Jump to content

  •  

CNers have asked about a donation box for Cloudy Nights over the years, so here you go. Donation is not required by any means, so please enjoy your stay.

Photo

Tons of Stuck/Hot Pixels on Canon 6D Images

astrophotography beginner CMOS dslr equipment imaging
  • Please log in to reply
17 replies to this topic

#1 benringel22

benringel22

    Lift Off

  • -----
  • topic starter
  • Posts: 12
  • Joined: 07 Jun 2019
  • Loc: Peoria, IL

Posted 24 August 2019 - 01:30 AM

Hello everyone,

 

I recently purchased a lightly used Canon 6D on eBay as my first real astrophotography DSLR (I am very much a beginner...). This camera, when I received it today, had less than 8,000 shutter actuations, and had no signs of wear or damage. The first thing I did when getting the camera was to test it for defects and issues, and it seemed to pass my initial (newbie) inspection.

 

It was not until tonight while attempting to image a nebula that I noticed the images I was capturing had tons of white, blue, and red speckles covering the image (which clearly was not normal noise). This obviously scared me, and I did much research to find out that the camera sensor has many stuck/hot pixels. I attempted many fixes, such as the popular manual clean sensor method; however, attempt after attempt left me with the same result. It seems that it is often the same pixels causing the issue, and this makes me think that the sensor may have more than just stuck/hot pixels.

 

I know that a few of these pixels may be unavoidable and can be fixed in post processing (with dark frames, etc.), but this many affected pixels seems to me to be a very serious issue, and I am out of ideas to solve the problem without spending more money (on what was supposed to be an affordable camera...). Does anyone know what may be wrong with my camera sensor and if there is a way I can fix it?

 

Thank you all so much in advance for your help!

Ben

 

P.S. Below is an attached image of a 30 second dark frame at ISO 1600. I'm not sure if the pixels will show up for others, but the file on my computer shows an obvious and large number of these pixels, even without zooming in....

 

Stuck/Hot Pixels (JPEG)

 



#2 cray2mpx

cray2mpx

    Vostok 1

  • *****
  • Posts: 176
  • Joined: 06 Jan 2013
  • Loc: Vienna, Austria

Posted 24 August 2019 - 02:06 AM

can‘t open the image to have look ...



#3 benringel22

benringel22

    Lift Off

  • -----
  • topic starter
  • Posts: 12
  • Joined: 07 Jun 2019
  • Loc: Peoria, IL

Posted 24 August 2019 - 02:09 AM

Hmm, okay how about this one? I think I may have had it set as a private album.
 
Stuck/Hot Pixels (JPEG)


#4 Xplode

Xplode

    Explorer 1

  • -----
  • Posts: 58
  • Joined: 16 Nov 2011

Posted 24 August 2019 - 02:21 AM

I can't see the image either.
You should upload a cr2 file, hot pixels gets removed when images are downscaled to jpg.

Did you do the test indoors?
Noise is lowered a lot when you go outdoors, dark noise halves for every 6-7C drop.
A solution to get rid of hot pixels and dark noise in DSLR's is to dither your images.
I've used the 6D for several years with great success, it's a good DSLR for astro imaging smile.gif


Edited by Xplode, 24 August 2019 - 02:24 AM.


#5 james7ca

james7ca

    Fly Me to the Moon

  • *****
  • Posts: 7338
  • Joined: 21 May 2011
  • Loc: San Diego, CA

Posted 24 August 2019 - 02:32 AM

With long exposures and at room or outdoor summer temperatures it's normal to have a "ton" of bright pixel artifacts. So, it's likely there is nothing wrong with your camera. However, if you calibrate your images with a good dark field master you'll likely see FAR fewer artifacts.


  • Jim Waters and RedLionNJ like this

#6 benringel22

benringel22

    Lift Off

  • -----
  • topic starter
  • Posts: 12
  • Joined: 07 Jun 2019
  • Loc: Peoria, IL

Posted 24 August 2019 - 02:33 AM

I've attempted to upload the CR2 file many times now; unfortunately CN says that the file cannot be uploaded because it is too big. I've attached a screenshot of the zoomed in version of the JPG image I showed earlier. For some reason, the JPG actually has more hot pixels than the CR2, and even this does not show what it looks like from my camera and computer.

 

Zoom In: Stuck/Hot Pixels

 



#7 benringel22

benringel22

    Lift Off

  • -----
  • topic starter
  • Posts: 12
  • Joined: 07 Jun 2019
  • Loc: Peoria, IL

Posted 24 August 2019 - 02:38 AM

Okay, this is good to know. It is mostly concerning because I am getting very bright and noticeable hot pixels on every long exposure frame with the Canon 6D, which never happened with a much lower quality Nikon D3300 I was using. I did side by side dark frames with each camera, and the Nikon D3300 never produces a single hot pixel this bright... To me this still seems slightly alarming, but maybe it shouldn't be?



