Author Topic: Star Artifacts Post integration- Average Combination vs. Median SNR  (Read 3735 times)

Offline fulatoro

  • Newcomer
  • Posts: 36
I am seeing artifacts on all bright stars after integration using default parameters.  Interestingly enough, when switching the combination from "Average" to "Median" these disappear.  This is great except that PixInsight documentation at http://pixinsight.com/doc/tools/ImageIntegration/ImageIntegration.html#usage_003

States that:

Median. The output integrated image is the pixel-by-pixel median of all input images. Provides robust implicit rejection of outliers, but at the cost of about a 20% signal loss.

20% signal loss ....Can I fix this without resorting to Median, what is causing this weird behavior. I am attaching the result of the integration.


Offline Carlos Milovic

  • PTeam Member
  • PixInsight Jedi Master
  • ******
  • Posts: 2172
  • Join the dark side... we have cookies
    • http://www.astrophoto.cl
These are very weird artifacts. Does any of your individual frames shows what may be the source for this?
You may try bu changing the rejection thresholds, using smaller values, specially on the shadows.
Regards,

Carlos Milovic F.
--------------------------------
PixInsight Project Developer
http://www.pixinsight.com

Offline sreilly

  • PixInsight Padawan
  • ****
  • Posts: 791
    • Imaging at Dogwood Ridge Observatory
A single image may be a better clue. It looks to me like the collimation is off but that may just look like that from the resulting combined image. Looking at the dimmer stars shows some wierd shapes as well.

Steve
www.astral-imaging.com
AP1200
OGS 12.5" RC
Tak FSQ-106ED
ST10XME/CFW8/AO8
STL-11000M/FW8/AO-L
Pyxis 3" Rotator
Baader LRGBHa Filters
PixInsight/MaxIm/ACP/Registar/Mira AP/PS CS5

Offline fulatoro

  • Newcomer
  • Posts: 36
Three configurations

1-Default Configuration - Average Combination
2-Median Combination, less artifacts but more overall noise
3-Default Configuration - Average Combination with SIGMA high set to 1.0, less artifacts, less noise but still artifacts...

Could this be as a result of poor guiding in one or more of the frames where star alignments essentially ends up with 1 or more frames with bigger or more noisy stars.  Interestingly enough, artifacts only appear on ONE side of the star which might indicate guiding issues for some of the frames...Or I am completely clueless....


Moussa

Offline fulatoro

  • Newcomer
  • Posts: 36
Median Combination

Offline fulatoro

  • Newcomer
  • Posts: 36
Average Combination with Sigma high = 1.0

Offline fulatoro

  • Newcomer
  • Posts: 36
Looks like Guiding was the issue, I identified 6 frames that were worst than the others. I am assuming this is what led to the averaging being worst than the median in terms of star artifacts.

I am attaching a bad guiding frame and the resulting integration with all bad frames removed.

What do people usually do for "grading" frames? I can't be looking at each frame before integrating...:-(

Moussa

Offline fulatoro

  • Newcomer
  • Posts: 36
Integration after removing bad guiding frames.

Offline fulatoro

  • Newcomer
  • Posts: 36
I just discovered the SubframeSelector script...That answers my question about choosing frames...Awesome!!!!