npx skills add MZqk/deepsky-skill/siril-moon-stacking What it is. A professional lunar astrophotography / lucky-imaging post-processing skill . It takes camera RAW (.ORF, .CR2, .NEF, .ARW, …) or FITS/SER sequences and produces high-resolution natural-detail JPG/TIFF plus geologically-rich “mineral moon” color renders.
In one line: an AI-driven lunar lucky-imaging stacker that pairs Siril’s native processing with a Python subpixel-registration/ADC plugin, engineered for real optical fidelity rather than synthetic detail.
Hi ,
Thanks for the feedback! It’s great to hear that Siril’s development version is bringing native stacking capabilities.
To clarify, the primary focus of siril-moon-stacking goes well beyond just the alignment/stacking step. It is designed as an end-to-end processing pipeline tailored specifically for lunar and mineral moon imaging, which includes:
Automated Optical Inference & Deconvolution: Deriving telescope focal length via lunar limb fitting to drive physically accurate Airy disk PSF deconvolution.
Artifact Damping: Sub-pixel anti-ringing damping to eliminate dark halos (doughnut artifacts) along steep crater edges.
Geological Mineral Moon Pipeline: L/RGB chrominance separation, terminator phase-reddening defense, and atmospheric extinction gradient compensation.
Mosaic Calibration Locking: Anchor profile locking across multi-panel tiles to ensure zero-seam photometric and color consistency.
Once the native stacking feature in the dev version matures and is exposed to the CLI, we’d love to integrate it as a backend option in our pipeline!
Just to clarify, this project is built as an AI Agent Skill rather than a standalone replacement. Its key advantage is leveraging LLMs (Large Language Models) to intelligently and adaptively generate optimized Siril .ssf scripts on the fly based on image metadata, moon phase, and seeing conditions—acting as an AI co-pilot that automates and orchestrates Siril rather than replacing it.
