How to Help with Bloated Stars

What can I do to help with stars becoming too large? See my latest (Pac-Man).

Is this due to too long of exposures? Processing technique? Atmospheric issues?

Any insight would be appreciated.

Thanks.

Too long exposures certainly contributes to them being overly large as you saturate more and more pixels (and your FWC+ is small to start with). I shall also add that using ED achromats,and fast ones to boot, is a recipe for bloating stars. You can rule out seeing as the 71mm is just too small to suffer much bout it. Star reduction technique could be employed as well as RGB substitution. I'd suggest you moderate your ambitions in terms of image scale by a factor of 2 as this will help with both the bloating and the perceived contrast of the image.

shorter exposures will help. Have a similar setup as yours just different flattener. You can have look at my version. In this one i had backfocus issues and pinched optics still to work out but starbloat is not nearly as strong as yours https://www.astrobin.com/jk8cct/

A lot of things can make stars bloat. Focus, dew, seeing, clouds, guiding, quality of lenses, quality of filters, collimation, you name it. I would suggest first to look if it's a one-time issue or consistent in your images. If it only happens once in a while, there is nothing to worry about. It happens to all of us. Seeing is the most common issue, and sometimes there are high clouds too thin to tell in the dark but affect the image. If you shoot in a high-humidity environment, a dew-removing device is a must. If you consistently get big stars, I would suggest starting by examining the imaging train. The quality of the telescope lens is unfortunately an important factor as well. After all, there is a reason why top-tier refractors are so expensive.

Exposure is not so much a factor, in my opinion. You can get over-saturated stars by overexposing a frame but it doesn't necessarily make stars a lot bigger, at least not to a very noticeable extent. But that may also depend on your camera, I suppose.

And finally, there are ways to make stars smaller in post-processing. Many solutions are available in various software. But the most effective way is the new Bill Blanshan's method in PixInsight. It really works like magic! If you use PixInsight, you can watch this video to learn more: https://www.youtube.com/watch?v=rM3-yAcAbZc&t=0s

I would try 300" subs.

I would suggest removing the stars from your integrated stacks and then use a separate version of the same stacks stretched just enough to bring out the stars without bloating and then layer those on top of your starless processed image. You could also take a separate stack of shorter images taken using RGB rather than narrowband filters, e.g. 30 x 30s in each of R,G and B.

For me it's seeing. At times the seeing induced bloat is so bad I can't even platesolve. But when processing, I do exactly as Roger posted above so far as a separate stretch for the stars. I've tried shooting for stars separately as well, but I think my exposures were too short and at my fairly dark site I didn't swamp the read noise. Anyhow, the lesser stretch for stars technique is essentially what's used in the video linked to above. The one difference is that the tools in the video reduce the stretch in a non-linear state.

Cheers,
Scott

I think the crop played a big factor in this image especially but I also think there has been a bit of haze in the skies of some sort that may be playing a factor. I’ll try some variables.

A lot of it can be dealt with in processing. For nebulae like the pacman I would extract the stars from a low stretch and then stretch the starless inage up via (e.g PixInsight) Hyperbolic and local histograme equalisation etc and then add the (relatively unstretched) stars back at the end.

Tim

Tim Hawkes:
A lot of it can be dealt with in processing. For nebulae like the pacman I would extract the stars from a low stretch and then stretch the starless inage up via (e.g PixInsight) Hyperbolic and local histograme equalisation etc and then add the (relatively unstretched) stars back at the end.

Tim

I’m intrigued by this process but use APP currently. Conceptually. I get it but how do I create a starless version?

In my opinion 600s is too much for the 183's well capacity. As the pixels fill up with photons, they begin to spill over onto neighboring pixels, resulting in bloated, blown out stars. I'd try shorter exposures, 180 or maybe 300s.

Could also be your guiding. What kind of numbers are you getting? The fine details are fuzzy in the packman suggesting a lot of noise reduction was used. Bad guiding can give you poor signal to noise ratios after stacking. Shorter sub exposures can also compensate somewhat for poor guiding.

You can only reduce stars so much with processing before it becomes noticeable. Check out StarXterminator photoshop plugin.