#8 Xplode

Xplode

    Explorer 1

  • -----
  • Posts: 58
  • Joined: 16 Nov 2011

Posted 24 August 2019 - 02:48 AM

Nikon cheats by supressing the noise when you take darks so darks are pointless for many Nikon cameras.
In my opinion darks aren't needed for 6D either as long as you dither your images.
You can see the effect temperature has on dark noise on the 6D here (tests done by myself)
https://cdn.astrobin...0_watermark.jpg
 
Same image with gamma +3 to increase the difference for shorter exposures and colder temperatures
https://cdn.astrobin...0_watermark.jpg
  • Jim Waters likes this

#9 benringel22

benringel22

    Lift Off

  • -----
  • topic starter
  • Posts: 12
  • Joined: 07 Jun 2019
  • Loc: Peoria, IL

Posted 24 August 2019 - 02:52 AM

Thank you for the post! That is similar to what my frames look like; however, I am getting significant and noticeable hot pixels at 30 seconds and ISO 1600. Would you say that is normal as well?

 

Thanks again!



#10 Jim Waters

Jim Waters

    Mercury-Atlas

  • *****
  • Posts: 2783
  • Joined: 21 Oct 2007
  • Loc: Phoenix, AZ USA

Posted 24 August 2019 - 02:55 AM

What was the ambient temperature?  The 6D is relatively good.  You are lucky you don't have a T3i.



#11 Jim Waters

Jim Waters

    Mercury-Atlas

  • *****
  • Posts: 2783
  • Joined: 21 Oct 2007
  • Loc: Phoenix, AZ USA

Posted 24 August 2019 - 02:56 AM

Nikon cheats by supressing the noise when you take darks so darks are pointless for many Nikon cameras.
In my opinion darks aren't needed for 6D either as long as you dither your images.
You can see the effect temperature has on dark noise on the 6D here (tests done by myself)
https://cdn.astrobin...0_watermark.jpg
 
Same image with gamma +3 to increase the difference for shorter exposures and colder temperatures
https://cdn.astrobin...0_watermark.jpg


Really like your post / images.  Yes - always dither........

#12 benringel22

benringel22

    Lift Off

  • -----
  • topic starter
  • Posts: 12
  • Joined: 07 Jun 2019
  • Loc: Peoria, IL

Posted 24 August 2019 - 02:59 AM

It was a much cooler night than usual (which is why I was surprised). The ambient temperature was around 16C (60F). I have also heard that the Canon 6D is a much less noisy sensor than many others, which is also why I am rather shocked to get noticeable hot pixels at such short exposures.



#13 t_image

t_image

    Gemini

  • -----
  • Posts: 3154
  • Joined: 22 Jul 2015

Posted 24 August 2019 - 09:11 AM

So there is a lot of 'informational noise' on the internet.
I don't own a Canon so can't test myself, and maybe models differ also.

I've seen talk about "remapping" and also sometime that the "remap" is conveyed through the RAW but not jpg.

Looks like mention of CR2 mean you are viewing RAW after the touted "fix" procedure.

I've also read some "fix" procedures on cameras may be a matter of selective photosite discharge or re-calibration...

Additionally I've also seen some procedures can only be done at a repair shop via connection to some proprietary setup.

Hard to tell the accuracy of all this, but could explain why doesn't resolve your issue,

besides the differences between stuck,dead,leaky pixels,and one's that are holding excessive charges without discharging, etc.....

 

I wouldn't worry about the sensor being an issue unless problematic pixels were recognizable in terrestrial images or in 1:1 ROI live video renderings (maybe possible using BYEOS).....

There is a reason why this modding firm would mutilate the 6D for cooling purposes to get better astrophotos:

http://www.centralds...c-cooled-eos-6d



#14 Coconuts

Coconuts

    Vostok 1

  • *****
  • Posts: 194
  • Joined: 23 Sep 2012

Posted 24 August 2019 - 09:25 AM

I am just getting started in digital astrophotography, but in my first wide field image of Orion, Astro Pixel Processor automated making a Bad Pixel Map from my darks and flats.  I was startled to see just how many hot and cold pixels my Canon 6D had.  But the map took care of them.

