[rfortson]

If you didn't do a lens cap test, I think it is permissable to assume that your unit wouldn't be the single exception known to date. Probably, your shot didn't contain enough dark area to spot the hot pixels. Most of them are pretty dim actually...

That very well could be the case. I took a shot of a dark (but not black) fabric couch in my office while I was sitting at the computer. I got a nice shot of fabric grain but not a lot of hot pixels.

I can see your point that mine may not qualify as "unaffected" by your methods. Of course, my point is that I'd rather not go look for reasons to make myself unhappy with my camera. I'll also happily apply the updated firmware when it comes out.

[falconeye]

I'd rather not go look for reasons to make myself unhappy

You are a wise person.

[gkopeliadis]

codiac,

The thing is that you don't specifically try those pixels to appear. Most likely in normal shooting conditions they don't appear.

DPReview simply are looking for problems. They are measurbating, torturing camera and measurbating again. And oh, look, hot pixels. They hinted that they got pixels after LiveView use. That's very understandable. During LiveView sensor heats up and chances that hot pixels appear increases very much.

Nikon D300 also have exactly the same problem (Nikon D300 Review by Thom Hogan), however DPReview didn't bother to mention it in their review.

Now it seems, they are preparing ground for another not so favourable review: K10D had "soft jpegs" (although soft D300 jpegs were OK), K20D will have "hot pixels issue"...

As far as I can see, YOU are the one biased against DPR
Things tend to be in the gray area, neither black nor white. I've red quite bad comments for products by any brand in that site. I don't think DPR has any bias against Pentax. Making mistakes is another mater.

[falconeye]

I have carried out a short study of the hot pixel phenomenon using the mirror lock-up trick.

The result is available here for download:

http://www.livis.de/download/pub/photo/k20d-hotpx.pdf

The main result is that hot pixels are dim, they are many. I compared with the Pentax *istDS and both are very similiar in this regard (at 2s w/o dark frame subtraction). so probably, nothing to be concerned about.

Sometimes, not reproducible though, there can be 1000s which are visible. I am wondering if the last phenomenon isn't what has been observed by DPR, c't and others. And is more or less pronounced from camera to camera.

---

Last edited by falconeye; 05-30-2008 at 05:24 AM. Reason: Added istDS comparison

[Vlad]

when are they gonna release the new firmware?

and is the firmware gonna solve the problem?

[gkopeliadis]

when are they gonna release the new firmware?

and is the firmware gonna solve the problem?

The answer may be here

[falconeye]

I updated the study I posted earlier.

The main addition is a comparison with a *istDS. Using this comparison, there is no significant difference between how the K20D and the *istDS produce hot pixels during long exposures w/o dark frame subtraction.

The answer may be here

Ok, this person currently carries out tests with the only camera (his K20D) having the new firmware applied as a patch already, at German forum DFN.

Preliminary results indicate that the hot pixels are gone, indeed and that no additional 2s dark frame subtraction is carried out. The same for 3s remote shutter release delay.

[Reps]

check this out also: NEWS! - Pentax K20D hot pixel bug confirmed, tests confirm limited scope. and see the images taken at their forum. It does not bug me at all as I find K20D a superb cam for my own needs and NONE of product photos taken with 2 sec MLU do not suffer from it (RAW of course)- so I am lucky and MUST not do vol. 2 retake :-)
Best and happy shooting, JR

[flyer]

There is a firmware upgrade coming soon for the hot pixels issue:
Pentax K20D Hot Pixels Problem Confirmed | PentaxLife

[BeerCan]

Just for curiosity sake I have a question. I know the k20d has a feature to map out hot pixels. Is it possible to "un map" them? Like a reset function to start over?

[flyer]

I think this is the best question I heard so far on this topic, and you raise a very interesting point. Normally, you wouldn't have to "unmap" those pixels, since they are dead, but if they are mapped out due to a glitch in the software, I see that it could be a problem after some use.

[falconeye]

The map can't be undone, as far as I know. Which is why I never used the function. The missing "reset pixel mapping" function, IMHO is a bug in the firmware.

[RiceHigh]

The map can't be undone, as far as I know. Which is why I never used the function. The missing "reset pixel mapping" function, IMHO is a bug in the firmware.

Eventually, the whole image frame could be ruled out! ;-)

[jeffkrol]

I think this is the best question I heard so far on this topic, and you raise a very interesting point. Normally, you wouldn't have to "unmap" those pixels, since they are dead, but if they are mapped out due to a glitch in the software, I see that it could be a problem after some use.

since it's just a 2 sec timer issue 1) Can pixels be mapped at this point.
2) Subsequent re-mappings may involve going to zero and starting over again. Who or where is it stated this process is "cumulative"?
I wouldn't design it this way. Permanent pixels will always be there so no real need to not remap the whole sensor.
There seems to be a lot of judgements.. hint hint.. based on little facts.

[jasonwarth]

Hi All,

I've perused some of the posts in this thread, but didn't real all 9 pages, so please forgive me for that.

I appear to have a slight hot pixel issue with my K20D. As best I can tell, it's a single pixel -- or maybe 2 pixels at the most. It's not something that bothers me, as it rarely (if ever) shows up in my "final product." I shoot exclusively in RAW and use Lightroom for post processing. I usually only notice this issue when flipping through images in Lightroom. If I pause long enough for the image to "settle," the pixel often disappears.

Regardless, if anyone's interested, I used Snapz Pro X to record a small movie file of the affected area as I flipped through pictures in Lightroom. The movie file is about 3.5 Mb in size, so I compressed it a little to allow it to fit into a .zip file that would attach to this post. I'm happy to share the original if anyone wants it.

For the record, I have no interest in taking any sides on what appears to be a somewhat contentious issue. I'm only offering this in case it helps the community arrive in some way. Overall, I'm thrilled with my K20D, and am 100% confident that I made the right choice for my first DSLR.

Cheers,

-Jason