›

Wondering why CC isn’t in WBPP

Marsha Ann Wilcox
in PixInsight
I’ve found myself wondering why Cosmetic Correction isn’t a module in WBPP.  I often start to set up WBPP and need to exit, set up a CC instance, then return to the WBPP setup.  I know I could set up an in instance at startup and solve it that way.  

Thinking there’s a good reason it’s not part of the script, just wondering why.

Thanks!
Marsha

Comments

  • BlockHeadBlockHead
    There are a couple of reasons- but one is that is has a handful of methods that are not all compatible with WBPP. You implicitly desire the automatic method of hot pixel removal.

    I do agree, it would be a convenience... but this is one I didn't feel is worth the battle... bigger battles to win. 

    -the Blockhead
  • Marsha Ann Wilcox
    Just updated PI, and to my pleasant surprise, CC is now in WBPP!

    Grateful for the convenience.

    M.
  • BlockHeadBlockHead
    Can you show me a screenshot of what you are seeing?
    Thanks,
    -the Blockhead
  • Marsha Ann Wilcox
    File attached.

    Options to change the hot pixel sigma or point to a template.
    WBPP w CC 8-5-24 at 1.57 PM.jpg
    2696 x 1844 - 576K
  • BlockHeadBlockHead
    OK! Indeed, this appears to have happened in June while I was on vacation.
    Nice... now Auto is available.
    Thanks!
    -the Blockhead

  • Craig Weinert
    Cosmetic correction is now in the newest WBPP but it's default for hot pixels is 10 sigma.  I've watched you do your own CC and you usually have a much lower sigma--2-3.  Assuming my statistics are correct, a sigma of 10 is very strict- a neighboring pixel will have to have a very, very different value to identify a hot pixel, whereas a sigma of 2 is very permissive--that will identify a lot more as hot pixels.  

    Why the big difference between your method and the default PI sigma?  Is there an advantage to strict versus permissive? 
    craig
  • BlockHeadBlockHead
    Craig,

    I am not certain, I just tried today to get the latest version- but the server does not appear to be working (for me). However, unless the Universe has changed and I am in bizarro world... 
    Smaller sigma values are MORE aggressive. A sigma value of 10 would do virtually NO hot pixel removal. 

    So unless they changed something- I would continue to set it to the previous default of 3 sigma. 

    In this case the sigma value signifies the pixel needs to be more than the chosen sigma threshold to be removed. 10 sigma is a pretty big value!

    If you run WBPP...it will tell you how many pixels are being cosmetically corrected. If the value is a few hundred or less out of millions of pixels..that is a very small amount.

    -the Blockhead
  • BlockHeadBlockHead
    Oh...I just looked into it...
    The WBPP method is *completely different* .
    I will need to make a little video to explain it!

    -the Blockhead
  • BlockHeadBlockHead
    edited August 2024
    These sigma units are with respect to the median (filtered) of the image (the master dark).
    So the farther you are away (bigger sigma) the FEWER pixels will be tagged as being hot. And the smaller the sigma (close to the noise floor) the MORE pixels will be considered hot.

    So it does make sense. The "what" the sigma refers to has changed.  I will make a video.

    -the Blockhead
  • Marsha Ann Wilcox
    Thank you for the video, Adam. It helped me understand the new implementation of cosmetic correction in WBPP. The default value of 10 seems a little high to me. I guess I’ll have to do some experimentation to find the right value for me.
    Thank you again.

  • KEVIN WADDELL
    Another interesting note regarding the new Auto CC.  If you use this method,  the suffix  _cc  is no longer present in the filename
  • BlockHeadBlockHead
    Yes, this is because the dark frame method of CC is part of the image calibration process. Literally ImageCalibration... which is why it is just _c .

    -the Blockhead
Sign In or Register to comment.