Star reduction in astrophotography is a technique used to reduce the appearance of stars in an image. This technique is often used to reduce the visual impact of large, bright stars, which can be distracting or overpowering in an image. It can also create a smoother, more even background in an image, which can help highlight the more subtle details of the photographed object. When reducing stars, the goal is to reduce the impact the stars make on your final image; you don’t want to eliminate smaller, weaker stars since that affects your image’s accuracy and truthfulness yet you want to reduce the impact the stars make on your image.

There are a few ways to reduce the stars in your image in PixInsight:

  • BlurXTerminator (must be applied early in the workflow)
  • MorphologicalTransformation
  • EZ Star Reduction script – NOTE: The EZ Processing Suite is no longer available

Each process has its advantages and disadvantages that become apparent when you observe the changes each process makes to an enlarged part of your image. The effects are subtle; however, depending on how you plan to present your final image, the effect can be quite significant. The biggest differentiator among all the processes is their effect on the smaller and weaker stars. Except for one technique, the smaller stars range from being completely attenuated to being so reduced and color changed that the effect actually works to reduce the finer details of your images.

The rest of this article presents Bill Blanshan’s Star Reduction technique and demonstrates why it is superior to other star reduction techniques. At the end of the article, I explain how to obtain the PixelMath expressions and how to use them. By adopting this approach, photographers can achieve superior results and make their images stand out from the rest.

Star Reduction Technique

Bill Blanshan’s Star Reduction technique uses PixelMath expressions to reduce stars. The following is an image where the stars have been reduced using Blanshan’s Star Reduction technique, method #3:

Swipe over the before (left) and after (right) several times to see the difference between the two images.

When it comes to reducing the appearance of stars in an astrophotography image, it’s important to strike a balance between removing unwanted visual distractions and preserving the finer details. One method that has been shown to be effective is a technique that reduces the overall appearance of stars while preserving even the smallest ones in the image.

When using this method, the difference between the original image and the modified version is intentionally subtle. The goal is to reduce the prominence of stars in the image without completely removing them. This way, the image retains its level of detail while still having a cleaner overall appearance.

By preserving the smallest stars in the image while reducing the overall appearance of the stars, this technique strikes a balance that can really make astrophotography images stand out.

The effect is more apparent when you zoom into part of the image and take a 3-D plot of a preview area, as I have done below:

If you look closely at the before (left) and after (right) images and compare the smaller and weaker stars, you’ll notice that this technique does a great job of preserving these stars while reducing the appearance of the larger ones. The brightness of the stars is also maintained, which helps preserve the overall look of the image.

This approach has been shown to be superior to other methods that are outlined in the rest of the article. To demonstrate this, I’ve included before and after 3-D plots of the same image area, so you can see the difference in the results.

Note for Patreon Subscribers

Patreon Pro and Premium subscribers can download a complete PixInsight project that includes all of the before and after 3-D plots, as well as the original images and preview area I used to create the plots, making it easy to experiment with the various approaches. The project includes a ReadMe that documents all of the artifacts in the project to make it easier to use.

Comparison of Star Reduction Methods

I compare and contrast the following star reduction techniques

  • BlurXTerminator
  • MorphologicalTransformation
  • EZ Star Reduction
    • Morphological Selection
    • Adam Block’s Method

In each case, I present a 3-D plot of the same region within the same image so that you can compare the before and after 3-D plots.

BlurXTeminator

You need to use BlurXTerminator when your image is in its linear state and has minimal processing applied to it. The following is a before (left) and after (right) 3-D plot of a preview of the linear image:

Since the image is linear, it has very little information – the stars are pretty much all you can make out in the rendering because the image has not been stretched.

The effect on the larger stars is to make them smaller without affecting their color and this is great. The problem is with the smaller stars. For example, the two stars in the center: if you repeatedly swipe over them, you’ll notice that the color of the stars changes from mostly blue to different colors; the stars also become much smaller and are much less visible as a result.

Let’s look at the MorphologicalTransformation process next.

