Normally, I rarely read reviews at Steves Digicams, owing to the non-professional and non-scientific / laymen nature of their reviews in which comments are often made too brief and subjective. However, a DPR forum user today has just pointed out that he found "awful lot of hot pixels" in the latest K20D review at the site. And as pointed out by the first few replies in the thread, I've visited the sample page of the review:-
http://www.steves-digicams.com/2008_reviews/pentax_k20d_samples.html
If you look at those "studio" pictures, it is not difficult to see there are various obvious green hot pixels across the frame which could appear or disappear from picture to picture. The most worrisome thing is that those hot pixels could even appear just at ISO 100 and with in-camera jpegs. Just see these two samples:-
http://www.steves-digicams.com/2008_reviews/pentax_k20d/samples/IMGP0311.JPG
http://www.steves-digicams.com/2008_reviews/pentax_k20d/samples/IMGP0312.JPG
Furthermore, I think the hot pixels are not generated as a function of (longer) exposure time (which should be normal then) as the second picture was taken with flash at 1/45th second. If we are to going to examine the photos in the gallery one by one, it can be seen that even a higher shutter speed could still give (even more) hot pixels.
Despite Steve does not mention about the hot pixels issue in his review (for either he was unaware of the issue or he opted not to mention it), his sample pictures say it all for the issue.
All in all, this is the third review sites who tested the K20D have found or shown the hot pixels, regardless if they reported that or not, or if they have published their reviews or not. Up till now, the Polish site got 3 K20D production units and found all of them had this issue whereas Dpreview had got two production units and found both of them were the same (so that they opted to delay the K20D review instead of telling their readers formally in a published review about their solid findings). Now, Steve's direct out of the camera sample photos show the same. I think now that the issue has now been proven to be 100% real. Let's wait and see what Pentax/Samsung will do to address this problem.
Since that hot pixels issued first surfaced, I have seen that some people said that the hot pixels will not affect image quality. But I am totally disagreed with this. The main reason to use a higher pixel count DSLR is to get more room for cropping. With such hot pixels issue, how can those K20D users to "freely" crop their pictures? Do they need to shoot RAW all the time and use software to remove all those hot pixels afterwards?
Sorry but Steve's digicams is just as scientific as your site, although it has no pretensions of being scientific.
ReplyDeleteHow many are the pixels on a K20 sensor? 14.something million ! You need a lot of hot pixels to make a difference among that many pixels.
My feeling is that you are seeing too many trees, not enough forest.
Bjorn
On examination of the two frames you link to, I can see four hot pixels, which appear in the *same* locations in both files.
ReplyDeleteYou mention many random hot pixels - what is your definition of 'many' and 'random'??
I have tried to keep an open mind about your views on Pentax products, but I now believe that you are totally biased. This is sad, as your technical knowledge could be put to much better use.
David
Compare more samples in the gallery and you will know. They are not sticky dead pixels afterall. Even if these are dead pixels, is this acceptable? Do note that they are visible on the final photos.
ReplyDeleteParticularly, if you still won't believe, inspect also this and compare (and other outdoor photos as well).
As for the two frames I link to, it mainly about that I found it most worrisome to see these hot pixels even at the ISO 100 low speed and with in-camera jpeg, which is something that has not been reported before.
Seems to me that Richigh is getting a little desperate for some juicy new material. So sad. All this time spent on pseudo research, long winded diatribes, and for what? to get 3-4 ridiculing comments in return? wow. how sad is that?
ReplyDeleteSeems some have found the VERY specific circumstances that cause the "numerous hot pixel" phenomenon.
ReplyDelete............
The *key* it seems is the choice of drive mode chosen.
Use the 2 sec. self timer mode and get the 'pixels' or use the single
frame shooting mode and get no hot pixels...........................
Further testing by others
...................................
Just FYI this also occurs when using the wireless remote with 3 second delay--the other mode that locks the mirror up in advance. Seems certain the mirror up has something to do with it. I cannot duplicate the results in any other mode/combination where the mirror is not up prior to shutter firing (including live mode w/o 2 second lockup delay)....................
CONCLUSION: Apparently there is something buggy w/ 2 sec delay and the sensor/processing
CURRENT SOLUTIONS as proposed by others:
1)To avoid this problem with 2s mirror lockup when shooting JPG on a tripod, one may want to switch long-exposure NR to ON (as opposed to AUTO). This should avoid the problem in this particular situation.
2)They could certainly work around it by activating the dark frame NR whenever the 2-sec delay is used, and even take the dark frame during the 2-sec. delay rather than after the exposure.
NOTE: Unconfirmed but it is looking like most of the hot pixels are NOT transient even w/ the 2 sec delay "bug" but it has to be verified further.
................................
The hot pixels appear in exactly the same areas,some premapping frame glitch i suspect,firmware should be able to solve
.................................
http://forums.dpreview.com/forums/read.asp?forum=1036&message=28062243
Anything between dot lines are quotes.
