To all concerned with K20D/Hot Pixel Issue

PentaxPoke · 05-14-2008, 08:04 AM

Originally posted by mconder

Do you think it's a DPR/Canikon conspiracy?

I wouldn't say a conspiracy. More a subconcious bias towards thier biggest "sponsors."

Lance B · 05-14-2008, 06:51 PM

Originally posted by PentaxPoke

I wouldn't say a conspiracy. More a subconcious bias towards thier biggest "sponsors."

That is a good way to put it, IMO.

reddream · 05-15-2008, 02:48 AM

I found this issue is easy to reproduce. In fact, I rarely shoot in RAW (2 out of 2500 pics). But when I converted my only 2 dng (normal shoot without live view) files into tiff using DCRAW today, I did find out several hot pixels.
When I view the result in Picasa, I found Picasa doesn't remove hot pixels (of raw) either. So checking raw in Picasa seems to be the easiest way to verify this problem.

mconder · 05-15-2008, 08:42 AM

Originally posted by reddream

I found this issue is easy to reproduce. In fact, I rarely shoot in RAW (2 out of 2500 pics). But when I converted my only 2 dng (normal shoot without live view) files into tiff using DCRAW today, I did find out several hot pixels.
When I view the result in Picasa, I found Picasa doesn't remove hot pixels (of raw) either. So checking raw in Picasa seems to be the easiest way to verify this problem.

Will you post the exiff data? I just purchased a k20d over a Canon 40D, so I hope that this is not a normal occurrence for this sensor and I hope I do not get one of these bum copies of the camera.

benjikan · 05-15-2008, 09:14 AM

What Issue (Problem)?

mconder · 05-15-2008, 10:24 AM

Originally posted by Photo Tramp

All camera manufactures have the same 0.2-3.0 percentage of hot pixel acceptance from brand new.

Ya, but I have owned several Canon's and I can say from experience, a brand new Canon very rarely has even one hot pixel. This is the standard Pentax should and will be held to.

KungPOW · 05-15-2008, 05:44 PM

I understand the K20D has a feature to map out hot pixels. Can the K20D turn this feature on and off, so that if you wanted to, you could see the hot pixels?

Is it possible that the hot pixel issue is just a matter of the pixel map being turned off on some cameras?

Another question: Did Pentax add the pixel map feature because they knew the CMOS sensor was prone to hot pixels, or did they add the mapping, as a feature, and are now being burnt by their own upgrade? The question is a litle retorical.

My gut tells me that this hot pixel issue is more of a matter that various imaging software packages need to be updated to deal with K20D files. The software needs to be updated to repair the hot pixels found in the K20D.

Having said that, I do not think that the K20D has a greater or lesser chance of a HP then does the "Golden Can*n" (sounds of angels singing), or the "Mighty Nik*n" (sounds of victory trumpets). I expect the chance of a HP on a Canikon is about the same as on a Pentax.

Lance B · 05-15-2008, 05:59 PM

Originally posted by mconder

Ya, but I have owned several Canon's and I can say from experience, a brand new Canon very rarely has even one hot pixel. This is the standard Pentax should and will be held to.

I have not had any of the dreaded hot pixels with any of my Pentax cameras either.

You say you haven't had any on your Canons either, but is that because they are automatically removed by software, either in camera or in post process?

Maxington · 05-15-2008, 06:27 PM

Originally posted by KungPOW

I understand the K20D has a feature to map out hot pixels. Can the K20D turn this feature on and off, so that if you wanted to, you could see the hot pixels?

Is it possible that the hot pixel issue is just a matter of the pixel map being turned off on some cameras?

Another question: Did Pentax add the pixel map feature because they knew the CMOS sensor was prone to hot pixels, or did they add the mapping, as a feature, and are now being burnt by their own upgrade? The question is a litle retorical.

My gut tells me that this hot pixel issue is more of a matter that various imaging software packages need to be updated to deal with K20D files. The software needs to be updated to repair the hot pixels found in the K20D.

Having said that, I do not think that the K20D has a greater or lesser chance of a HP then does the "Golden Can*n" (sounds of angels singing), or the "Mighty Nik*n" (sounds of victory trumpets). I expect the chance of a HP on a Canikon is about the same as on a Pentax.

I'd guess that its similar to the AF adjust feature, its an ability that has always been available to repairers, but now Pentax has enabled it for end users.

Which is a very handy thing to do. I doubt its just because of the new sensor.

Maybe they listened to everyone wanting to use the repairers K10D AF adjust that got removed in 1.30 firmware? I suspect that they wouldn't mind users adjusting AF, they just didn't have a failsafe setup for the K10D, too risky.

*isteve · 05-15-2008, 07:33 PM

Originally posted by codiac2600

Until there is hard proof of this being a true problem, which as of right now it's not, let's not have 50 million threads on the same stupid subject.

