Any hot pixels on your K-5?

falconeye · 11-10-2010, 05:50 AM

Originally posted by Class A

I see your point that the original hot pixel map cannot be used.

However, the original hot pixel map could be used by using a mapping of the subsampled pixel positions to their original positions.

A less space-efficient but quicker method would be to use a dedicated hot pixel map for video (would probably have to be computed on the fly at the beginning of a video recording because it differs for all video modes).

You are right. But it's a bit more awkward because the nearest sensel of the same color which is read out too, can be far away and the pixel mapping algorithm would be different. I think they simply didn't do it. Or maybe, Pentax now did for the K-5 and Nikon didn't.

John Shriver · 11-10-2010, 08:31 PM

Every sensor from every manufacturer is tested for pixel uniformity. There are selection grades for how uniform they are. That's why they are so expensive -- lots of them go in the "too uneven" bin (trash).

Making analog chips consistently is much harder than digital chips -- and CCD and CMOS sensors are most definitely analog chips.

Kubicide · 11-12-2010, 12:29 AM

I'm new to the Pentax club having just purchased a K-5, and it has a small cluster of hot pixels visible during video. Disappointing but it happens. It's going back for exchange. Pixels are redish in color located near the middle lower left of center and actually quite visible during playback.

Didn't bother mapping since it is new and I simply want to see a better sensor out of the box. Mapping the cluster (if it worked since the algorithms usually limit the size of the correction as to not map out a false area) means they will show false info or copied colors from neighboring pixels which could turn this cluster into what looks like a dust spot.

(Edit: new camera looks good - no hot pixels.)

falconeye · 11-12-2010, 08:58 AM

Originally posted by Kubicide

I'm new to the Pentax club having just purchased a K-5, and it has a small cluster of hot pixels visible during video. Disappointing but it happens. It's going back for exchange. Pixels are redish in color located near the middle lower left of center and actually quite visible during playback.

And here goes my theory of supreme in-production line sensor testing :ugh:

Keep us posted about the next one

JayBee · 11-13-2010, 07:27 PM

Just checking some long exposures from tonight; in one shot 23 sec ISO 400, I count 3 hot pixels. DFS was turned off. I don't consider that to be a serious problem. I had a few hot pixels in my K0D; never noticed any with my K20D, but that was probably because DFS got rid of them before the image could be displayed.

falconeye · 11-13-2010, 07:45 PM

Originally posted by JayBee

Just checking some long exposures from tonight; in one shot 23 sec ISO 400, I count 3 hot pixels. DFS was turned off.

That's not the sort of hot pixels we're talking about. They can and should be mapped out. Every camera has a few. Some old threads exist where users counted the #pixels above a threshold in lens cap shots, done around the time the K20D came out.

The hot pixels talked about here are those which are can be distracting in a video with dark footage - and cannot be mapped out with the current firmware.

So far, only one such report and camera returned for replacement.

Raybo · 11-13-2010, 08:08 PM

I'll be honest.

Having owned a K100DS and K20D I have not noticed" any hot OR stuck pixels, but than again I have not "peeped" to find any.
If I just happened to be the owner of the D7000 that I refered to in my original post I think that I would have noticed them without "peeping".

JayBee · 11-13-2010, 08:48 PM

Originally posted by falconeye

That's not the sort of hot pixels we're talking about. They can and should be mapped out. Every camera has a few. Some old threads exist where users counted the #pixels above a threshold in lens cap shots, done around the time the K20D came out.

The hot pixels talked about here are those which are can be distracting in a video with dark footage - and cannot be mapped out with the current firmware.

So far, only one such report and camera returned for replacement.

Ah. I was responding to the OP's original question in general - any hot pixels in the K-5?. I have seen posted D7000 images (stills, not videos) that had a lot of hot pixels.

But... knowing about the hot pixels that showed up in my stills tonight, I shot a few low-light videos just now to see how the K-5 fared. Just fine. Even in the areas of the shot where the hot pixels had been visible in the stills from earlier tonight.
(Again - those are not a problem - they disappear as soon as you open the image in any photo editing software. If you open in a photo viewer such as FastStone, then you get to see the pixels)

65535 · 11-14-2010, 01:06 AM

That's pretty bad. However I still find the photo stunning.

**rod_grant** · 11-14-2010, 05:32 AM

Yes, it is a stunning photo. Normal size there appear to be a few stars, and it looks natural. Enlarged they look a bit too rectangular for stars!!

