Back to forums

Anybody Know What Happened Here?

Pleiades Astrophoto PixInsight Artifacts
14 replies•1.3k views
Jerry Gerber avatar
Jerry Gerber avatar
I shot 3 nights with an OSC cooled camera, Bortle 1 sky, no moon, good seeing and transparency.  Each night with its own flats. 

First night:   Askar Sii-Oiii filter
Second night:  Askar Ha-Oiii filter
Third night:  UV/IR Cut filter

I pre-processed them in WBPP and ended up with 3 unstretched images.  When I combined the Ha-Oiii and Sii-Oiii images, everything is good.  But when I added the 3rd image, the UV/IR Cut integrated unstretched image, I got this:




More strange, is here below is the unstretched integration of only the UV/IR image.  It looks fine. 



I tried combining the Ha-Oiii and the UV-IR Cut images and they look fine.  I also tried combining the Sii-Oiii and the UV-IR Cut images and they look fine too.  But when I combine all three these weird artifacts appear and I have ruled out dust motes because the individual integrated images do not have any calibration issues.  Why can't I combine all three images in Pixinsight's Image Integration Process?


Thanks,
Jerry
Well Written Engaging
Tim Eriksen avatar
Tim Eriksen avatar
Jerry Gerber:
I shot 3 nights with an OSC cooled camera, Bortle 1 sky, no moon, good seeing and transparency.  Each night with its own flats. 

First night:   Askar Sii-Oiii filter
Second night:  Askar Ha-Oiii filter
Third night:  UV/IR Cut filter

I pre-processed them in WBPP and ended up with 3 unstretched images.  When I combined the Ha-Oiii and Sii-Oiii images, everything is good.  But when I added the 3rd image, the UV/IR Cut integrated unstretched image, I got this:




More strange, is here below is the unstretched integration of only the UV/IR image.  It looks fine. 





Thanks,
J

Why are you using luminance with narrowband?

If you need a lum just make a synthetic from your data minus your lum filter
Dave Boddington avatar
Dave Boddington avatar
looks like the UV_IR image registration is incorrect hence stars being rejected..  possibly 180 degree out
Dan Brown avatar
Dan Brown avatar
Something appears to be rotated 180 degrees. The dark spots correspond with the bright stars but rotated 180. Did you blink everything? Did you subtract stars?
Dan
Adam Block avatar
Adam Block avatar
Dan Brown:
Something appears to be rotated 180 degrees. The dark spots correspond with the bright stars but rotated 180. Did you blink everything? Did you subtract stars?
Dan

Dan has it. Rotated images in the stack. Rejection... and yuck.
Scott A avatar
Scott A avatar
Tim Eriksen:
Jerry Gerber:
I shot 3 nights with an OSC cooled camera, Bortle 1 sky, no moon, good seeing and transparency.  Each night with its own flats. 

First night:   Askar Sii-Oiii filter
Second night:  Askar Ha-Oiii filter
Third night:  UV/IR Cut filter

I pre-processed them in WBPP and ended up with 3 unstretched images.  When I combined the Ha-Oiii and Sii-Oiii images, everything is good.  But when I added the 3rd image, the UV/IR Cut integrated unstretched image, I got this:




More strange, is here below is the unstretched integration of only the UV/IR image.  It looks fine. 





Thanks,
J

Why are you using luminance with narrowband?

If you need a lum just make a synthetic from your data minus your lum filter

Since he's using a OSC camera, I'm assuming he's trying to capture the RGB then adding the SHO data to build a SHORGB image.
Jerry Gerber avatar
Jerry Gerber avatar
Thanks all!  You're right, I didn't notice the rotation error.   I rotated and aligned (registered) the 3 images and that allowed the image integration to succeed without any artifacts. 

Live and learn!

Jerry
Well Written Respectful
James Peirce avatar
James Peirce avatar
Aside from the suggestion mentioned above, what did you mean by “when you add”? Did you have a mask active that could have created an issue?
Jerry Gerber avatar
Jerry Gerber avatar
James Peirce:
Aside from the suggestion mentioned above, what did you mean by “when you add”? Did you have a mask active that could have created an issue?

By "add" I mean I added the image to the Image Integration Process.  But because all three images were not registered and one was rotated 180 degrees, the integration produced those artifacts in the first image I posted above.  After rotating the image to match the others and after registering them, they integrated successfully without issue.
Well Written Helpful Concise
James Peirce avatar
James Peirce avatar
Jerry Gerber:
James Peirce:
Aside from the suggestion mentioned above, what did you mean by “when you add”? Did you have a mask active that could have created an issue?

