New star removal tool

Tony GondolaWillem Jan DrijfhoutAloke Palsikarpatrick cartou
24 replies629 views
Tony Gondola avatar

There’s finally a new star removal tool out there that’s worth looking at. It’s called Zenith-V1 by Syqon. It’s available as a script for both Siril and SetiAstroSuite, not sure about PI but I can’t imagine it won’t make its way there at some point.

The first time I tried it on some M-51 Ir data, I thought it wasn’t working well. My resulting starless image had a lot of small dots left behind. My first thought was “fail” but when I looked closer I realized that those dots were all very small field galaxies that are normally mistaken for stars and removed by Starnet++ so that impressed me. I tried it on a few other images and I have to say it worked very well. The biggest overall difference is that it leaves a very clean starless image behind. Almost no clumps or ghosts of the brighter stars in the field. If you’ve spent time with blemish blaster or a clone stamp cleaning your starless images up, you’ll know what I mean. The only thing to keep in mind is that it is used post stretch, not pre. It doesn’t need to be the final stretch but it needs at least a light statistical stretch to work properly.

Well Written Helpful Insightful Engaging
John Stone avatar
Just bought it and tried it out.

It does work better than StarXTerminator in the sense that it doesn't remove nebula around the star

StarLess v1






StarXterminaor

Tony Gondola avatar

That’s what I’ve found to be it’s greatest strength. That and making a very clean starless image, at least when compared to Starnet++

John Stone avatar
Have you found a way for the tool to produce a stars-only image?

The image above I had to create with PixelMath expression subtracting the normal image from the starless image produced by the tool to create the stars-only image.

This isn't as nice as StarXTerminator since the background does not end up pure-black (0 ADU count)
Well Written Insightful Respectful Concise
Tony Gondola avatar

The version I use in Siril has an check box option to create a star mask. The commercial version must have something similar.

Willem Jan Drijfhout avatar
Wouldn't the need for a stretch before star removal make it a bit pointless? The whole reason to eliminate the stars is to stretch the non-star areas without having to worry about the stars (clipping etc).
Well Written Insightful Respectful Engaging
SonnyE avatar

I’m curious if anybody has published a starless image on Astrobin to see if it can plate solve the image.

I’d try it myself, but I don’t drastically alter my images.

Salvatore Iovene avatar

SonnyE · Feb 10, 2026, 06:21 PM

I’m curious if anybody has published a starless image on Astrobin to see if it can plate solve the image.

AstroBin cannot plate-solve starless images. Plate-solving works by detecting stars and triangulating positions against reference databases.

Well Written Concise
andrea tasselli avatar
Hence the name: plate-solving.
SonnyE avatar

Salvatore Iovene · Feb 10, 2026, 06:23 PM

SonnyE · Feb 10, 2026, 06:21 PM

I’m curious if anybody has published a starless image on Astrobin to see if it can plate solve the image.

AstroBin cannot plate-solve starless images. Plate-solving works by detecting stars and triangulating positions against reference databases.

As I suspected, Salvtore. Thank you for clarifying that.

Tony Gondola avatar

Willem Jan Drijfhout · Feb 10, 2026, 05:52 PM

Wouldn't the need for a stretch before star removal make it a bit pointless? The whole reason to eliminate the stars is to stretch the non-star areas without having to worry about the stars (clipping etc).

Not a full stretch or you’d be largely correct. That said, I’m finding that it also works fine pre-stretch as well.

Mark Fox avatar

This appears to be great stuff, from what I am reading on the SyQon website. I’ve always been a bit disappointed with the blotchiness of starless output.

As of today 10Feb, it appears that the software has been rebranded to SyQon Starless. To get this in Siril, look for the Python script SyQon_starless.py. I didn’t see any mention of the “Zenith” name.

Helpful Concise
Cyril Richard avatar

Tony Gondola · Feb 5, 2026, 12:18 AM

There’s finally a new star removal tool out there that’s worth looking at. It’s called Zenith-V1 by Syqon. It’s available as a script for both Siril and SetiAstroSuite, not sure about PI but I can’t imagine it won’t make its way there at some point.

Hello.

Zenith is a model developed exclusively for Siril from scratch :). We now have version 1.3 of the model available. Therefore, neither SetAstro nor PI will be able to use it because SyQon has drafted a very specific license regarding this matter.

Well Written Helpful Concise
Victor Van Puyenbroeck avatar

Willem Jan Drijfhout · Feb 10, 2026, 05:52 PM

Wouldn't the need for a stretch before star removal make it a bit pointless? The whole reason to eliminate the stars is to stretch the non-star areas without having to worry about the stars (clipping etc).

I would need to ask the author, but I believe Starnet++ was trained on non-linear data.

I always assumed that, if the ‘linear data’ checkbox is ticked in PixInsight, Starnet will apply a stretch before the star removal step. Once complete, an inverse stretch is used to transform the starless image back to a linear state.

Well Written Concise
Willem Jan Drijfhout avatar
Victor Van Puyenbroeck:
Willem Jan Drijfhout · Feb 10, 2026, 05:52 PM

Wouldn't the need for a stretch before star removal make it a bit pointless? The whole reason to eliminate the stars is to stretch the non-star areas without having to worry about the stars (clipping etc).

I would need to ask the author, but I believe Starnet++ was trained on non-linear data.

I always assumed that, if the ‘linear data’ checkbox is ticked in PixInsight, Starnet will apply a stretch before the star removal step. Once complete, an inverse stretch is used to transform the starless image back to a linear state.