Phoenix Smasher · 11-15-2010, 12:44 AM

Originally posted by falconeye

So far, we have no reports of hot pixels in video.

You do now.

You Tube

I purchased it about a week ago, so far I've loved it, but tonight I noticed that when shooting HD video, I have these 2 pixels, one in the upper right and another in the lower left of center. I will be calling their warranty and repair center tomorrow afternoon.

farhagh · 11-15-2010, 07:51 AM

Originally posted by buaobao

Wow ! it produce "so many stars".

I guess the stars in the sky, and even the reflection in water are OK. Am not sure about the stars on the beach though...

farhagh · 11-15-2010, 07:53 AM

Originally posted by Cannikin

I believe the appropriate quote is:

"My God, it's full of stars!"

and that's why all superstars prefer Nikon!

melander · 11-15-2010, 09:52 AM

Originally posted by Phoenix Smasher

You do now.
YouTube - Pentax K-5 image sensor problem.
I purchased it about a week ago, so far I've loved it, but tonight I noticed that when shooting HD video, I have these 2 pixels, one in the upper right and another in the lower left of center. I will be calling their warranty and repair center tomorrow afternoon.

Do a manual pixelmapping in the camera menu?

Should do the trick

Phoenix Smasher · 11-15-2010, 10:49 AM

Originally posted by melander

Do a manual pixelmapping in the camera menu?

Should do the trick

I've tried it over and over. Still there.

11-12-2010, 12:29 AM	#33
Kubicide Junior Member Join Date: Nov 2010 Posts: 25	I'm new to the Pentax club having just purchased a K-5, and it has a small cluster of hot pixels visible during video. Disappointing but it happens. It's going back for exchange. Pixels are redish in color located near the middle lower left of center and actually quite visible during playback. Didn't bother mapping since it is new and I simply want to see a better sensor out of the box. Mapping the cluster (if it worked since the algorithms usually limit the size of the correction as to not map out a false area) means they will show false info or copied colors from neighboring pixels which could turn this cluster into what looks like a dust spot. (Edit: new camera looks good - no hot pixels.) Last edited by Kubicide; 11-12-2010 at 12:40 PM. Reason: Updated with new camera

11-13-2010, 07:27 PM	#35
JayBee Senior Member Join Date: Sep 2008 Location: Manitoba Photos: Gallery Posts: 119	Just checking some long exposures from tonight; in one shot 23 sec ISO 400, I count 3 hot pixels. DFS was turned off. I don't consider that to be a serious problem. I had a few hot pixels in my K0D; never noticed any with my K20D, but that was probably because DFS got rid of them before the image could be displayed. Last edited by JayBee; 11-13-2010 at 07:39 PM.

11-14-2010, 05:32 AM	#40
rod_grant Site Supporter Join Date: Apr 2010 Location: Wangaratta, Victoria Photos: Gallery Posts: 6,951	Yes, it is a stunning photo. Normal size there appear to be a few stars, and it looks natural. Enlarged they look a bit too rectangular for stars!! Last edited by rod_grant; 11-15-2010 at 12:00 AM.

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Hot pixels - K7	shininglight	Troubleshooting and Beginner Help	1	10-31-2009 08:13 PM
Hot Pixels?	prinze18	Troubleshooting and Beginner Help	4	10-09-2009 11:22 AM
Hot Pixels	GregK8	Troubleshooting and Beginner Help	8	03-04-2009 07:17 AM
hot pixels?	laissezfaire	Pentax DSLR Discussion	14	10-02-2008 07:22 PM
Hot pixels	richard64	Pentax DSLR Discussion	4	08-05-2008 05:50 PM

11-10-2010, 05:50 AM	#31
falconeye Veteran Member Join Date: Jan 2008 Location: Munich, Alps, Germany Photos: Gallery Posts: 6,871	Originally posted by Class A I see your point that the original hot pixel map cannot be used. However, the original hot pixel map could be used by using a mapping of the subsampled pixel positions to their original positions. A less space-efficient but quicker method would be to use a dedicated hot pixel map for video (would probably have to be computed on the fly at the beginning of a video recording because it differs for all video modes). You are right. But it's a bit more awkward because the nearest sensel of the same color which is read out too, can be far away and the pixel mapping algorithm would be different. I think they simply didn't do it. Or maybe, Pentax now did for the K-5 and Nikon didn't.

