What star reduction does
Star reduction makes stars smaller or dimmer so the nebula, galaxy or cluster behind them gets more attention. It is a cosmetic step that many astrophotographers apply late in processing, once the image has been stretched. Done lightly, it is hard to notice. Done heavily, it can leave dark rings, flat discs or an unnatural look.
PixInsight offers several routes. Which one suits you depends on your data, your tools and how much control you want.
Before you start
A few habits help whichever method you choose.
- Work on a non-linear image. Star reduction is most commonly applied after stretching, because that is when bloated stars become visible. Some people remove stars at the linear stage instead, so check which workflow you follow.
- Keep a copy. Duplicate the image, or use the History Explorer, so you can step back.
- Use a preview. Test settings on a small preview that contains both bright and faint stars before applying to the whole image.
- Judge at 100%. Zoom in to check the edges of stars, not just the full-frame view.
Good star reduction also depends on how tight your stars are to begin with. Poor focus, guiding or collimation makes stars bloated, and reduction only hides that. If you are unsure how sharp your data is, our guide to FWHM in astronomy explains how to measure it.
Method 1: Morphological Transformation with a star mask
This is the traditional approach.
- Build a star mask, for example with the StarMask process, so that stars are white and the background is black. Adjust the settings until the mask covers the stars you want to shrink without including nebulosity.
- Soften the mask slightly, for example with a small blur or by reducing its contrast, so the edit blends in. Hard-edged masks tend to leave visible rings.
- Apply the mask to the image.
- Open MorphologicalTransformation and choose Erosion (some users prefer Morphological Selection). Use a small structuring element, such as 3x3, and a low amount.
- Apply once, check, and repeat only if needed.
Small steps with a gentle amount usually look better than one aggressive pass. Remember to remove the mask before further processing so later steps do not affect only the stars.
Method 2: Star removal and recombination
Star removal tools split an image into a starless image and a stars-only image. Common choices are StarXTerminator (a commercial plug-in from RC Astro) and StarNet2. Workflow outline:
- Run the removal tool to produce a starless image and a stars-only image.
- Process the starless image on its own, with a stronger stretch or colour boost than stars could tolerate.
- Reduce the stars image, for example by lowering its brightness with PixelMath or a curves adjustment.
- Recombine the two. A screen blend is common. With PixelMath, a typical expression is
~(~starless * ~stars), wherestarlessandstarsare the image identifiers. Check the result and adjust the star brightness before blending.
This gives the most control, since stars and background can be stretched separately. The cost is extra steps and the risk of artefacts, covered below.
Method 3: Community scripts
Several free scripts wrap star reduction into one dialog. Bill Blanshan’s StarReduction script is a well-known example, and other scripts exist. They usually offer a live preview and a built-in mask for protecting small stars. Scripts are installed through PixInsight’s repository system, and the author’s page gives the current repository address. Because scripts change over time, read the author’s instructions for current options.
Choosing a method
| Situation | Often a good fit |
|---|---|
| Moderate star bloat, no extra tools | Morphological Transformation with a star mask |
| Dense star fields, faint nebulosity needing strong stretching | Star removal and recombination |
| Want fast results with a preview | A community script |
These are general leanings, not rules. Many people try more than one on the same image.
Common problems
- Dark rings or halos around stars. Often caused by a mask that is too tight or too hard-edged. Widen and soften the mask, or lower the amount.
- Flat, blobby stars. Too many erosion passes. Reduce the number of iterations.
- Blotchy starless background. In dense fields, star removal can leave uneven patches, and aggressive stretching exaggerates them. Gentler stretching and noise reduction before the final stretch may help.
- Colour loss in stars. Stars can lose colour if they are stretched hard. Processing the stars-only image separately and adding gentle saturation can restore it.
- Residual star fragments. Removal tools may leave pieces of very large stars. Cloning or touch-up may be needed.
Should you reduce stars at all?
Opinions differ. In forum discussions, some imagers remove stars routinely for faint nebulae and then blend them back gently. Others argue that stars give context to a nebula and prefer to leave them alone, or only reduce them lightly. Both are valid choices. A reasonable approach is to reduce stars just enough that they stop distracting, and then compare against the unreduced version before settling.
Where this fits in your workflow
Star reduction normally comes near the end, after calibration, stacking (see deep sky astrophotography for beginners), background correction, colour calibration and stretching. If your images come from automated capture software, our EKOS guide covers one option for gathering data.