If you have a K20D with said problem please post in this thread a Jpeg straight from the camera with proof of hot pixels so we can look into it. BTW, EXIF data must be intact in order to make sure everything is on the up and up.

So far I tested all 5 K20D's we have in the store and no hot pixels, had friends and costumers email me jpegs and nothing. My tally is officially 14 no hot pixel cameras and the likely hood DPR got two of them in a row is looking slimmer and slimmer.

So if you have proof show it and I don't care if you have a blog saying it's true, we are looking for proof to either make a case about it or drop it at that.

Clearly his phenopmenon is happening under some circumstance, but in Phils uinhelpful way we dont know...

1. What cicumstances? I have used live view for minutes at a time when doing lens tests and I have no hot pixels. Strikes me you have to use it for extended periods before these pixels show up. Its warns against this in the manual.

2. If the pixels are different in each shot, do they eventually reset when it cools down or do some of them stay hot? If th sensor can recover its hardly a problem is it?

Phils rather flippant remark about using LV "a bit" is clearly disingeneous and particularly unhelpful. I suspect if we knew the exact circumstances it would make it easier to replicate.

**jfdavis58** · 05-15-2008, 08:18 PM

..........

Wow!

..........

rfortson · 05-16-2008, 08:10 AM

Originally posted by falconeye

If you read my post #86, you'll see me claiming that PPL removes hot pixels as well. That's part of the reason I posted my testing method.

My claim is actually based on test: I don't see the hot pixels in PPL that I see in PhotoMatix (cf. my images in quoted post).

Updated info: If you make a large 100% loupe window (which is awfully slow to update), you'll see intermediate processing stages within PPL (due to its slowness). Then I can clearly state here that my hot blue pixel (cf. above) (when push brightness up a bit):

Appears, then disappears in the loupe window during processing in PPL.

This clearly proves that PPL eliminates hot pixels in one of its processing stages.

(Maybe PPL didn't do that in the past.)

That's good to know. Are you sure you didn't have any of the noise reduction enabled? I know it didn't do it in the past because I took a couple of long (multi-second) exposures with my K100D and had a couple of stuck pixels. PPL didn't remove them, but ACR did.

I wish the PPL software had better documentation. I like it and find more and more that it does. I just have to find out by trial and error, or by user reports like yours. Personally, I think it's a good thing that software removes hot pixels. I suspect lots more hot pixel removal is going on in all cameras. Manufacturers just don't advertise it as it's seen as a "defect".

rfortson · 05-16-2008, 08:14 AM

Originally posted by mconder

Ya, but I have owned several Canon's and I can say from experience, a brand new Canon very rarely has even one hot pixel. This is the standard Pentax should and will be held to.

I can say by the same experience that the several Pentax cameras that I've owned have rarely had even one hot pixel. My K20D has exactly none.

falconeye · 05-16-2008, 10:10 AM

Originally posted by rfortson

Are you sure you didn't have any of the noise reduction enabled?

I used it in its default setting.

There it has:
- Noise suppression: 0.
- Correction of wrong color information: >0 (dont remember exactly, but ON).
(both translated back from my German GUI)

I don't know what the second setting really means. If it removes hot pixels: fine

rfortson · 05-16-2008, 12:01 PM

Originally posted by falconeye

I used it in its default setting.

There it has:
- Noise suppression: 0.
- Correction of wrong color information: >0 (dont remember exactly, but ON).
(both translated back from my German GUI)

I don't know what the second setting really means. If it removes hot pixels: fine

I don't really know what that does either. It's been there for several versions, so I don't think it's the "magic hot pixel removal" slider. More likely, as you said, they added it to the converter and just didn't mention it.

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Pentax K20D - vertical bands, noise, and hot pixel problems... :-(	pentaxmz	Pentax DSLR Discussion	26	04-04-2012 04:33 AM
K20d Hot Pixel on v1.03	drndrw	Pentax DSLR Discussion	3	03-10-2010 06:40 PM
K20D - Worth upgrading firmware if I have no hot pixel issues?	NicholasN	Pentax DSLR Discussion	6	01-07-2009 01:24 PM
Hot pixel issue duplicated	BeerCan	Pentax DSLR Discussion	1	05-26-2008 05:41 AM
Hot Pixels Issue of the K20D	RiceHigh	Pentax DSLR Discussion	30	05-09-2008 09:25 PM

05-14-2008, 08:04 AM	#106
PentaxPoke Veteran Member Join Date: Jan 2008 Location: Oklahoma Posts: 1,411	Originally posted by mconder Do you think it's a DPR/Canikon conspiracy? I wouldn't say a conspiracy. More a subconcious bias towards thier biggest "sponsors."

