Auto-StarMask Studio

Auto-StarMask Studio is the place where you build star masks and starless layers that behave well under real processing: deconvolution, star reduction, aggressive stretches and creative compositions. It uses multiscale detection so small, medium and large stars can be controlled separately.

Quick Start (5-10 min)

Load a linear or gently stretched deep-sky image with reasonably clean stars.
Choose a detection mode (all stars, small/medium/large emphasis) depending on your goal.
Adjust sensitivity/threshold so stars are detected without grabbing most of the nebula.
Tune core/halo balance and optional feather to control how soft or hard the mask looks.
Toggle between Mask and Starless previews to validate the result.
Export the mask, the starless image, or both, and use them in your processing workflow.

Controls

Control	Default / Range	What it does	When to use
Detection mode	All stars	Chooses which scales (small/medium/large) are emphasized in detection.	Use “All” for general masks, or target small/large stars for specialized workflows.
Sensitivity / threshold	Medium	Controls how aggressively pixels are classified as stars.	Increase to catch faint stars; decrease to avoid grabbing nebulosity or noise.
Core emphasis	Balanced	Adjusts how strongly star cores appear in the mask.	Emphasize for decon helper masks; reduce when you want very gentle star reduction.
Halo emphasis	Medium	Controls how much of the star’s outer halo is included.	Increase for big, bloated stars; decrease to avoid clipping diffuse halos around bright stars.
Feather / smooth	On	Softens mask edges to avoid ringing and harsh transitions.	Keep enabled for most workflows; reduce only when you need very tight masking.
Starless mode	Off	Generates and previews a starless version of the image using the computed mask.	Use for creative nebula processing or to inspect whether bright structures are being eaten.
Export mask / starless	-	Exports the star mask, the starless image, or both.	Use when you are happy with the preview and ready to integrate the outputs into your pipeline.

Example workflows

Deconvolution helper mask

Load a linear luminance (or L-equivalent) image.
Use a mode that emphasizes small and medium stars.
Increase core emphasis slightly and keep halos modest.
Export the mask and use it to protect star cores during deconvolution.

Gentle star reduction

Create a mask that includes most stars but with moderate halo emphasis and good feathering.
Apply a mild star-reduction process (inside or outside AIT) with this mask active.
Check that bright nebula structures stay untouched while the star field becomes cleaner.

Aggressive starless nebula work

Enable Starless mode and fine-tune sensitivity so stars disappear but nebula stays intact.
Export both starless and star-only layers.
Process nebula detail on the starless layer, then recombine with softened stars later.

Tips & notes

Linear or only gently stretched data produces cleaner masks; very hard stretches can make detection noisy.
If nebulosity is being eaten, reduce sensitivity and/or adjust mode to de-emphasize very large scales.
Always zoom in and pan across the frame in the preview; small artifacts are easier to spot than in the final composite.

Troubleshooting

Mask includes too much nebula → lower sensitivity, use a detection mode that favors smaller scales, or consider masking bright nebula regions before detection.
Small stars missing from the mask → increase sensitivity or switch to a mode emphasizing small stars.
Starless image shows obvious artifacts → refine the mask (especially around bright stars) and regenerate the starless layer.

Updated 2025-11-15