›

Using a custom background for DBE

Wade Hilmo
in PixInsight
Hello,

I have some old data of NGC6888, which includes images for red, green, blue, OIII and Ha channels.  The data is properly bias and dark calibrated, but the flat calibration is not perfect.  It appears that dust motes calibrated out mostly (enough so that they are fainter than any of the data in which I'm interested), but the integrated master frames still show a significant radial gradient.  The data was collected in under very dark skies and there is virtually no gradient other than what appears to be over-corrected vignetting.

Using DBE, I am able to get a good background model from the blue channel.  For other channels though, particularly Ha, there is no part of the frame that is free of interesting signal.

Since the nature of the illumination error seems to be mostly the same in all channels, I would like to be able to use the background extracted from the blue channel to correct the illumination in all of the other channels.

Is there a way to do this with DBE, or will I need to build a PixelMath expression and do it manually?

Any suggestions?

Thanks,
-Wade

Comments

  • BlockHeadBlockHead
    Hi Wade,

    If I understand correctly.. you don't just want to use the same samples in terms of position... (this would just require you create a new instance of DBE and apply to another image)- but instead you want the actual calculated coefficients you determine in one image to be applied to another.

    I have not done this...but I believe if you convert each sample to Fixed (check the box)- that coefficient will be locked in. So when you create your new instance... it should have fixed values for you to apply.

    I have NO idea if this would actually work! But a good experiment if I am on the right track.

    -the Blockhead
  • Wade Hilmo
    Thanks for the suggestion.  It appears that worked.

    In the better images that I see on the web, the Ha to the south of the Crescent are a bit brighter than the ones to the north.  My Ha master using fixed samples that I selected on the blue channel has that same characteristic.

    When I use the same sample points, but not fixed (so that they take the values from the Ha master), the area immediately to the north of the Crescent is too bright in Ha, so it's not properly corrected that way.  I could probably select better samples by finding background spots in the Ha master, but there just aren't any good places.  This is a pretty narrow field shot, so there's at least some Ha everywhere.
Sign In or Register to comment.