Jump to content


CNers have asked about a donation box for Cloudy Nights over the years, so here you go. Donation is not required by any means, so please enjoy your stay.


Supplement good seeing data with bad seeing data?

  • Please log in to reply
3 replies to this topic

#1 bokemon


    Viking 1

  • *****
  • topic starter
  • Posts: 748
  • Joined: 28 Jul 2020
  • Loc: Silicon Valley, California

Posted 21 June 2021 - 07:50 AM

Hello folks, 

A few days ago, I got a night of "good seeing", but was only able to get in about an hour of imaging time.  The stars are very sharp and clean, but the galaxy is kind of faint, and thus kind of noisy.  A few days prior, I got in a few hours on the same target, but the seeing was rather bad.  I'm fine with just keeping the stars background image from the "good" dataset, but I wonder if there is any way to take the galaxy data from the "bad" dataset and and paste it in, while trying to remove the stars from it, and maybe some kind of sharpening?

Or is it just a matter of throwing away the bad seeing dataset and just trying to process the good data with better noise reduction, contrast enhancement, etc?


I currently use AstroPixelProcessor, followed by Nikon CaptureNXD for various Lightroom-like image tweaks and "cheating".

At some point, I will probably buy Star Tools and / or Pixinsight.


These are 75% crops of the Fireworks galaxy taken with a 10" F4 Newtonian.  Good day vs bad day.


fireworks_galaxy_good bad seeing_resize.jpg

Edited by bokemon, 21 June 2021 - 07:52 AM.

  • limeyx likes this

#2 bobzeq25



  • *****
  • Posts: 25,567
  • Joined: 27 Oct 2014

Posted 21 June 2021 - 08:25 AM

The gain in signal to noise ratio with total imaging time is more than offset by the loss in quality.


There is no magic here.  The data is bad.  Even the galaxy data.  It's been blurred by seeing.


When my FWHM goes above 4 pixels and my guiding above 1.0 because of seeing, I just stop imaging.  No point to it.

Edited by bobzeq25, 21 June 2021 - 08:27 AM.

  • AhBok and limeyx like this

#3 AhBok



  • *****
  • Posts: 3,320
  • Joined: 02 Dec 2010
  • Loc: Lakeland, TN

Posted 21 June 2021 - 08:35 AM

This is a lesson I’ve learned begrudgingly! Good data + Bad data = Bad data. Lately, I’ve been culling out more and more subs that I would have kept until recently.
  • bobzeq25 and limeyx like this

#4 Midnight Dan

Midnight Dan

    Voyager 1

  • *****
  • Posts: 14,652
  • Joined: 23 Jan 2008
  • Loc: Hilton, NY, Yellow Zone (Bortle 4.5)

Posted 21 June 2021 - 09:18 AM

The problem with "good" vs. "bad" is there are 2 criteria: noise and sharpness.  The bright parts of the image are generally where you want sharp data.  Since the signal is strong there, noise is much less apparent and SNR is higher.  The dim parts are where you want less noise, and sharpness is less important.


There's a technique I've heard mentioned here on CN, although I have to admit I haven't tried it yet.  Stack you subs twice.  The first time, cull heavily keeping only the sharpest images.  The second time, stack all the subs, leaving out only those that have a lot of noise due to being too close to sunrise/set, moonglow, etc.  Process the first stack to maximize sharpness, process the second to minimize noise.


Then comes the tricky part - blending the two.  I don't know APP, but in PI you could use pixel math and a mask to do it.  If you have Photoshop you can use a mask and layers.  Again, I haven't actually done it, but I suspect the blending part would take some experimentation.  But the idea is to use the brighter parts of the sharp image, blended with the darker parts of the low-noise image.



CNers have asked about a donation box for Cloudy Nights over the years, so here you go. Donation is not required by any means, so please enjoy your stay.

Recent Topics

Cloudy Nights LLC
Cloudy Nights Sponsor: Astronomics