Pixel issue - Forums

Topic starter

Feb 6, 2016 · Edited: Feb 6, 2016

Need help trying to figure out how to remove these RGB hot/cold pixels. The image is posted along with my other images. This is after complete processing, including cosmetic correction, which did not fix the problem.
I'm seeing them on the raw image, and it's there on short exposures and long exposures. I tried the manual sensor cleaning for the Canon camera, but that did not correct the problem. By the way, it's a Hutech modified Canon 40D.
Thanks in advance for any input.
Mike

Scott

Feb 6, 2016

Dithering usually takes care of this problem.

Mike Pelzel

Topic starter

Feb 6, 2016

Yes, I use dithering, alos calibrate with darks, flats, and bias.

Scott

Feb 6, 2016

How old are your darks? Maybe you need to take some new ones as the noise pattern of a sensor will change over time.

Well Written Concise

Mike Pelzel

Topic starter

Feb 8, 2016

Shot a new set of darks, that seems to have helped.

Daniel.P

Feb 9, 2016

I am using a QHY8L color CCD camera. When cooled at -15°C, it shows negligible dark signal but it had some few hor pixels.
Initially, dithering didnt help me to fully remove the hot pixels ater stacking, especially the brighter ones.
I eventually found that I had to used cosmetic correction on the raw data (after bias and flat correction). I do this in Pixinsight using the CFA option, which process each CFA components separately.
Then I debayer, register and stack.
I am a beginner with forum, but hope this will be helpful for you.
I check your photos, there are very nice !
Daniel

Helpful Concise Supportive

Edoardo Luca Radice (Astroedo)

Feb 10, 2016 · Edited: Feb 10, 2016

I am using a QHY8L color CCD camera. When cooled at -15°C, it shows negligible dark signal but it had some few hor pixels.
Initially, dithering didnt help me to fully remove the hot pixels ater stacking, especially the brighter ones.
I eventually found that I had to used cosmetic correction on the raw data (after bias and flat correction). I do this in Pixinsight using the CFA option, which process each CFA components separately.
Then I debayer, register and stack.
I am a beginner with forum, but hope this will be helpful for you.
I check your photos, there are very nice !
Daniel

I completely agree, another trick in PixInsight (but also present in other preprocessing software) is using a stronger pixel rejection setting (for example a smaller sigma for the bright pixels when sigma clip type algorithms are used).

Seeing your image, IMHO, you are using a value of dithering too small, because your hot pixels are very near to each other.
Having hot pixels more distant makes easier to the pixel rejection algorithm to get rid of the anomalous pixel values.

bye

Edoardo

Helpful Insightful Respectful

Daniel.P

Feb 13, 2016

I completely agree, another trick in PixInsight (but also present in other preprocessing software) is using a stronger pixel rejection setting (for example a smaller sigma for the bright pixels when sigma clip type algorithms are used).

Edoardo,
I like the idea of increasing the pixel rejection for bright pixel, but how do you achieve this in PI ?
Daniel

Edoardo Luca Radice (Astroedo)

Feb 14, 2016

It is very simple:
In ImageIntegration choose a PixelRejection algorithm (In Pixel Rejection (1) section) based on how many subframes you are integrating (most of the times I use Winsorized Sigma Clip).

Set The sigma value for rejection in Pixel Rejection (2) section: remember that the lower is the value of sigma the higher is the rejection strenght.

Use a value low enougth to reject your bad pixel without loosing too much signal.

Read this good presentation by Jordi Gallego

http://astrosurf.com/jordigallego/articles/Image_integration_JGallego.ppt

If you are using Batch Preprocessing, you have similar settings as Integration Parameters...

Bye

Edoardo

Helpful

Daniel.P

Feb 20, 2016

Thank you very much

Daniel