›

Comet Alignment not removing comet from comet with stars pics

Rich Temen
in PixInsight
I have been following Adam's steps on comet C/2022 E3.
I am now entering the Comet Subtraction session.
When I set up the files and put in the operand, the comet is not subtracted.

I need some assistance on how to debug this problem.
So far, all of the previous steps seem to have been working, so I do not know how to debug this.

Thanks,
Rich

Comments

  • BlockHeadBlockHead
    Hi Rich,

    You may want to look at the other recent thread on this in this forum.
    A member had the some question... 

    1. You need to make certain you used exactly the same reference that your star aligned images used for your comet alignment reference. 
    2. If the subtraction is aligned by not good enough, you should try turning off Linear Fit.
    3. You need to make certain the "Comet Aligned" button is checked for the operand.

    Some screenshots might also be helpful.

    -the BLockhead
  • Rich Temen
    I am attaching my Comet.xisf file which is the result of the previous session.
    It looks beautifully cleaned up.
    Also is a jpg of my PI screen. I have also tried turning Linear fit off, and it makes no difference.
    The comet stays this clear and obvious in all the comet removed files.
    I have no idea what to check for to find what is wrong.

    Also, you mentioned another post for the same problem. I haven't found it yet, so if you can point me to it, that would help.

    Thanks.
    Rich
    oops, the system says my comet.xisf is not allowed to upload.
    Suggestions?
    CometNotRemoved.jpg
    2084 x 871 - 397K
  • BlockHeadBlockHead
    I am concerned there is an error ... Do you see how PI is complaining about the PSF Fitting?
    Are you 100% certain you clicked on the nucleus of the comet in the three frames? 
    If so... your next step would be to turn off the PSF Fitting...

    I show in one of my comet videos how to use the Comet Mask output to see the positions of what CometAligment is calculating and where you comet actually is. You did not show enough of your screen... but I think you can also see the offsets/shifts in the table and they will be non-sensical if there is the problem which appears to be the case.You can also simply Pin EVERY position in your list (you only have 20). It CANNOT screw this up since you are forcing it. 

    -the Blockhead


  • Rich Temen
    I pinned every one of the twenty. It makes no difference.
    Maybe we should make sure that the star and comet pics are the right ones.
    Or that they are made properly.

    How can we check that?

    Rich

  • BlockHeadBlockHead
    Well.. you can simply open your reference image (that StarAligned one) that has the comet.
    Then open your starless integrated image comet (the operand image). If you compare/blink them... they have to be aligned. If not... you made an error. In the previous thread the member promised me he did not make a reference error... but then discovered he did. :)

    Check your reference images... and report back!
    -the Blockhead
  • Rich Temen
    They are spot on!
    I guess what I asked before was to be sure that the files that I am trying to subtract the comet from are set up properly.
    Maybe I am choosing the wrong files or didn't generate the files properly.
    After following the tutorial this far, there are now several folders with the images with stars and comet.
    Maybe one of them are wrong???

    Also, my pics are from a OSC while Adam's tutorial is with separate red, blue and green files.
    I was watching his process and made assumptions that if I followed his steps for only one of his colors, it would work.
    Maybe this is the problem.
    What do you think?

  • BlockHeadBlockHead
    It does require some good brain tracking to make sure you are not assuming you are using certain files rather than others. So you opened the file *from the list of files you have in CometAlignment* and you compared this to your operand image?

    You turned off PSF fits?
    You outputted the comet mask and compared the outputted aligned images?

    Also, I do have a an OSC example- but there are some additional complications to doing OSC correctly (fully). You should eventually watch that lesson. 

    Basically the way to approach this is to post screenshots (or video) that "proves" everything is correct. You need to demonstrate the files all match (the files you have loaded!). You need to show the comet path in your images matches the comet mask output. All of this kind of stuff...  So I think some screenshots would be helpful.

    I do look forward to figuring this out. 

    -the Blockhead
  • Rich Temen
    Is this error message important?

