Back to forums

Drizzle and Sigma Kappa stacking, can it be combined?

Russell Croman Astrophotography BlurXTerminator Pleiades Astrophoto PixInsight Acquisition Drizzle Deep Sky
10 replies•539 views
Stephan Linhart avatar
Stephan Linhart avatar
Dear AB community,

Lately I started to drizzle my images as it is recommended by BlurXTerminator even wehen oversampling. I found that combined with a sigma kappa, or similar, integration in Pixinsight satellite trails will remain visible.

Searching the net I found a statement that drizzling and sigma kappa stacking methods are not compatible.

I like to continue using drizzle also for undersampled images, but having zillions of muskoids on my images would kill my plans ;)

See here for a result with satellite trails remnants.
They are weak, but it bothers me:
https://www.astrobin.com/1ixbcx/C/

Best regards,

Stephan
Engaging
andrea tasselli avatar
andrea tasselli avatar
Stephan Linhart:
Dear AB community,

Lately I started to drizzle my images as it is recommended by BlurXTerminator even wehen oversampling. I found that combined with a sigma kappa, or similar, integration in Pixinsight satellite trails will remain visible.

Searching the net I found a statement that drizzling and sigma kappa stacking methods are not compatible.

I like to continue using drizzle also for undersampled images, but having zillions of muskoids on my images would kill my plans ;)

See here for a result with satellite trails remnants.
They are weak, but it bothers me:
https://www.astrobin.com/1ixbcx/C/

Best regards,

Stephan

Not sure what sigma-kappa "integration" is supposed to mean. Are you talking about some pixel-rejection filter?  At any rate, have you tried Large Scale Pixel Rejection (High)?
Stephan Linhart avatar
Stephan Linhart avatar
Hi Andrea,

yes, I was referring pixel rejection.
I did not try large scale pixel rejection, and will give it a shot.
Stephan Linhart avatar
Stephan Linhart avatar
Tried it, but no change.
andrea tasselli avatar
andrea tasselli avatar
Can you share an image of your stacking arrangements (i.e., the ImageIntegration window) with all the options visible? Also, have you LN'd your light frames and is it to a frame *without* any trail? Because in my experience I don't get to see any trail with 46 frames to stack up.
Helpful Concise
Stephan Linhart avatar
Stephan Linhart avatar
Hi Andrea,

I always use local normalization, I used a reference frame without trailing.
It is also not a general problem I see. Here is the image integration options as I used it on the WBPP script. As said before tried again with the large scale pixel rejection high checked.

andrea tasselli avatar
andrea tasselli avatar
Ditch WBPP. Apply the ImageIntegration process to the registered images (adding the  normalization files) as shown below:

Stephan Linhart avatar
Stephan Linhart avatar
Will try it out and let you know!
Brian Puhl avatar
Brian Puhl avatar
In WBPP I've always left the rejection to auto.   46 frames might not be enough for WBPP to reject it, and also, IMO that's not really enough frames for an OSC image.   Chances are there's not enough signal in the blue yet, especially at your focal ratio, to overcome the satellites.
Helpful Insightful Concise
Elmiko avatar
Elmiko avatar
Have you tried Windsorized clipping?
Guiem Kimi avatar
Guiem Kimi avatar
I usually dither and drizzle to 2x. Both Linear fit and Windsorized sigma clipping, with default settings, do the job with no issues in my case. No hot pixels and no satellite trails.

I choose between the two algorithms depending on the amount of subs I have. Last images, with about 240 subs, Linear fit was perfect. With 50 subs I use Windsorized sigma clipping algorithm.

Guiem.
Helpful Concise