05-14-2008, 06:51 PM	#107
Lance B Veteran Member Join Date: Nov 2006 Location: Sydney, Australia Photos: Gallery Posts: 593	Originally posted by PentaxPoke I wouldn't say a conspiracy. More a subconcious bias towards thier biggest "sponsors." That is a good way to put it, IMO.

05-15-2008, 02:48 AM	#108
reddream New Member Join Date: Apr 2008 Posts: 17	I found this issue is easy to reproduce. In fact, I rarely shoot in RAW (2 out of 2500 pics). But when I converted my only 2 dng (normal shoot without live view) files into tiff using DCRAW today, I did find out several hot pixels. When I view the result in Picasa, I found Picasa doesn't remove hot pixels (of raw) either. So checking raw in Picasa seems to be the easiest way to verify this problem.

05-15-2008, 08:42 AM	#109
mconder New Member Join Date: May 2008 Location: Saint George, Utah Posts: 6	Originally posted by reddream I found this issue is easy to reproduce. In fact, I rarely shoot in RAW (2 out of 2500 pics). But when I converted my only 2 dng (normal shoot without live view) files into tiff using DCRAW today, I did find out several hot pixels. When I view the result in Picasa, I found Picasa doesn't remove hot pixels (of raw) either. So checking raw in Picasa seems to be the easiest way to verify this problem. Will you post the exiff data? I just purchased a k20d over a Canon 40D, so I hope that this is not a normal occurrence for this sensor and I hope I do not get one of these bum copies of the camera.

05-15-2008, 09:14 AM	#110
benjikan Veteran Member Join Date: Jan 2007 Location: Paris, France Photos: Gallery Posts: 4,308	What Issue (Problem)?

05-15-2008, 10:24 AM	#111
mconder New Member Join Date: May 2008 Location: Saint George, Utah Posts: 6	Originally posted by Photo Tramp All camera manufactures have the same 0.2-3.0 percentage of hot pixel acceptance from brand new. Ya, but I have owned several Canon's and I can say from experience, a brand new Canon very rarely has even one hot pixel. This is the standard Pentax should and will be held to.

05-15-2008, 05:44 PM	#112
KungPOW Veteran Member Join Date: Nov 2007 Location: Vancouver, Canada Photos: Gallery Posts: 1,699	I understand the K20D has a feature to map out hot pixels. Can the K20D turn this feature on and off, so that if you wanted to, you could see the hot pixels? Is it possible that the hot pixel issue is just a matter of the pixel map being turned off on some cameras? Another question: Did Pentax add the pixel map feature because they knew the CMOS sensor was prone to hot pixels, or did they add the mapping, as a feature, and are now being burnt by their own upgrade? The question is a litle retorical. My gut tells me that this hot pixel issue is more of a matter that various imaging software packages need to be updated to deal with K20D files. The software needs to be updated to repair the hot pixels found in the K20D. Having said that, I do not think that the K20D has a greater or lesser chance of a HP then does the "Golden Cann" (sounds of angels singing), or the "Mighty Nikn" (sounds of victory trumpets). I expect the chance of a HP on a Canikon is about the same as on a Pentax.

05-15-2008, 05:59 PM	#113
Lance B Veteran Member Join Date: Nov 2006 Location: Sydney, Australia Photos: Gallery Posts: 593	Originally posted by mconder Ya, but I have owned several Canon's and I can say from experience, a brand new Canon very rarely has even one hot pixel. This is the standard Pentax should and will be held to. I have not had any of the dreaded hot pixels with any of my Pentax cameras either. You say you haven't had any on your Canons either, but is that because they are automatically removed by software, either in camera or in post process?

05-15-2008, 06:27 PM	#114
Maxington Senior Member Join Date: Sep 2007 Location: Australia Posts: 255	Originally posted by KungPOW I understand the K20D has a feature to map out hot pixels. Can the K20D turn this feature on and off, so that if you wanted to, you could see the hot pixels? Is it possible that the hot pixel issue is just a matter of the pixel map being turned off on some cameras? Another question: Did Pentax add the pixel map feature because they knew the CMOS sensor was prone to hot pixels, or did they add the mapping, as a feature, and are now being burnt by their own upgrade? The question is a litle retorical. My gut tells me that this hot pixel issue is more of a matter that various imaging software packages need to be updated to deal with K20D files. The software needs to be updated to repair the hot pixels found in the K20D. Having said that, I do not think that the K20D has a greater or lesser chance of a HP then does the "Golden Cann" (sounds of angels singing), or the "Mighty Nikn" (sounds of victory trumpets). I expect the chance of a HP on a Canikon is about the same as on a Pentax. I'd guess that its similar to the AF adjust feature, its an ability that has always been available to repairers, but now Pentax has enabled it for end users. Which is a very handy thing to do. I doubt its just because of the new sensor. Maybe they listened to everyone wanting to use the repairers K10D AF adjust that got removed in 1.30 firmware? I suspect that they wouldn't mind users adjusting AF, they just didn't have a failsafe setup for the K10D, too risky.

