user controlled D800 hot pixel remapping: details

soloryb · Aug 24, 2013

Thanks for the shot.

OK Bernard, though I'm not fully convinced, you've got me concerned enough not to dismiss the cosmic ray sensor pixel killing theory completely. I'm motivated to do some more research now. And, I double cleaned my sensor, just in case.

BTW, couldn't those dark side of the moon 'bad' pixel just be hot pixels from an extra-long exposure? I've had such hot pixels when shooting moon images with long exposures.

Bernard Delley said:
Bernard Delley wrote:

Bernard Delley said:

soloryb wrote:

OK post your shots. I'm not afraid to learn something new.

Bernard Delley said:

Bernard Delley wrote:

If I dig them out, I can post my findings for a recent flight forth and back to Japan.

Of course these effects from radiation are not dramatic, its a few pixels that become sick out of millions. Also for humans its not dramatic, air personnel does not have a noticeably shorter life than the rest of us.

Click to expand...

Click to expand...

looks like the only copy of the data from that test are on my recently died computer...

You should not expect pictures though. (pictures of black frames have been shown before, with proper contrast they look like a starry sky except you do not find constellations and the milky way)

Anyway, this image shows a very strong hot pixel showing up in jpg after a transatlantic flight. Was not there the evening before the flight. This false red light was always in the same location with respect to the sensor, not the image. It could not be seen in low ISO images when this spot was not too dark. Remapping at the service center made it finally disappear.

This post by somebody else shows many hot pixels aka stars

http://www.dpreview.com/forums/post/52000388

Why? There are false stars shining through the dark side of the moon! The folse stars are clearly sharper than the slightly blurred moon - optically imaged stars should have the same blur! Real stars could not be recorded with the mentioned exposure!

For technical purposes it is better to "look" through loss-less raw files using a computer program to spot the few to hundreds of excessively bright pixels in the millions of well behaved pixels. Of course a bright pixel may hit any of the Bayer RGB colors.

Bernard Delley · Aug 25, 2013

soloryb said:
soloryb wrote:

OK Bernard, though I'm not fully convinced, you've got me concerned enough not to dismiss the cosmic ray sensor pixel killing theory completely. I'm motivated to do some more research now. And, I double cleaned my sensor, just in case.

BTW, couldn't those dark side of the moon 'bad' pixel just be hot pixels from an extra-long exposure? I've had such hot pixels when shooting moon images with long exposures.

hot pixels act as with a built in constant light source: exposure goes in proportion to time and ISO. If you put the good thermal noise pixels below the black level, you may see lots of the hot pixels as int that moon example. Most of these tend to be relatively weakly hot pixels. Nikon now removes the brightest, the ones below threshold remain. Since you report haven seen such hot pixels, you can easily set up a test shot -- leaving the moon away. Unfortunately, you have already removed the ones above threshold. You can see the fainter now, but you cannot check the effectiveness of the procedure. -- Just wait a year or till after an intercontinental flight and you will likely have one, two ... fresh quite hot ones.

Bernard Delley · Aug 26, 2013

The table below shows the summary of hot pixels found in raw black frames of a Nikon D7000, 3 frames taken before a long distance flight to Japan, another 3 were taken in Japan before the flight back and then 3 frames at home right after the flight back.

The top header number is the image number for the column below. Columns x y are the pixel location. The main colums show the lossless raw values (14bit) of that pixel at x y. A "-1" means "black: a low number of no interest", characterized by the typical distribution RMS=20 at ISO 1600)

The first line shows a pixel going "hot" during flight #1. After the return flight #2 , two further hot pixels appeared. Hot pixels usually stay hot, though the table shows that their variation over longer periods of time can be larger than expected from Poisson statistics. In more rare instances hot pixels can even become normal again (room temperature annealing of defect).

With Nikons user procedure, the main theme of this thread, 8 hot pixels would be remapped in this example after flight #2, while 3 listed here would remain untouched below the threshold. If -- if the procedure were available on D7000, which is not the case.

Bernard Delley · Aug 28, 2013

labalaba said:
labalaba wrote:

I'd like to hear

I could test D300 black frames: the double sensor clean procedure hat no effect on its hot pixels - unlike new models.

from my experience with sensor cleaning at the swiss Nikon service center (Excellent!) at the same time hot pixels were remapped and firmware updated without asking for this. Perhaps its standard procedure.

rubank · Aug 28, 2013

Bernard Delley said:
Bernard Delley wrote:

Bernard Delley said:

labalaba wrote:

I'd like to hear

Click to expand...

