly in too deep to get out , theres also so many more Nikon lenses out there in the used market and to rent....pentax rentals are zero where I am....sorry for getting OT


but all the tests Ive seen so far..the K5 images are better than the D7000..so Id rather get a K5...my head will explode soon im sure of it

[Fer]

The shots I took tonight just illustrate the issue as defined by Pentaxeros. The only surprising aspect was that the effect was triggered just by focusing on different subjects in the same room. To the eye, the lighting on the diploma and box wasn't very different so that was not expected.

In the back of my mind is the quick tests earlier in the day that I was shooting at +10 for accurate focus. This coupled with the shots posted (with 0 adjustment) would indicate there isn't just 2 scenarios in play, but possibly many based on the amount or type of light. I'm not drawing any conclusions about that because the quality of my tests don't warrant it.

I hate doing this stuff, but really want to understand it and figure out how to best use the camera for now. Hopefully a fix will be available soon and I can just forget about it.

Well Pentaxeros has not posted any images, so we have to guess it is the same issue.
However, your images mirror exactly the ones I have posted in that you can immediately tell which shots are taken with LV focusing (always sharp) or VF focusing (always blurred). So yes, the problem is very much there.
Thanks for the samples.

[znieh]

@tommot
I very well understand your situation K5/D7000. First I had K5 - 4 times, and I returned them one after the other mainly because of FF issue described in this thread. After this experience, having also only two lenses, I bought the D7000. It is a great camera, and I cannot subscribe to your statement that K5 IQ is better. AF is a really good feature (and after the K5 traumatic experience to me was the main point!) and some other details are really nice. But: This cam had hot pixel issue at slower exposure times with higher ISO.
Finally, after my store guy told me that Pentax Europe (Hamburg) had acqknowledged the issue and confirmed Pentax is working on that (note: not yet clarified if hard- or firmware issue), I gave back the D7000 and got the last K5.
--- Now I'm waiting for the fix!!! And getting along with unperfect work-arounds like USER1 w/ -10 adjustment (which I named "TUNGSTEN" ... ) and LV-Focusing. A pity that there are many situations where the adjustment would be in between ...

[Fer]

@tommot
...
Finally, after my store guy told me that Pentax Europe (Hamburg) had acqknowledged the issue and confirmed Pentax is working on that (note: not yet clarified if hard- or firmware issue) ...

znieh, while the fact that many users are not complaining about the AF issue would suggest it might be a hardware issue, it is quite clear now that it is a firmware issue, and people who don't have it are simply not shooting in low tungsten light, or not looking at their images close enough.
If you read the previous two pages, you will see that Pentaxeros runs an online shop in Spain and says ALL of the K5 units he has tested have this issue. What's more, I posted the reply I got from Pentax Europe support stating that they are working on a firmware to fix this.
So while it's not entirely unconceivable that there might be some hardware issue in the mix, this seems to be mainly a firmware issue.

[novsky]

Hi Fer, did u think that it could be a Firware issue, like using the contrast detection of the liveview (for tungsten light only)?? I really hope that it won't be an hardware issue, don't want to send it back again after the stains on captor solved.

[bjan]

Hallo Znieh. In case you do a -10 adjustment with a zoomlens please test this on several distances. Good change that you have good focus on 1 metre and bad focus at 6 metres for example. Adjusting FF or BF with microadjustments is very risky with ZOOMLENSES .
You will get from one problem into an other . ON a fixed lens this is no issue though

[pcarfan]

I wanted to see whether my impressions of the K-7 doing a better job than the K-5 in Tugsten light was correct or not. To my surprise the initial test I did showed even the K-5 not to have any problems as the images taken with VF and LV looked the same. Then I realized I had setup the target right under the bright light, and where as I had seen the problem previosuly in much dimmer light, so I repeated the test after lowering the Tungsten light and was able to easily document the difference.

According to my tests the K-5 has this Tungsten AF issue ONLY in dim light, in decent light the K-5 does fine with Tungsten light focusing

Here is the evidence that the K-7 does a better job even in low light Tugsten light

First K-7 VF focus


K-7 LV


K-5 VF focus


K-5 LV focus


EXIFS Intact

[Tommot1965]

@tommot
I very well understand your situation K5/D7000. First I had K5 - 4 times, and I returned them one after the other mainly because of FF issue described in this thread. After this experience, having also only two lenses, I bought the D7000. It is a great camera, and I cannot subscribe to your statement that K5 IQ is better. AF is a really good feature (and after the K5 traumatic experience to me was the main point!) and some other details are really nice. But: This cam had hot pixel issue at slower exposure times with higher ISO.
Finally, after my store guy told me that Pentax Europe (Hamburg) had acqknowledged the issue and confirmed Pentax is working on that (note: not yet clarified if hard- or firmware issue), I gave back the D7000 and got the last K5.
--- Now I'm waiting for the fix!!! And getting along with unperfect work-arounds like USER1 w/ -10 adjustment (which I named "TUNGSTEN" ... ) and LV-Focusing. A pity that there are many situations where the adjustment would be in between ...

crikey mate , you have been through the wars with your D7000 v K5 experience ,

I ended up buying a K5 today ... lets just hope firmware 1.02 sorts all this out..