    G:/Astronomy/C2022/2023_02_08/comet_only/aligned/Light_C-2022 E3_120.0s_Bin1_20230208-195223_0009_c_cc_starless_ca.xisf

    19 : 2297832 19.649% ( 2231048 + 66784 = 19.078% + 0.571%)

    G:/Astronomy/C2022/2023_02_08/comet_only/aligned/Light_C-2022 E3_120.0s_Bin1_20230208-195425_0010_c_cc_starless_ca.xisf

    20 : 3541606 30.285% ( 3362371 + 179235 = 28.752% + 1.533%)


    Total : 60087888 25.691% ( 56115963 + 3971925 = 23.993% + 1.698%)


    Computing scale and location parameters: done

    PSF signal evaluation (ch 0):done

    Noise evaluation (MRS): done

    Computing noise scaling factors: done


    Scale estimates : 1.159714e-02

    Location estimates : 5.512360e-02

    Noise estimates : 1.3378e-03

    SNR estimates : 1.2393e+01

    PSF signal weights : -nan(ind)

    PSF SNR estimates : 0.0000e+00

    PSF fit counts : 0


    ** Warning: Inconsistent Instrument:Telescope:FocalLength (FOCALLEN keyword) value(s) - metadata not generated.

    20.435 s



    * CometAlignment: Loading 20 file(s)

    Loading image: w=4144 h=2822 n=3 RGB Float32

    12 image properties

    179 FITS keyword(s) extracted

    Loading image: w=4144 h=2822 n=3 RGB Float32

    12 image properties

    179 FITS keyword(s) extracted

    Loading image: w=4144 h=2822 n=3 RGB Float32

    12 image properties

  • BlockHeadBlockHead
    No.. that warning isn't the issue.
    -the Blockhead
  • Rich Temen
    I get these errors when trying to subtract the comet.

    179 FITS keyword(s) extracted

    ** Warning: PSF fitting error: No significant data. Using interpolated centroid coordinates.

    cx = 2263.34, cy = 1495.74

    dx = +137.57, dy = +35.96

    Writing output file: G:/Astronomy/C2022/2023_02_08/registered/Light_BIN-1_4144x2822_EXPOSURE-120.00s_FILTER-NoFilter_RGB/comet_subtracted/Light_C-2022 E3_120.0s_Bin1_20230208-195023_0008_c_cc_d_r_cs.xisf

    Writing image: w=4144 h=2822 n=3 RGB Float32

    184 FITS keyword(s) embedded

    12 image properties embedded


    Loading target file: G:/Astronomy/C2022/2023_02_08/registered/Light_BIN-1_4144x2822_EXPOSURE-120.00s_FILTER-NoFilter_RGB/Light_C-2022 E3_120.0s_Bin1_20230208-195223_0009_c_cc_d_r.xisf

    Loading image: w=4144 h=2822 n=3 RGB Float32

    12 image properties

    179 FITS keyword(s) extracted

    ** Warning: PSF fitting error: No significant data. Using interpolated centroid coordinates.

    cx = 2273.89, cy = 1498.50

    dx = +148.12, dy = +38.71

    Writing output file: G:/Astronomy/C2022/2023_02_08/registered/Light_BIN-1_4144x2822_EXPOSURE-120.00s_FILTER-NoFilter_RGB/comet_subtracted/Light_C-2022 E3_120.0s_Bin1_20230208-195223_0009_c_cc_d_r_cs.xisf

    Writing image: w=4144 h=2822 n=3 RGB Float32

    184 FITS keyword(s) embedded

    12 image properties embedded


    Loading target file: G:/Astronomy/C2022/2023_02_08/registered/Light_BIN-1_4144x2822_EXPOSURE-120.00s_FILTER-NoFilter_RGB/Light_C-2022 E3_120.0s_Bin1_20230208-195425_0010_c_cc_d_r.xisf

    Loading image: w=4144 h=2822 n=3 RGB Float32

    12 image properties

    179 FITS keyword(s) extracted

    ** Warning: PSF fitting error: No significant data. Using interpolated centroid coordinates.

    cx = 2284.55, cy = 1501.29

  • BlockHeadBlockHead
    Yes, I sent you an e-mail...
    -the Blockhead
Sign In or Register to comment.