MorphologicalTransformation

The following is a before (left) and after (right) 3-D rendering of a preview area – swipe over the image to view the changes:

The process has the biggest effect on the smaller stars – they are attenuated quite a bit, and much less visible as a result. The larger stars are slightly reduced, but the effect of MT is greatest on the smaller stars where some of the colors have changed too.

Next, we’ll look at the EZ Star Reduction script.

EZ Star Reduction

Note: This is no longer available

The EZ Star Reduction script operates on your stretched image and requires a star mask as well as a starless image to work properly. The script implements two star reduction techniques: Morphological Selection and Adam Block’s method. While both methods appear to work well, when you zoom into part of the image, and take a 3-D plot, the changes become more apparent.

Morphological Selection

The following comparison block demonstrates the before (left) and after (right) 3-D renderings of a part of the image before and after applying the script:

You’ll find that the smaller stars are attenuated the most and their color changes. The larger stars are also attenuated around the base where some details have been removed completely.

Adam Block’s Method

This was the recommended way of reducing stars and is still quite good at what it does; however, you’ll notice that the effect is uneven over stars of various sizes:

For the smaller stars, they get reduced in intensity as well as change color and size. The larger stars become narrower at their base while also losing some details; however, the overall effect is star reduction yet the smallest stars are affected the most.

Blanshan’s PixelMath Star Reduction

The following is a comparison of methods one and two of Bill Blanshan’s Star Reduction techniques; compare these two results to the first and note the significant differences.

Method 1

This is called the transfer method of reducing stars. The method reduces the exposure of the star image and merges that with the starless image to generate the final result. This is a comparison of the before and after zoomed-in 3-D plots:

You’ll find that the larger stars do indeed become smaller at their base and they do not lose intensity. The effect is most noticeable on the smallest stars which get reduced a lot; some barely rise above the noise threshold (left side of the 3-D plot). While this effect reduces the stars overall, it does so with the loss of the smallest stars, and depending on how you plan to present your image, this effect could be significant.

Method 2

This method is referred to as the halo method because it reduces the stars’ halo. Swipe over the 3-D plot t view the differences:

The process reduces the stars’ bases while preserving intensity for the brightest stars. The smaller stars also get reduced and change in color and this effect could be significant, depending on how you intend to present your image.

Comparing Results

Based on these results, it’s clear that method #3 is the best overall choice. This technique does a great job of preserving the smallest stars in your image while also reducing the size of the larger stars, which helps your image maintain its level of detail. It’s worth noting that the differences between the different methods can be quite significant, particularly if you plan to present your image in a large format or at full resolution. Overall, method #3 is the best option for preserving the finer details in your astrophotography image.

Acquiring Blanshan’s PixelMath Expressions

Blanshan made his PixelMath expressions into a set of process icons for you to download; you can download them from his YouTube channel, where there’s a link to his Google Drive, and I have made the file available at the link below:

Download Bill Blanshan’s PixelMath Expressions

Using the PixelMath Expressions

You’ll find four PixelMath expressions in the file that you downloaded:

  • CloneForStarless
  • Star Reduction Method 1
  • Star Reduction Method 2
  • Star Reduction Method 3

Use the following directions to use the expressions:

  1. Merge the process icons with those in PixInsight your workspace.
  2. Use StarNet, StarNet V2, StarXTerminator or another process to produce a starless version of your image. When you finish, you should have the original and starless images available on your PixInsight desktop.
  3. Drag the CloneForStarless PixelMath expression to your starless image (this just clones your image and names it ‘Starless’ and makes it easier to work with each of the three methods).
  4. Drag one of the Methods to your original image to execute it.

You can compare the effects of the different methods by cloning your original image and using each method on your images.

YouTube Video

You can learn about all of the details of the PixelMath expressions in the following YouTube video; however, note that I have covered the highlights in this article.

Conclusion

In this article, you learned about star reduction, and the effect it has on your overall image, you learned about Bill Blanshan’s PixelMath Expressions, learned how to acquire and use them, and learned which of the approaches is best for your images.