11-10-2010, 08:31 PM	#32
John Shriver Senior Member Join Date: Apr 2009 Location: Arlington, MA Posts: 160	Every sensor from every manufacturer is tested for pixel uniformity. There are selection grades for how uniform they are. That's why they are so expensive -- lots of them go in the "too uneven" bin (trash). Making analog chips consistently is much harder than digital chips -- and CCD and CMOS sensors are most definitely analog chips.

11-12-2010, 08:58 AM	#34
falconeye Veteran Member Join Date: Jan 2008 Location: Munich, Alps, Germany Photos: Gallery Posts: 6,871	Originally posted by Kubicide I'm new to the Pentax club having just purchased a K-5, and it has a small cluster of hot pixels visible during video. Disappointing but it happens. It's going back for exchange. Pixels are redish in color located near the middle lower left of center and actually quite visible during playback. And here goes my theory of supreme in-production line sensor testing :ugh: Keep us posted about the next one

11-13-2010, 07:45 PM	#36
falconeye Veteran Member Join Date: Jan 2008 Location: Munich, Alps, Germany Photos: Gallery Posts: 6,871	Originally posted by JayBee Just checking some long exposures from tonight; in one shot 23 sec ISO 400, I count 3 hot pixels. DFS was turned off. That's not the sort of hot pixels we're talking about. They can and should be mapped out. Every camera has a few. Some old threads exist where users counted the #pixels above a threshold in lens cap shots, done around the time the K20D came out. The hot pixels talked about here are those which are can be distracting in a video with dark footage - and cannot be mapped out with the current firmware. So far, only one such report and camera returned for replacement.

11-13-2010, 08:08 PM	#37
Raybo Veteran Member Join Date: Dec 2007 Posts: 872 Original Poster	I'll be honest. Having owned a K100DS and K20D I have not noticed" any hot OR stuck pixels, but than again I have not "peeped" to find any. If I just happened to be the owner of the D7000 that I refered to in my original post I think that I would have noticed them without "peeping".

11-13-2010, 08:48 PM	#38
JayBee Senior Member Join Date: Sep 2008 Location: Manitoba Photos: Gallery Posts: 119	Originally posted by falconeye That's not the sort of hot pixels we're talking about. They can and should be mapped out. Every camera has a few. Some old threads exist where users counted the #pixels above a threshold in lens cap shots, done around the time the K20D came out. The hot pixels talked about here are those which are can be distracting in a video with dark footage - and cannot be mapped out with the current firmware. So far, only one such report and camera returned for replacement. Ah. I was responding to the OP's original question in general - any hot pixels in the K-5?. I have seen posted D7000 images (stills, not videos) that had a lot of hot pixels. But... knowing about the hot pixels that showed up in my stills tonight, I shot a few low-light videos just now to see how the K-5 fared. Just fine. Even in the areas of the shot where the hot pixels had been visible in the stills from earlier tonight. (Again - those are not a problem - they disappear as soon as you open the image in any photo editing software. If you open in a photo viewer such as FastStone, then you get to see the pixels)

11-14-2010, 01:06 AM	#39
65535 Senior Member Join Date: Dec 2008 Location: Canada Posts: 187	That's pretty bad. However I still find the photo stunning.

11-15-2010, 07:51 AM	#42
farhagh Veteran Member Join Date: Nov 2008 Location: Ohio Photos: Albums Posts: 339	Originally posted by buaobao Wow ! it produce "so many stars". I guess the stars in the sky, and even the reflection in water are OK. Am not sure about the stars on the beach though...

11-15-2010, 07:53 AM	#43
farhagh Veteran Member Join Date: Nov 2008 Location: Ohio Photos: Albums Posts: 339	Originally posted by Cannikin I believe the appropriate quote is: "My God, it's full of stars!" and that's why all superstars prefer Nikon!

11-15-2010, 09:52 AM	#44
melander Veteran Member Join Date: Jul 2008 Location: Sweden, Umea Posts: 876	Originally posted by Phoenix Smasher You do now. YouTube - Pentax K-5 image sensor problem. I purchased it about a week ago, so far I've loved it, but tonight I noticed that when shooting HD video, I have these 2 pixels, one in the upper right and another in the lower left of center. I will be calling their warranty and repair center tomorrow afternoon. Do a manual pixelmapping in the camera menu? Should do the trick

11-15-2010, 10:49 AM	#45
Phoenix Smasher New Member Join Date: Nov 2010 Posts: 4	Originally posted by melander Do a manual pixelmapping in the camera menu? Should do the trick I've tried it over and over. Still there.