Anonymous,
ReplyDeleteI would not consider him to be "desperate" because of this notice. Couple of friends of mine claims that a few hot pixels really does not matter. No-one of them suffers from hotpixels on their (older) pentax bodies. I do...
I have two bright green hot pixels near the centre, they are on all shots even iso100 and 1/2000, they usually get aliased through next four pixels and they do not vanish during resize. This effectively make me to run all picture through the spot healing software even in case when any other processing is not needed or wanted... For me this is an issue and I believe pointing it out is a right thing
Seems you've fallen to the level of all the forums you've been booted from. I had 2 good rebuttals posted and "surprise" there gone.
ReplyDeleteYou should leave your yellow journalism to China.
What? I have never deleted any comments posted here. What a lie!
ReplyDeleteMind you also that don't make attacks to a country or race. If it happens again, it will be the only reason I delete *your* post - which is totally off-topic and not camera related.
My apologies, for some reason my posts disappearred temporarily on my computer. May have been a cache issue, but I'm not sure.
ReplyDeletePlease accept one of the few apologies you'll get from me. ;)
OK save your yellow journalism for yourself.....
It was meant to be "leave it at home" not as a specific hit at any country. The US has it's share of "yellow press" though I think it we collect statistics I have a feeling that things wouldn't be that equal. but who knows...
What amazes me is that there has been 'measurbators' that have been able to isolate the issue and apparently point the cause to the 2 sec. MLU, as has been commented before, and the blog owner still doesn't mention it in the article body or in a new article, not even as a comment...what happens? you don't trust these tests? is the issue not enough worriying for you now? Come on, it is certainly a Pentax bug, please, inform us !! ;-)
ReplyDeleteWhy should he mention it.. it's not
ReplyDeleteCOLLOSAL, or STUPENDOUS.. or SENSATIONAL enough.. nor is it easy to spin 5 paragraphs of half truths and bad pseudo-science. RH believes one test is enough to condemn a whole generation of things. 3-10 independent testers only seeing an issue (more hot pixels than there should be) in a VERY SPECIFIC situation certainly isn't GRANDIOUS as one obscure test done by some methodology challenged individual..
Funny thing is, this issue is abit more critical to me than all the handful of random hot pixels would be otherwise. MLU is one of the things I like to use and shouldn't be penalized for it's use.
As written up by RH...YAWN...
Not saying that the issue does not exist -- however I have used ML function many times with K20 and have yet to encounter a single hot pixel. strange yes? I shoot DNG raw exclusively so perhaps ACR is wiping them away for me.
ReplyDeletein any case, this has been a complete non-issue for me though I don't suppose RH cares to document that it is not a universal problem. Nah that would be too fair and balanced.
Actually, I think that the problem is universal, but very localized (and, if one shoots RAW, the converter will eliminate the hot pixels for you).
ReplyDeleteSo, the sequence is:
- Some reviewers notice about hot pixel (HP) issue. As usually, Ricehigh (RH) quickly devote a blog entry to it.
- Another reviewer DOES NOT notice about HP, but RH does find HP in some of the sample pictures and devotes a NEW blog entry to it.
- A thread in a forum seems to have perfectly located the issue, that seems to be related to the 2sec MLU. Lots of users try to reproduce the issue succesfully. RH DOES NOT MENTION IT.
- Someone mentions the 2 sec. MLU finding as a comment in this RH blog (see above).
RH AGAIN DOES NOT REPLY/COMMENT IT (but instead he reacts very quickly when someone mentions his Country, so he indeed reads his blog comments)
Frankly, it is surprising how quickly RH reacts to some events and how slowly reacts to others.
My opinion about his honesty about pointing out the Pentax equipment issues has dramatically changed looking at his behavior with this particular issue.
I am disappointed with him but even with me for not having realized before.
It seems that it is a firmware error that will be fixed in the next firmware update:
ReplyDeletehttp://forums.dpreview.com/forums/read.asp?forum=1036&message=28085894
Dear Mr. Ricehigh,
ReplyDeleteWe eagerly await your response to the above url'd report that the hot pixel "problem" is, and will be, fixable via firmware upgrade. Remember to be humble, classy, and above all, honest with yourself and your audience.
Oh, he already answered...in a Forum, not in his blog:
ReplyDeletehttp://www.pentaxforums.com/forums/254460-post143.html
No news: Extremely openminded when possible problems arise, extremely skeptical when possible solutions arise.
As I said before, what a enormous deception.
C'mon Ricehigh!
ReplyDeleteI want to see you turn a silk purse into a sow's ear! You're going to have to better than "New Super Aggressive Hot Pixel Ruling Out Algorithm". I am sure that you can bet more negative than that.
We're waiting...
I don't know how true that *rumour* is and when it will actually happen even if it is really true. However, not until there is an official announcement from Pentax and that the new firmware has been tried out for a while, nobody will know if it really works or not. So, let's wait and see. Of course, it's yet a good thing *if* Pentax can really learn to admit problems and most importantly, to react fast to issues.