I could test D300 black frames: the double sensor clean procedure hat no effect on its hot pixels - unlike new models.

from my experience with sensor cleaning at the swiss Nikon service center (Excellent!) at the same time hot pixels were remapped and firmware updated without asking for this. Perhaps its standard procedure.

http://www.dpreview.com/forums/post/26855191

worked for me (really)

Bernard Delley · Aug 30, 2013

rubank said:
rubank wrote:

http://www.dpreview.com/forums/post/26855191

worked for me (really)

If you still have the D300, It would be interesting to know if it can be made to work. The procedure in the post you mention is clearly different, so nothing is proved by my findigs for somebody else's D300 using the D800 procedure. That former post met a lot of criticism -- I will try the 'old' procedure on my D7000.

To know for sure one needs a black frame before and a second one after the procedure.

exposure 1/5 sec ISO 1000 NEF/RAW is good for this test D300. No big deal to take such black frames. You probably can see some false stars by ViewNX with exposure +2 . If they are still there for the frame after the procedure: tough luck - does not work.

If you can make such black frames, and know how ftp works, please deposit the two NEF at

ftp ftp.psi.ch "anonymous" cd PSI_incoming/delley put ...

please let me know if you did - before the ftp data gets erased again.

I will "look" through the numbers in the raw and give feedback.

Bernard Delley · Sep 10, 2013

a new D5200 had 3 hot pixels (stronger than 1000 NEF/raw units at at ISO1600 1/5s)

after a trip Zurich-Vancouver-Zurich it had 5 additional hot pixels,

which were removed by the procedure of this thread :-D

MartinPhotographer · Jun 9, 2014

Bernard Delley said:
a new D5200 had 3 hot pixels (stronger than 1000 NEF/raw units at at ISO1600 1/5s)

after a trip Zurich-Vancouver-Zurich it had 5 additional hot pixels,

which were removed by the procedure of this thread :-D

Hot pixel removal procedure did worked for Nikon d5300. From very first tests it seems that my hot pixel (I had only one at that moment) is gone. Its still visible during super long exposure (~1 min), super high iso (Hi 1) and with cap on test, but most pixels seem hot as a result of this test anyway. Most importantly, its not visible on normal photos which I took quickly just after the cleaning procedure for quick tests.

MrocArt · Apr 26, 2015

I thank you for this post. Saved me from spending hundreds on either buying a new camera or for the repair to my sensor. Hotpixels are gone. Thanks so much! Glad I googled how to and stumbled onto this site and forum. Now I'm a member because this meant so much!

Alan Mermelstein · May 17, 2016

I know this is an old post but I just want to thank you. I had 2 hot pixels in my new D500 and this technique cleared them up!!!

Oswaldo Ortega · Nov 15, 2016

I dont understand where are the information for the proceadure you are talking about, can you please tell me how you fix your D5200 hot pixels?

Bernard Delley · Dec 14, 2016

Just tested the procedure on a D7200 with 32 hot pixels.

It turns out, that a simplified procedure has been implemented on later models since the original post. But, never mind the old procedure works also on new models.

The simplified procedure to remap hot pixels is:

menu / setup menu / Clean image sensor / clean now (activate once)

In addition to the recommended automated (dust shake) clean at startup/shutdown, the "clean now" function includes the hot pixel detection + remapping procedures

the procedure for older models is (activate "clean now" twice in quick succession)

Bernard Delley · Dec 18, 2016

Bernard Delley said:
Just tested the procedure on a D7200 with 32 hot pixels.

It turns out, that a simplified procedure has been implemented on later models since the original post. But, never mind the old procedure works also on new models.

The simplified procedure to remap hot pixels is:

menu / setup menu / Clean image sensor / clean now (activate once)

In addition to the recommended automated (dust shake) clean at startup/shutdown, the "clean now" function includes the hot pixel detection + remapping procedures

the procedure for older models is (activate "clean now" twice in quick succession)

The D500 works with the simplified procedure.

Tony Vinto · Jul 5, 2018

Glad I ran into this thread! My D500 seems to have developed a hot pixel so I will be doing this tonight.

One question... does the camera give any feedback that it did something? ...or do I just test for the bad pixel to see if it is gone?