By "add" I mean I added the image to the Image Integration Process.  But because all three images were not registered and one was rotated 180 degrees, the integration produced those artifacts in the first image I posted above.  After rotating the image to match the others and after registering them, they integrated successfully without issue.

Gotcha. Sounds like the commenters above intuited this.
Jerry Gerber avatar
Jerry Gerber avatar
James Peirce:
Jerry Gerber:
James Peirce:
Aside from the suggestion mentioned above, what did you mean by “when you add”? Did you have a mask active that could have created an issue?

By "add" I mean I added the image to the Image Integration Process.  But because all three images were not registered and one was rotated 180 degrees, the integration produced those artifacts in the first image I posted above.  After rotating the image to match the others and after registering them, they integrated successfully without issue.

Gotcha. Sounds like the commenters above intuited this.

Yes, there's many very helpful people around this forum.
apoorva.iyer avatar
apoorva.iyer avatar
Run StarAlignment. Profit.
Jerry Gerber avatar
Jerry Gerber avatar
Scott A:
Tim Eriksen:
Jerry Gerber:
I shot 3 nights with an OSC cooled camera, Bortle 1 sky, no moon, good seeing and transparency.  Each night with its own flats. 

First night:   Askar Sii-Oiii filter
Second night:  Askar Ha-Oiii filter
Third night:  UV/IR Cut filter

I pre-processed them in WBPP and ended up with 3 unstretched images.  When I combined the Ha-Oiii and Sii-Oiii images, everything is good.  But when I added the 3rd image, the UV/IR Cut integrated unstretched image, I got this:




More strange, is here below is the unstretched integration of only the UV/IR image.  It looks fine. 





Thanks,
J

Why are you using luminance with narrowband?

If you need a lum just make a synthetic from your data minus your lum filter

Since he's using a OSC camera, I'm assuming he's trying to capture the RGB then adding the SHO data to build a SHORGB image.

Yep , that's exactly right!
TiffsAndAstro avatar
TiffsAndAstro avatar
Jerry Gerber:
I shot 3 nights with an OSC cooled camera, Bortle 1 sky, no moon, good seeing and transparency.  Each night with its own flats. 

First night:   Askar Sii-Oiii filter
Second night:  Askar Ha-Oiii filter
Third night:  UV/IR Cut filter

I pre-processed them in WBPP and ended up with 3 unstretched images.  When I combined the Ha-Oiii and Sii-Oiii images, everything is good.  But when I added the 3rd image, the UV/IR Cut integrated unstretched image, I got this:




More strange, is here below is the unstretched integration of only the UV/IR image.  It looks fine. 



I tried combining the Ha-Oiii and the UV-IR Cut images and they look fine.  I also tried combining the Sii-Oiii and the UV-IR Cut images and they look fine too.  But when I combine all three these weird artifacts appear and I have ruled out dust motes because the individual integrated images do not have any calibration issues.  Why can't I combine all three images in Pixinsight's Image Integration Process?


Thanks,
Jerry


Looks like starnet++ artefacts to me from here.
Could try stretching more (or less) before removing stars.
Or maybe touching them up by had in gimp using heal.
If you have pixinsight there is a repair tool. I think?
Jon Rista avatar
Jon Rista avatar
Its definitely something flipped 180 degrees, and then subtracted. In the top image, you can see that every feature has effectively been mirrored. The dark arc of dust from the fish (basically its gills)… You can see it in the first image twice. And its 180 degrees rotated in the "subtraction"…

The darker structures below the fish, you can also see a similar thing. Its a bit less clear, due to the nature of the structures there, but you can also tell its been mirrored by the "subtraction." 

I would just try to do a 180 degree rotation of the UV/IR cut, and then re-try whatever you were doing. It may not be a perfect registration, but at least it would tell you if the issue truly is a 180 degree rotation. I would then just register the UV/IR cut to one of the other masters as a reference, and then combine it. Once registered, that should resolve any rotation issue.

EDIT:

On closer inspection…the problem may be that the UV/IR is just flipped horizontally, not actually rotated 180 degrees. Regardless, registering that (or those subs) to the SAME reference you used for all the other channels, should resolve the issue and allow you to stack or otherwise combine them without problems.
Helpful