›

Cosmetic correction

Sedat Bilgebay
in PixInsight
Hi Adam, 

Thanks for this excellent series of tutorials.

I was watching "Home » PixInsight » Fundamentals » The Workflow: The Example of Beautiful NGC 3486 » Workflow Batch PreProcessing" last night and I noticed that you used the same Cosmetic Correction icon for both 1x1 and 2x2 data (@ 07:40) . My guess is that the script skipped the cc operation for the 2x2 data.... please kindly comment. 

Best wishes

Sedat Bilgebay


Comments

  • BlockHeadBlockHead
    Hi Sedat,

    The "auto" method of Cosmetic Correction is independent of image size (I think, I am writing this not looking at my computer). This is why I can get away with this. All I need to do is specify the sigma rejection parameters. If I were to use the Master Dark method of identifying hot pixels... then I need to match darks of 1x1 with 1x1 data and 2x2 darks with 2x2 data. This isn't the case for the "Auto" method. 

    So I think everything is OK...but I will look more closely at it this evening to make certain.

    By the way, I hope others will chime in if they know the answer to things! (I will not always know...but together we can figure it out )

    -the Blockhead
  • Sedat Bilgebay
    This is a valid point Adam. I missed that "auto" detail. It should work fine in this case, I believe.  I always use the master dark method hence my confusion.

    Thanks.

    Sedat
  • Scott Iver
    I would be very interested to know sort of the practical difference between "auto" vs master dark.  I was not aware of the "auto" method being able to work like that this seems like a really good time-saver.  Are there any trade-off's or gotcha's in using auto instead of the darks? 
  • BlockHeadBlockHead
    In my case "auto" is my best choice because I tend to use the same master dark frame from many weeks if not months. The hot pixel population changes (in the raw/new data) over time scales like this. So "auto" identifies hot pixels by a statistical likelihood of being an "outlier" - whereas the master dark is an accurate fixed pattern- but temporally it might not characterize things as well in the future. "Auto" may not be as precise in identifying hot pixels with a conservative rejection criteria- but I consider Cosmetic Correction only an incremental step in hot pixel removal. Dithered data with subsequent rejection will take care of anything CC misses.

    That is my logic anyway. 

    -the Blockhead
  • Scott Iver
    Very nice,  thank you!
Sign In or Register to comment.