ReplyDeleteRiceHigh won't address the firmware fix because it is a "rumor". Yet he has no qualms about broadcasting unsubstantiated rumored "problems". your approach is assymetrically misguided at best. dishonest and maliciously agenda-laden at worst.
ReplyDeleteYou say the K20D hot pixels issue is a "rumor"? Are you kidding?? Are all those posted pictures with hot pixels shown up by reviewers and users FAKE???
ReplyDeleteWell, what a really silly and stupid fanboyism which can make man blind like this..
My god you are a twit. do you not read your own blog? hahaha. go back and read some of your previous entries numbskull.
ReplyDelete"Breaking News! one K10D user in zimbabwe reports a faulty shutter button! Major design flaw!"
"News flash! a single K100D user in Nova Scotia reports flaking body paint! Pentax QC is to blame!"
Ricehigh, you don't need to comment about a new firmware that it is still a rumour.
ReplyDeleteBut at least you could mention that the problem has been isolated and that lots and lots of users have verified it.
Please, read your own blog. You have recently devoted an entire entry about an issue found by a SINGLE user and we have never heard of it again:
http://ricehigh.blogspot.com/2008/05/k20d-banding-problem-reported-at-low.html
So possible issues find its way to your blog immediatly, no matter if they are universal or reported by a single user, but possible explanations/solutions (specially if they disminish the seriousness of the issue) must be not only verified by tens of users, but also officially acknowledged by Pentax.
The disappointment keeps growing...
Hi Rice,
ReplyDeletebelieve me - the problem is really fixed - i was this week at the PENTAX Europe Headquarter in Hamburg and the chief technician Mr. Petersen flashed my K20D with a brandnew partial new firmware and we both tested it -- no hot-pixels anymore. After quality testing phase this will be implemented in the next firmware update - soon.
Cord
Hot pixels will be fixed:
ReplyDeletehttp://www.imaging-resource.com/NEWS/1212174178.html
"Pentax USA has confirmed the issue, themselves noting that it only seems to occur when the 2-second self-timer is employed. They said that the engineering team was working on a firmware fix for the problem, and that a fix would be publicly available at some point in the future. No ETA was available for the fix yet, as the engineering team had apparently only just recently confirmed the problem and its cause.
Stay tuned, we'll be sure to update our readers as soon as we hear from Pentax that a fix is available. In the meantime, the workaround is dead simple: Just don't use the 2-second self-timer and you should never see the problem."
Seems odd that such a fervent user/ supporter of the Pentax brand can do nothing other than pick holes in evey product now :(
ReplyDeleteimaging-resource.com not good enough for you? C'mon RiceHigh.
ReplyDeleteThe problem with hot pixels were also found by Polish reviewers in RAW files they shoot with K20D (see: http://www.optyczne.pl/69.7-Test_aparatu-Pentax_K20D_Szumy_i_jakosc_obrazu_w_RAW.html).
ReplyDeleteThe problem does not existed in JPEGs, because dead pixels were mapped-out. They processed RAWs with dcraw (they used it in all their tests). As they wrote, they check several K20D's bodies and they all showed hot pixels in RAWs.
Now, the firmware is official...waiting for RiceHigh comment
ReplyDeletehttp://www.pentax.jp/english/support/digital/k20d_s.html
http://tinyurl.com/4ls3j9
Don't hold your breath "Disappointed", else you are likely to be even more disappointed.
ReplyDeleteRice has a clear agenda. Facts are irrelevant here.
Why all the attention to pentax's hot pixel probs. Canon and Nikon also have their fair share of hot pixels issues. Try google.
ReplyDeleteSimple. Just because the K20D is much hotter than the others. Period.
ReplyDeleteGuess I'll have to post here since you didn't update this blog w/ my testing results for this.
ReplyDeleteJust FYI, I'm OCD about testing and pixel peeping but I tested my K20D when I first got it quite a bit (including testing for sensor/viewfinder tilt, low temperature banding testing, etc.).
The hot pixels are really weird because they look like pixels with rings around them and there were quite a few that moved around. Higher ISOs made it worse, but long 30sec exposures had no problems (probably because dark frame subtraction removed them). I took probably 50 photos to test this.
The kicker is that all this was fixed by running the built-in pixel mapping feature in the menu. I pixel peeped the next set of images all the way up to the useless ISO6400 and the hot pixels were gone.
I've since put it through a photoshoot of over 100 images w/o problems and had to use shake reduction in low light for 1/4 of the shoot.
So, yes, it's weird that they move around but the pixel mapping seems to fix it so that it's really a non-issue...
According to dpreview the hot-pixel and banding issue on K20D is a problem. Is there any verification that Pentax firmware update has resolved the issue on K20D?
ReplyDeleteThere was an update already but the problem is not totally eliminated. Search my Blog for more relevant reports.
ReplyDelete