Satellite Train Removal Question / Problem [Deep Sky] Processing techniques · Benjamin DeHaven · ... · 12 · 298 · 3

mcchurchmouse 0.00
...
· 
·  Share link
I imaged the Witch Head Nebula, NGC1909 a while back. Haven't done much because I have a fuzzy satellite train in it that I can't seem to stack out. I have tried several different rejection Algorithms including GESD. I have Large-Scale Pixel Rejection High enabled and I've tried both layers and growth at 1, 2, and 6. Nothing seems to even make it different. I'm not keen to throw away this data but I can't stand the fuzzy trail:

Sat Train Starless.JPG

Here is the zoomed in blink from 110 of the 216 files...



Can anyone help me? Thanks!

Benjamin
Like
ashastry 2.81
...
· 
·  Share link
I have found PixInsight's WBPP to not always deal well with trails, it's shoddy a lot of times to be honest even with good dithering, and I need to fix them later in the processing workflow.

I don't know what your processing workflow looks like, but a simple application of the Remove Tool in Photoshop should fix this just fine: just create a new empty layer on top of your base layer that has the image with the trails (you will want to export 16-bit non-linear tiffs), click the remove tool, check "sample all layers", uncheck the "remove after each stroke", click at the start of the trail, then hold down Shift and click again at the end of the other trail, and you should see a brush line overlapping the trail, and then click the apply button. This flow makes it non-destructive and you can choose to either keep the portions or the entirety of the "remove layer".
Edited ...
Like
a26413768 0.00
...
· 
·  Share link
https://pixinsight.com/forum/index.php?threads/new-in-pixinsight-1-8-5-new-large-scale-pixel-rejection.10683/
Like
WhooptieDo 10.40
...
· 
·  Share link
Welcome to the celestial equator.      Geosync satellites will be the bane of your time here.      While I have no great suggestions,  clone stamp or photoshop healing brush is your friend.

You could also attempt to manually reject them and restack a separate image, blend with photoshop in the end.  Chances are you will be close to rejecting everything though.  Your rejection algos are doing exactly what they're supposed to be doing.    You'll just need to get creative here.
Like
mcchurchmouse 0.00
Topic starter
...
· 
·  Share link
Shun-Chia yang:
https://pixinsight.com/forum/index.php?threads/new-in-pixinsight-1-8-5-new-large-scale-pixel-rejection.10683/

I agree the pictures in this post are exactly what I want to do. Aren't I already using this tool by enabling Large-Scale Pixel Rejection in WBPP? I'm playing around with the ImageIntegration tool right now but honestly, I use WBPP and am having to learn to use ImageIntegration from scratch... This pre-release post on the pixinsight site seems big on promises and short on "how to use the tool". But at least I know it can be done, it's user error on my end.
Like
mcchurchmouse 0.00
Topic starter
...
· 
·  Share link
Brian Puhl:
Welcome to the celestial equator.      Geosync satellites will be the bane of your time here.      While I have no great suggestions,  clone stamp or photoshop healing brush is your friend.

You could also attempt to manually reject them and restack a separate image, blend with photoshop in the end.  Chances are you will be close to rejecting everything though.  Your rejection algos are doing exactly what they're supposed to be doing.    You'll just need to get creative here.

Forgive the noob question. Judging by stellarium, the sky is fixed in regards to the celestial equator. So any time of year from my location at least, the Witch Head and Orion for that matter are very close the to the celestial equator. Meaning I can never shoot them without the train of satellites?
Like
Stephen.J 1.43
...
· 
·  Share link
Maybe you could split the R G B channels and use the clone stamp on each; then combine again.
Like
mcchurchmouse 0.00
Topic starter
...
· 
·  1 like
·  Share link
I think I got it... Forgive the WBPP, I am trying to duplicate my ImageIntegration settings. But if you look at the integration2 starless image you can see the trail is basically gone. I switched back to GESD (ESD) and upped my ESD Outliers to 1.00 and ESD Significance to 1.00 as well... And in Lare-Scale Pixel Rejection I turned on High and put both factors at 2.

Sat Train Success.JPG
Like
mcchurchmouse 0.00
Topic starter
...
· 
·  Share link
Benjamin DeHaven:
I think I got it... Forgive the WBPP, I am trying to duplicate my ImageIntegration settings. But if you look at the integration2 starless image you can see the trail is basically gone. I switched back to GESD (ESD) and upped my ESD Outliers to 1.00 and ESD Significance to 1.00 as well... And in Lare-Scale Pixel Rejection I turned on High and put both factors at 2.


Well, my attempt to get WBPP to give me the same results failed. The streak was still there. So apparently it only works in ImageIntegration. The only setting I can't find to duplicate is "Clip low pixels", "clip High Pixels", "Clip low range", and "Clip High Range". Now to figure out how to Drizzle after ImageIntegration...
Like
smcx 3.61
...
· 
·  Share link
Man i hope someone figures this out. I don’t want to do all the steps manually, and I also find I have to cull images with trails because I can’t get WBPP to remove them. Even faint trails get stacked ¯\_(ツ)_/¯
Like
ScottBadger 7.63
...
· 
·  1 like
·  Share link
I find the Winsorized algorithm to work a lot better on satellite trails than GESD, though it's possible my GESD settings aren't optimal. Also, there's a new Seti Astro script called Blemish Blaster (sort of an auto clone-stamp) that has just been updated with a satellite trail removal tool. https://www.setiastro.com/pjsr-scripts (scroll to the bottom)

Cheers,
Scott
Like
mcchurchmouse 0.00
Topic starter
...
· 
·  Share link
So I can get the trail out but when I drizzle in DrizzleIntegration I lose a ton of contrast and detail in the Witch Head dust compared to the WBPP drizzle integration with the trail. Any idea why this might be?
Like
mcchurchmouse 0.00
Topic starter
...
· 
·  1 like
·  Share link
Ok, for those following along on the edge of their seats… I figured out that the high rejection is eating the detail from the Witch's Head leaving just formless clouds. So I'm starting to use the SKILL script to color over the satellite trails. 268 frames might take me a week but thats the price we pay for a good image, yes? So now I'm cranking down the high rejection and cranking up the low rejection… I'll report how it goes!

Benjamin
Like
 
Register or login to create to post a reply.