 

All the best,

 

Kevin



#15 benringel22

benringel22

    Lift Off

  • -----
  • topic starter
  • Posts: 12
  • Joined: 07 Jun 2019
  • Loc: Peoria, IL

Posted 24 August 2019 - 09:50 AM

I am just getting started in digital astrophotography, but in my first wide field image of Orion, Astro Pixel Processor automated making a Bad Pixel Map from my darks and flats.  I was startled to see just how many hot and cold pixels my Canon 6D had.  But the map took care of them.

 

All the best,

 

Kevin

When you say remap are you referring to performing a manual sensor clean? Unfortunately this has no effect for me...



#16 Coconuts

Coconuts

    Vostok 1

  • *****
  • Posts: 194
  • Joined: 23 Sep 2012

Posted 24 August 2019 - 10:33 AM

No, a bad pixel map is created by reviewing your flats to find all dark pixels, and your darks to find all hot pixels.  A map of these (which can amount to 3% or more of all of your pixels) then is created.  These defective pixels are then not used in your final image.  It fixes them by interpolation of adjacent pixels. It's all done automatically; I found it very easy to do.  And once you create a Bad Pixel Map, it can be used for years; these are essentially permanent, pixel-specific aspects of your camera's particular sensor.

 

To learn more, check out this summary: https://www.astropix...a-bad-pixel-map

 

and this video: https://www.youtube....h?v=u8nVAt7X4Lw


Edited by Coconuts, 24 August 2019 - 10:57 AM.


#17 garyhawkins

garyhawkins

    Sputnik

  • -----
  • Posts: 48
  • Joined: 16 Feb 2019
  • Loc: San Diego

Posted 24 August 2019 - 06:41 PM

Hello everyone,

 

I recently purchased a lightly used Canon 6D on eBay as my first real astrophotography DSLR (I am very much a beginner...). This camera, when I received it today, had less than 8,000 shutter actuations, and had no signs of wear or damage. The first thing I did when getting the camera was to test it for defects and issues, and it seemed to pass my initial (newbie) inspection.

 

It was not until tonight while attempting to image a nebula that I noticed the images I was capturing had tons of white, blue, and red speckles covering the image (which clearly was not normal noise). This obviously scared me, and I did much research to find out that the camera sensor has many stuck/hot pixels. I attempted many fixes, such as the popular manual clean sensor method; however, attempt after attempt left me with the same result. It seems that it is often the same pixels causing the issue, and this makes me think that the sensor may have more than just stuck/hot pixels.

 

I know that a few of these pixels may be unavoidable and can be fixed in post processing (with dark frames, etc.), but this many affected pixels seems to me to be a very serious issue, and I am out of ideas to solve the problem without spending more money (on what was supposed to be an affordable camera...). Does anyone know what may be wrong with my camera sensor and if there is a way I can fix it?

 

Thank you all so much in advance for your help!

Ben

 

P.S. Below is an attached image of a 30 second dark frame at ISO 1600. I'm not sure if the pixels will show up for others, but the file on my computer shows an obvious and large number of these pixels, even without zooming in....

 

I've never noticed hot pixels on my D60 until I changed from Deepstacker (non-real time processing, it removes hot pixels) to Sharpcap (near real-time processing) for EAA.  I could see these little trials of red moving very slowly across the stacked image at the error rate of the tracking.  When I looked closer there were blues and green ones as well.  It does not really bother me unless the zoom is up really high, but I was surprised how many hot pixels there were - hundreds probably.



#18 WadeH237

WadeH237

    Aurora

  • *****
  • Posts: 4685
  • Joined: 24 Feb 2007
  • Loc: Snohomish, WA

Posted 24 August 2019 - 08:55 PM

I don't normally follow this forum, and ended up here by mistake - but I have a comment on this subject.

 

Speckled noise on DSLR images is completely normal, including what's described above.  On a cooled and regulated camera, you can calibrate a lot of it out, but with DSLRs, it's hard to do that because the temperature is not constant.

 

The best way to deal with it is to do a large dither between exposures.  That way, when you integrate the images later, the hot pixels won't line up and the software can recognize and eliminate them.  Done properly, this completely addresses it.

 

Oh, and on the observation that the Nikon does not exhibit the same speckled noise that the Canon does, I think that this is expected, too.  I've owned many Canon cameras, and they all do this.  I've never owned a Nikon, but I've heard that the Nikon DSLRs made in the last 5 or 10 years have a much smoother noise pattern in underexposed images.  I don't know why this is the case (ie. whether it's a sensor difference, and internal processing difference, or a bit of both).




CNers have asked about a donation box for Cloudy Nights over the years, so here you go. Donation is not required by any means, so please enjoy your stay.


Recent Topics





Also tagged with one or more of these keywords: astrophotography, beginner, CMOS, dslr, equipment, imaging



Cloudy Nights LLC
Cloudy Nights Sponsor: Astronomics