›

Calibration files of M51 work-flow confusion

Paul Wilson
in PixInsight
Forgive questions that illuminate my ignorance ok?

I am unaccustomed to downloading others' data, so I am confused about where to put what when loading the M51 workflow flats, etc.
I understand that the server couldn't accommodate all the files in one session, so the calibration files are spit into 2 parts.
But there are LRGB flats, for example, of each color on both downloads.  I don't do "raw" flats vs regular flats, so what is the difference and where do they go? Is "raw" just for DSLRs?  Do the color filter flats for LRGB from both downloads get regrouped together in the loading?
Perhaps this is covered in the tutorial, but I had planned to stack in my familiar Astro-Pixel-Processor, and then pick up the tutorial post stacking.
I'm not getting why all the "L" files aren't together, or the other colors grouped together (for different cameras?).  Also, there are different headers within each file group.  What is the difference in those?

If the tutorial clears all this up, I am glad to try to go that route instead.  I would, for myself, have RGB with each color having its own lights/flats/dark flats and a group of library darks that apply to all.  Perhaps this exercise will show me how and when I should consider adding Ha to RGB! 

Comments

  • BlockHeadBlockHead
    The tutorial does clear all of this up. You really do need to follow along. 
    Thanks!
    -the Blockhead
  • Paul Wilson
    I tried my best.  I carefully followed along, trying to enter those badly jumbled files and reading the dates.  
    But I started running into problems that were not happening to you.  I blinked the frames as you did.  But when trying to load the Bias files, only the first 20 would go in, and PI said the remainder "did not appear to be FITS files"!  That I don't understand.  So I just went with 20 instead of 60. (I myself shoot "dark flats")

    But after the calibration, when I blinked them again, bad things had happened.  About 10 of the L frames had become so bright, you couldn't tell there was a galaxy in there.  I should have started over from scratch with those awful files again.

    In the star alignment process, because of his data, you got "1 failed".. I got "4 failed".  Perhaps I should have stopped right there..  But I tried to keep going.

    In the Integration of the red files, I followed along:  windsor, then sigma clipping.  Each time, I was getting magenta WARNING (Will Robinson), and then "Meta-Data Not Generated". So no integration.  That did it.

    So...I am hanging it up for that particular workflow.  I can only guess that in trying to isolate the correct flats,darks, etc, I got something in with something incorrectly.  Hopefully the NGC 300 has files I can load more easily.  You know, garbage in/garbage out.  
    I'm sure I learned a few things with 3 days of frustration, but I had hopes of producing a real image.
  • Paul Wilson
    I would suggest 2 things.

    1)  Since you are dealing with newbies that can barely follow along step by step, at least let us build our foundation on some less problematic data while we take our first steps.  OR,

    2)  Help us out with this guy and re-organize his 2 sets of data with matching calibration files so that someone
    less practiced can at LEAST get the data loaded correctly in order to try to follow the lesson.

  • Paul Wilson
    This morning, just to see what NGC 300 is about, I downloaded the files.
    I blinked them in PI, the lights, and saw no BIG issues, perhaps you will discover some in your tutorial.

    I loaded them up into APP, my comfort zone having used it for a year.  In 20 minutes, including blinking, I had a very nice looking stack.  Some vignetting in the corners, guess his flats weren't perfect.  He has the same problems as me with star halos, will be VERY instructive to see what you do.  His sky looks darker, BETTER than mine, where does he live?  

    I will look forward to starting your tutorial this afternoon!  I love/hate this hobby!  But it's who I am.  Now back to following the latest details of the StarShip engine ignition test at Boca Chica, the new Edwards.
  • Roger Smith
    edited August 2021
    Paul,
    January was awhile ago. Did you finish M51?  I am also working on this downloaded M51 data. I found all the same files that Adam opened in the video looked the same as the video. Even the same (bad) files that would not open had errors and would not open for me. 

    One key learning point for me, when doing my own imaging, is to strive to keep the data set simple, well organized, and grouped according to the calibration files. 

    For the site location, you can open any of the FITS headers and see the longitude and latitude, then plug it into Google Maps. In the FITS header you can also see he (Mr. Josep Drudis) is using a 600mm Planewave at 3962mm focal length. Add to that 20 min subs, and 9nm pixels and we must realize there is a huge amount photons gathered in each sub. At least compared to my travel scope103mm refractor with 4.78nm color camera.

    Roger
  • BlockHeadBlockHead
    I always get criticism on this front. If I use "nice" data- I am an elitist using the best stuff out there. 
    If I use other people's data ("Let's see what Adam can do with my light polluted images...etc etc)- then the data isn't perfect in organization or quality- and I still get criticism. 
    Can't win this one. lol
    -theBLockhead
  • Roger Smith
    Adam,
    For me, I did not criticize the data quality. Taken for what the purpose of the Fundamental Workflow videos it demonstrates the usage and workflow very well. And the final result is great in my opinion. So I actually think the data quality is quite good as I said above. But of course I understand you are appealing to a wide range of subscribers, and each one has a history and style. (Like me!)

    What is a bit messy is the matching of flats and lights data. With the data unclear new users can easily get confused/lost. This results in extra distraction on top of the core learning of this M51 Fundamentals videos.

    Here is my take on the data....
    Actually for L, the Apr 2020 data does not show any pickoff visible even with extra stretching. All the Mar2020 L data has an obvious pick off.  All the L Flats, from both data groups of flats, have a pick-off showing. The pick off was apparently moved to be out of sight in the several weeks between when the flats were taken (03Mar20) and the L data was taken (mid Apr).
    So none of the L Flats really matches the Apr2020 L data.  Luckily both sets of L flats are very close with only a couple of donuts different (one at bottom, one in the galaxy).  The result after calibration is the pickoff showing in the calibrated data, but it is at the edge so it is irrelevant. 

    All the PA240 flats are at PA240 in the FITS header.
    None of the L data is at PA240. It is at PA137 for the Mar2020 data, and at PA227 for the Apr2020 data. (From the FITS headers.)

    So it is fair to say the data provided to you is not easy (possible?) to 'correctly' organize and match up. But the result can still be a nice image. 

    So thank you for your hard work to teach us the Fundamentals and more.

    Roger
Sign In or Register to comment.