05-15-2008, 07:33 PM	#115
*isteve Veteran Member Join Date: Nov 2006 Location: London, England Posts: 1,187	Originally posted by codiac2600 Until there is hard proof of this being a true problem, which as of right now it's not, let's not have 50 million threads on the same stupid subject. If you have a K20D with said problem please post in this thread a Jpeg straight from the camera with proof of hot pixels so we can look into it. BTW, EXIF data must be intact in order to make sure everything is on the up and up. So far I tested all 5 K20D's we have in the store and no hot pixels, had friends and costumers email me jpegs and nothing. My tally is officially 14 no hot pixel cameras and the likely hood DPR got two of them in a row is looking slimmer and slimmer. So if you have proof show it and I don't care if you have a blog saying it's true, we are looking for proof to either make a case about it or drop it at that. Clearly his phenopmenon is happening under some circumstance, but in Phils uinhelpful way we dont know... 1. What cicumstances? I have used live view for minutes at a time when doing lens tests and I have no hot pixels. Strikes me you have to use it for extended periods before these pixels show up. Its warns against this in the manual. 2. If the pixels are different in each shot, do they eventually reset when it cools down or do some of them stay hot? If th sensor can recover its hardly a problem is it? Phils rather flippant remark about using LV "a bit" is clearly disingeneous and particularly unhelpful. I suspect if we knew the exact circumstances it would make it easier to replicate.

05-15-2008, 08:18 PM	#116
jfdavis58 Site Supporter Join Date: Nov 2006 Location: 13 S 0357397-3884316 Photos: Gallery Posts: 887	.......... Wow! ..........

05-16-2008, 08:10 AM	#117
rfortson Veteran Member Join Date: Sep 2006 Location: Houston TX Photos: Gallery Posts: 2,129	Originally posted by falconeye If you read my post #86, you'll see me claiming that PPL removes hot pixels as well. That's part of the reason I posted my testing method. My claim is actually based on test: I don't see the hot pixels in PPL that I see in PhotoMatix (cf. my images in quoted post). Updated info: If you make a large 100% loupe window (which is awfully slow to update), you'll see intermediate processing stages within PPL (due to its slowness). Then I can clearly state here that my hot blue pixel (cf. above) (when push brightness up a bit): Appears, then disappears in the loupe window during processing in PPL. This clearly proves that PPL eliminates hot pixels in one of its processing stages. (Maybe PPL didn't do that in the past.) That's good to know. Are you sure you didn't have any of the noise reduction enabled? I know it didn't do it in the past because I took a couple of long (multi-second) exposures with my K100D and had a couple of stuck pixels. PPL didn't remove them, but ACR did. I wish the PPL software had better documentation. I like it and find more and more that it does. I just have to find out by trial and error, or by user reports like yours. Personally, I think it's a good thing that software removes hot pixels. I suspect lots more hot pixel removal is going on in all cameras. Manufacturers just don't advertise it as it's seen as a "defect".

05-16-2008, 08:14 AM	#118
rfortson Veteran Member Join Date: Sep 2006 Location: Houston TX Photos: Gallery Posts: 2,129	Originally posted by mconder Ya, but I have owned several Canon's and I can say from experience, a brand new Canon very rarely has even one hot pixel. This is the standard Pentax should and will be held to. I can say by the same experience that the several Pentax cameras that I've owned have rarely had even one hot pixel. My K20D has exactly none.

05-16-2008, 10:10 AM	#119
falconeye Veteran Member Join Date: Jan 2008 Location: Munich, Alps, Germany Photos: Gallery Posts: 6,871	Originally posted by rfortson Are you sure you didn't have any of the noise reduction enabled? I used it in its default setting. There it has: - Noise suppression: 0. - Correction of wrong color information: >0 (dont remember exactly, but ON). (both translated back from my German GUI) I don't know what the second setting really means. If it removes hot pixels: fine

05-16-2008, 12:01 PM	#120
rfortson Veteran Member Join Date: Sep 2006 Location: Houston TX Photos: Gallery Posts: 2,129	Originally posted by falconeye I used it in its default setting. There it has: - Noise suppression: 0. - Correction of wrong color information: >0 (dont remember exactly, but ON). (both translated back from my German GUI) I don't know what the second setting really means. If it removes hot pixels: fine I don't really know what that does either. It's been there for several versions, so I don't think it's the "magic hot pixel removal" slider. More likely, as you said, they added it to the converter and just didn't mention it.