A call for help to the Pixinsight experts - I am unable to remove the satellite trails from the satellite highway near the celestial equator - Forums

Topic starter

Dec 28, 2022 · Edited: Dec 28, 2022

Hello Pixinsight experts,

I recently imaged The Angel Nebula. From my location it stands on the highway of the geosynchronous satellites. I am trying to remove the satellite trails in PI but unfortunately for me - I am unable to do so. I tried all the stacking algorithms (with the default settings) but in each of them some trails remain. Here is what I see:

I would like to kindly ask for your support in finding a solution to my problem. I am sharing some frames if they are needed for testing.

https://1drv.ms/u/s!AsUSLdoc-U2un3sYD5QbChX0VTcc?e=4sk1q9

The frames are calibrated, cosmetic corrected, debayered and star aligned. These are frames from the green channel of my camera - ASI2600 attached to a RASA8.

Thank you in advance!

CS,
Ivan

Well Written Respectful

Tom Engwall

Dec 28, 2022

Do you dither between each frame?

Ivan Raichev

Topic starter

Dec 28, 2022

Tom Engwall:
Do you dither between each frame?

Yes, I dithered between each two frames.

GalacticRAVE

Dec 28, 2022

downloaded the file, integrated with ESD rejection and standard paremeters - no tracks in the final image. Matthias

Ivan Raichev

Topic starter

Dec 28, 2022

downloaded the file, integrated with ESD rejection and standard paremeters - no tracks in the final image. Matthias

Thanks, but did you run a background neutralization and auto stretch? Can you share your result please.

I shared my result with the standard parameters of ESD in the photo above - second row, 3rd image from left to right - the trails are visible...

GalacticRAVE

Dec 28, 2022

When I ran a simple ABE, i see some very faint residual. Probably not critical. Setting significance to 0.2 gets rid of it as does using median.

Ivan Raichev

Topic starter

Dec 29, 2022

When I ran a simple ABE, i see some very faint residual. Probably not critical. Setting significance to 0.2 gets rid of it as does using median.

Thank you. I found where to input the significance setting but it didn't help

I still see the residual from the saturated trails... Where do I use the median setting? Is it part from some of the other stacking algorithms?

CS,
Ivan

GalacticRAVE

Dec 29, 2022

the first option under image integration - combination. you have the choice between min, max, average and median. From SNR point of view, average is the better method. Median is always good to reject/downweight outliers. So I would use it only when the other methods do not lead a satisfactory result. Alternatively, in particular if only a few frames are affected you can explicitly blacken them out (handywork with pixelpath ….) and then ignore them at stacking with the clip low range option, also works for occasional dust mots etc (Adam Block has a video on this on youtube)

Helpful

Bogdan Borz

Dec 29, 2022

Ivan,

You should activate the Large-scale pixel rejection button in the image integration menu. Then play with the parameters in order to capture the satellite trails. You can also try to modify the sensitivity of the Pixel rejection 2 settings in the above menu. Satellite trails are about pixel rejection when stacking, so you need parameters that place them outside the range of accepted neighbouring pixels.

CS,

Bogdan

Helpful Concise

Dave & Telescope

Dec 29, 2022 · Edited: Dec 29, 2022

Ivan,
Those do not look like satellite trails to me. Satellite trails I have seen are typically random. Maybe because of your location it's different but these are uniformally vertical which is unusual. Are we sure this isn't banding from the sensor? There is a section in the lights tab in WBPP where you can correct column defects typical for CCD (see attached). Not sure it will work on a CMOS sensor but you could try it.
Also you mention "debayering" in your workflow. If this is a mono camera with filters you don't have to debayer.

HTH
Dave

PI defects.png

Helpful Insightful Respectful

Ivan Raichev

Topic starter

Dec 30, 2022

Dear Dave,

Thank you for your reply. This isn't banding from the sensor. They are satellites for sure. Here is a screenshot from SkySafari - green dots are satellites. These are all the Astra, TurkSat, etc. satellite TV shit... As I said - a highway...

Here are the satellites as a GIF in order to see what I mean:

https://astrob.in/gdso6t/E/

As it was suggested by Bogdan - a solution is to increase the pixel rejection but unfortunately for me this decreases the brightness of the imaged object as well which sucks

CS,
Ivan

Helpful Concise

Dave & Telescope

Dec 30, 2022

Wow Ivan that's crazy! I've never seen anything like that ever. There are so many it has the net effect of looking exactly like ccd banding. Yes I agree then with Bogdan the only other thing I have done is use the large scale pixel rejection option. What I have done is to leave "small" at default, increase "large" to 3 to start off. If that isn't enough then I dial down the "high pixel rejection 2" slider until it's good. I do wonder though what would happen if you tried the column defect correction since it looks so much like it.

Anyway thanks for sharing that video! I learned something quite sobering. The whole satellite thing is a real menace for sure, way more than I thought.
Regards
Dave

Engaging Supportive

Bogdan Borz

Dec 30, 2022

Ivan,

That gif is absolutely crazy. We are clearly heading towards an extinction of the night sky in the 5-10 years to come when Starlink, Amazon Kuiper, Chinese etc. will have fully deployed their fleets.

It’s becoming clearer why you’re unable to fully eliminate them. They are thin structures and not too bright. Maybe one supplementary solution would be to choose a good signal frame, free from satellite trails and use it as a reference in the Normalise Scale Gradient script. It will probably decrease their brightness in the image.

Who knows, maybe the effort to reduce the Starlink satellites brightness without completely eliminating it, will create more problems for excluding them on images like in the present case.

Bogdan

Helpful Insightful Respectful Engaging

Victor

Mar 20, 2023

Hello Ivan,
I batch processed your images and may be of help.
If you have not had any solution I can send you a copy of my batch processed and image .
Vic
520-878-6860

Ivan Raichev

Topic starter

Apr 14, 2023

Victor:
Hello Ivan,
I batch processed your images and may be of help.
If you have not had any solution I can send you a copy of my batch processed and image .
Vic
520-878-6860

Hello Victor,

Yes, please send me your solution. I didn't find a way to remove the satellite trails without sacrificing details.

Thanks in advance!

CS,
Ivan

Well Written Respectful