Thanks

Bernard Delley · Jul 5, 2018

Tony Vinto said:
Glad I ran into this thread! My D500 seems to have developed a hot pixel so I will be doing this tonight.

One question... does the camera give any feedback that it did something? ...or do I just test for the bad pixel to see if it is gone?

Thanks

Tony Vinto · Jul 6, 2018

Thanks Bernard.

I am traveling this weekend so I will take you up on your offer next week.

Same FTP as you mentioned in some of your earlier posts?

Tony

Bernard Delley · Jul 8, 2018

Tony Vinto said:
Same FTP as you mentioned in some of your earlier posts?

sorry, that appears to be canceled now. I will inquire eventually it this is meant to be so, security paranoia.

Jundar73 · May 13, 2019

Bernard Delley said:
soloryb said:

OK Bernard, though I'm not fully convinced, you've got me concerned enough not to dismiss the cosmic ray sensor pixel killing theory completely. I'm motivated to do some more research now. And, I double cleaned my sensor, just in case.

BTW, couldn't those dark side of the moon 'bad' pixel just be hot pixels from an extra-long exposure? I've had such hot pixels when shooting moon images with long exposures.

Click to expand...

hot pixels act as with a built in constant light source: exposure goes in proportion to time and ISO. If you put the good thermal noise pixels below the black level, you may see lots of the hot pixels as int that moon example. Most of these tend to be relatively weakly hot pixels. Nikon now removes the brightest, the ones below threshold remain. Since you report haven seen such hot pixels, you can easily set up a test shot -- leaving the moon away. Unfortunately, you have already removed the ones above threshold. You can see the fainter now, but you cannot check the effectiveness of the procedure. -- Just wait a year or till after an intercontinental flight and you will likely have one, two ... fresh quite hot ones.

Have Nikon D800E since 2012, had 3 or 4 intercontinental flights (USE - Europe) , no dead pixels yet.

Bernard Delley · May 13, 2019

Jundar73 said:
Bernard Delley said:

soloryb said:

OK Bernard, though I'm not fully convinced, you've got me concerned enough not to dismiss the cosmic ray sensor pixel killing theory completely. I'm motivated to do some more research now. And, I double cleaned my sensor, just in case.

BTW, couldn't those dark side of the moon 'bad' pixel just be hot pixels from an extra-long exposure? I've had such hot pixels when shooting moon images with long exposures.

Click to expand...

hot pixels act as with a built in constant light source: exposure goes in proportion to time and ISO. If you put the good thermal noise pixels below the black level, you may see lots of the hot pixels as int that moon example. Most of these tend to be relatively weakly hot pixels. Nikon now removes the brightest, the ones below threshold remain. Since you report haven seen such hot pixels, you can easily set up a test shot -- leaving the moon away. Unfortunately, you have already removed the ones above threshold. You can see the fainter now, but you cannot check the effectiveness of the procedure. -- Just wait a year or till after an intercontinental flight and you will likely have one, two ... fresh quite hot ones.

Click to expand...

Have Nikon D800E since 2012, had 3 or 4 intercontinental flights (USE - Europe) , no dead pixels yet.

good for you! Dead and hot pixels are different things for me. Dead and stuck pixels are typical for screens. Easy to see and not so frequent any more fortunately. Hot pixels are a sensor issue. I am pretty sure yours must have some unless you regularly act with the sensor cleaning menu. See the rest of this thread if you are somehow interested in this theme. If you have not mapped the into hiding by "cleaning" your sensor via the menu, then I could tell you from a black frame raw file at 1/4s ISO 1600.

mjevans · May 11, 2021

I know this is a very old thread but my hot pixels (on my D500) don't seem to go away after 1 clean now or even 2 quick clean now processes.

I turned off long exposure NR.

Set ISO 6400 and 3s exposure.

Left lens cap on.

I see 1 large red pixel at 100% zoom in CNXD and 4 small white pixels that don't go away after the above process even after taking 5-10 shots after the cleaning process.

Am I doing something wrong?

BTW, I've seen the large red pixel in actual shots.

Martin

user controlled D800 hot pixel remapping: details

Senior Member

Senior Member

Senior Member

Attachments

Senior Member

Senior Member

Senior Member

Senior Member

New member

New member

Member

New member

Senior Member

Senior Member

Active member

Senior Member

Active member

Senior Member

Member

Senior Member

Leading Member

Similar threads

Keyboard shortcuts