IronMarcoPolo · Oct 20, 2025, 01:22 AM
Got it! Thanks! And thanks for putting your process too.
You’re most welcome. My workflow is even more detailed. If you want, can copy-paste the one I used on my images posted.
This page gives a good start with Siril, if you’re interested.
https://perfectastronomy.com/astrophotography/processing-stacking-fits-seestar-s50-siril/
That would be great if you are willing to share. I will check out your webpage too. Thanks!
This is the workflow I used for the Heart Nebula. I do get about 20 to 30 hours of data to reveal fainter details. (With all the explanations, the workflow is in the bottom of this comment).
I use the OSC preprocessing script, modified to include the true drizzle as the earlier one had a debayer step which was not working with 1.4.0 Beta 2. I am guessing the official script is modified suitably.
The main python scripts I use are GraXpert, DBXtract, Cosmic Clarity ones and Continuum subtraction. If they are not there, can be downloaded using “Get Scripts” menu or going to Preferences, scripts and downloaded there. For Cosmic Clarity, you might need to download it from Seti’s Astro website. Use the mirror site (google drive), if available, which is faster. Github is too slow.
(Step 4) DBXtract will extract Ha, Oiii and Sii, based on your sensor spectral response. You can select your astro camera sensor. Do not think it has DSLR sensors. For me Oiii signal is usually weak. I use an Antlia Triband RGB Ultra ii. SV220 will also work well for Ha and Oiii. Not sure about sulfur band.
Continuum subtraction, just followed the video (https://www.youtube.com/watch?v=6QQlwel2dAU).
After continuum subtraction, use Ha for R and Oiii for G and B for B for RGB recomposition (Siril menu). Have to do linear match, auto brightness adjust and align (all in the same menu panel).
For SPCC (I have the 10GB Siril SPCC data base downloaded). Found it somewhere and cannot find it again. May be this video will help (https://www.youtube.com/watch?v=kNRGhY6W3SA).
After RGB recomposition, I do the color calibration. You can get away to Photometric color calibration. For SPCC, Siril will complain that there is no header data. This is because, DBXtract probably remove it in the Red layer. If you have not closed Siril, Astrometry dialog will still have the same object name. Just do astrometry and SPCC. If Siril had been closed, just put in the object name (Sh or IC or NGC), focal length and camera pixel size in micron.
Before GHS, if nothing is visible on the screen, do the python script for statistical stretch (stretch factor 0.2, linked and normalize ticked ON). Will have a nice pre-stretch. Sometimes, I had to keep the stretch factor to 0.1, if 0.2 is too bright.
For GHS (Step 8 below), I follow Siril’s official tutorial (https://siril.org/tutorials/ghs/). Type in stretch factor of 2.0 and symmetry point of 0.5. Use the local stretch intensity slider (to negative) to make the curve close to hugging the diagonal.
In the first pass, adjust the highlights close to 0.5 to preserve them. In the second pass, adjust the low lights (shadows) and if required the highlights sliders.
In the Inverse GHS stretch, type in 0.5 in symmetry point. May be about 1 in stretch factor and adjust shadows and highlights.
For Step 9, star recomposition, get the starless that was processed and the starmask into the menu. Typically do not have to stretch starless. For the starmask, type in 7 in the stretch factor. Zoom in to your Siril object window (mouse wheel) to may be 400%. Adjust black point. Usually, 0.03 to 0.04 is enough. You should barely see the black circle around stars.
For post processing I use ON1 Raw Max. There is an AI denoise, which is NOT regenerative. in Raw Max 2025 edition, the AI Denoise 2024 (not the high detail one) works well. Lum denoise of 70 and enhance details of 50 (defaults) and color denoise (100, default) works well.
Yesterday I found that running GraXpert Denoise with a value of 1.0 on the each continuum subtracted layer and the B layer (which are used for RGB composition) produces an end product with very minimal noise. But, might
1. RGB Align
2. Astrometry, Crop
3. Background Extraction: Siril BG (RBF, 1.0, 100, 0.3, Dither ON), Siril Python Script, Graxpert background extraction (1.0)
4. DBXtract, RGB Extract, Continuum Subtraction, Recombine RGB, R=Ha, G=Oiii Continuum Subtraction
5. Spectrophotometric Color Calibration (GAIA)
6. Starnet
7. Starless - Siril GraXpert Denoise (1.0)
8. Siril GHS: 2 iterations, Inverse GHS (1X), Black Point, Histogram stretch, Cosmic Clarity non-stellar sharpening (Default), GraXpert Denoise (1.0)
9. A. Star Recomposition from Starmask and Starless.
10. Save as .png
11. ON1 RawMax 2025, png for web
The above will give you a good start. Eventually, you will find your own settings based on the acquisition and other factor. Hope you find it useful.
All the best and Clear skies.