Ah, yes, that makes sense. Thank you.
Willem Jan Drijfhout avatar
The website of the developer of this tool (Syqon) also shows some other products. NovaStudio, one of them, claims to be a complete processing suite. There is very little information in the community on this software, let alone a full review. The few comments I did come across were unanimously negative, to the level of 'does not work as advertised'. Also the company does not offer a trial period.

Does anyone have experience with this tool? And if so, what do you think of it?
Well Written Engaging
Cyril Richard avatar

They offer a free model in Siril. So better than a trial.

You can try it right now.

Aloke Palsikar avatar

I tried the new SyQon Star Removal tool in Siril and the result is very good. Does a bit of denoising and also the artefacts of Star Removal which one sees in StarNet are either not there or very minimal

I use a Intel Macbook and still the performance was decent. All the relevant scripts required for installation are given in the Comments and work beautifully


Good tool from Zenith for Siril

Willem Jan Drijfhout avatar

Cyril Richard · Feb 11, 2026 at 11:20 AM

They offer a free model in Siril. So better than a trial.

You can try it right now.

I was referring to Novastudio. Is that also available within Siril?

Well Written Respectful
patrick cartou avatar

Hi all

I’m just coming to test SyQon Star Removal tool and obtained this result:

15:00:08: Début du script /home/cartou/.var/app/org.siril.Siril/data/siril-scripts/processing/SyQon-Starless.py

15:00:08: Installing scipy. This may take a few seconds...

15:00:09: Collecting scipy

15:00:10: Downloading scipy-1.17.0-cp313-cp313-manylinux_2_27_x86_64.manylinux_2_28_x86_64.whl.metadata (62 kB)

15:00:10: Requirement already satisfied: numpy<2.7,>=1.26.4 in /home/cartou/.var/app/org.siril.Siril/data/siril/venv/lib/python3.13/site-packages (from scipy) (2.3.5)

15:00:10: Downloading scipy-1.17.0-cp313-cp313-manylinux_2_27_x86_64.manylinux_2_28_x86_64.whl (35.0 MB)

15:01:26: ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 35.0/35.0 MB 493.5 kB/s eta 0:00:00

15:01:26: Installing collected packages: scipy

15:01:31: Successfully installed scipy-1.17.0

15:01:31: Successfully installed scipy

15:01:31: Traceback (most recent call last):

15:01:31: File "/home/cartou/.var/app/org.siril.Siril/data/siril-scripts/processing/SyQon-Starless.py", line 95, in <module>

15:01:31: th = s.TorchHelper()

15:01:31: ^^^^^^^^^^^^^

15:01:31: AttributeError: module 'sirilpy' has no attribute 'TorchHelper'

15:01:31: Le processus Python (PID : 3066) s'est arrêté avec le statut 1

What does it mean ?

Thanks by advance.

Best regards.

Patrick

Aloke Palsikar avatar

Check the comment text in the script. You will need to install Torch through the 5 lines of script given in the text comments. Run it within Siril itself. This should work

patrick cartou avatar

Thanks a lot Aloke !

I don’t know really how and where exactly insert these 5 lines of script, so I’ll wait for the Version 1.4.2 of Siril…

Patrick

Respectful
Aloke Palsikar avatar

Watch this You Tube Video by Deep Space Astro and all your doubts will be clarified. A complete step by step process has been explained

https://www.youtube.com/watch?v=8cioH5lctvU

Hope this helps

ZigZagZebraz avatar

The earlier model did not work well for me. But, the Zenith works very well.

Almost no residue in Oiii of bright star compared to Starnet++. Easily manageable with stretch than resorting to Blemish Blaster.

Thanks Tony for the heads up, the Siril team and Syqon for a great tool.

Helpful Respectful Concise Supportive
patrick cartou avatar

Aloke Palsikar · Feb 11, 2026, 02:58 PM

Watch this You Tube Video by Deep Space Astro and all your doubts will be clarified. A complete step by step process has been explained

https://www.youtube.com/watch?v=8cioH5lctvU

Hope this helps

Thanks again, Aloke !

Patrick

Related discussions
PixInsight New Tool - Multiscale Gradient Correction Test
Yesterday a new version of PixInsight 1.9.0 was released. The release contains some nice features, but by far the most important one is the MultiscaleGradientCorrection tool (I): https://www.youtube.com/watch?v=BHE3J069NCo It is supposed to do backgr...
Dec 21, 2024
Both posts discuss new astronomy software tools that initially didn't perform well with their default settings when the users first tested them.
M33. Should there be any green in the image at all?
Especially since the introduction of SPCC colour calibration into PixInsight I have given up using colour noise reduction tools to remove green pixels from images – because it ought to be unnecessary and will only alter the colour balance. Howe...
Jan 7, 2025
Both posts discuss digital image processing tools and techniques used in astrophotography to enhance astronomical images.
Fixing scattered light issue
Hi all After few days of Ha & Oiii acquisitions I finally checked my result. And of course as always I found an issue on integrated images I have quite nasty scattered light in right top corner. Of course it's on both images so I don't ha...
Jun 28, 2025
Both posts discuss technical problems encountered during astrophotography image processing.
What is real and what is not? How to know what to include...
Hi, as I always stated, image processing isn't my thing. But while forcing myself to progress in this field, I ran into some issues no tutorial seems to care about. Let's explain the problem. For my latest images I took the time to collect at...
Jul 11, 2024
Both posts discuss challenges and experiences with astrophotography image processing techniques and software tools.