Tim Hawkes:
A lot of it can be dealt with in processing. For nebulae like the pacman I would extract the stars from a low stretch and then stretch the starless inage up via (e.g PixInsight) Hyperbolic and local histograme equalisation etc and then add the (relatively unstretched) stars back at the end.

Tim

I’m intrigued by this process but use APP currently. Conceptually. I get it but how do I create a starless version?

StarXTerminator or Starnet does this.

As part of my normal processing workflow, I remove stars while the image(s) are still linear then stretch the nebulosity and the stars separately. I find if I do it while the images are linear there are fewer artifacts and I have more control of color and contrast when stretching each separately. At the nonlinear stage, I process each as desired and reintegrate nebulosity and stars at the end of the processing workflow.

Most likely the reason is what Blaine Gibby said in previous post - 183MM having low well depth.
Try using the subexposure tables for 183MM for your sky quality, f ratio and narrowband filters here:
https://www.cloudynights.com/topic/616524-sub-exposure-tables-for-the-zwo-asi183mm-and-qhy183m-and-colour-versions/
For 183MM, I would not go any longer than calculated using the subexp tables due to the well depth... IMX183 is not 294C or IMX571

I had similar issues with Borg 71 f/3.7 + 183MM combo: https://www.astronom.si/galerija/index.php?n=5768
I think this is more pronounced with APOs that have significant CA.

Peter

A lot of good advice here. For me I'd advise two things:

1. Shorter exposures. You probably only need 2 or 3 minutes per subexposure with your rig.

2. Install and learn how to use Starnet++ with Photoshop. Separating the stars from the nebula, processing them independently, and then recombining near completion, will allow you much greater control. You can stretch the nebula data more, and avoid overstretching the star data. StarXterminator works even better in my opinion, but its not free like Starnet++.

A lot of good advice here. For me I'd advise two things:

1. Shorter exposures. You probably only need 2 or 3 minutes per subexposure with your rig.

2. Install and learn how to use Starnet++ with Photoshop. Separating the stars from the nebula, processing them independently, and then recombining near completion, will allow you much greater control. You can stretch the nebula data more, and avoid overstretching the star data. StarXterminator works even better in my opinion, but its not free like Starnet++.

Completely agree. Just to add one point for clarification it can also be worth having two sets of exposures. A shorter set to extract star images from (under any skies) and a set of longer exposures (preferably under darker conditions) to get the faint nubula details making for a higher dynamic range overall.

Blaine Gibby:
In my opinion 600s is too much for the 183's well capacity. As the pixels fill up with photons, they begin to spill over onto neighboring pixels, resulting in bloated, blown out stars. I'd try shorter exposures, 180 or maybe 300s.

Could also be your guiding. What kind of numbers are you getting? The fine details are fuzzy in the packman suggesting a lot of noise reduction was used. Bad guiding can give you poor signal to noise ratios after stacking. Shorter sub exposures can also compensate somewhat for poor guiding.

You can only reduce stars so much with processing before it becomes noticeable. Check out StarXterminator photoshop plugin.

I am getting about 3-5 on guiding. Should I try focusing in on my guide scope to hone this in or is that just the HEQ5 limitation?

My pacman was also heavily cropped.

2. Install and learn how to use Starnet++ with Photoshop. Separating the stars from the nebula, processing them independently, and then recombining near completion, will allow you much greater control. You can stretch the nebula data more, and avoid overstretching the star data. StarXterminator works even better in my opinion, but its not free like Starnet++.

One trick here, use starnet2 or starXterminator to remove the stars, but use this PixelMath formula to create a stars only image: ~(~full_image/~starless). And then ~(~Stars*~Starless) to combine stars and starless. These are also in the linked to video above.

Cheers,
Scott

Tim Hawkes:
A lot of good advice here. For me I'd advise two things:

1. Shorter exposures. You probably only need 2 or 3 minutes per subexposure with your rig.

2. Install and learn how to use Starnet++ with Photoshop. Separating the stars from the nebula, processing them independently, and then recombining near completion, will allow you much greater control. You can stretch the nebula data more, and avoid overstretching the star data. StarXterminator works even better in my opinion, but its not free like Starnet++.

Completely agree. Just to add one point for clarification it can also be worth having two sets of exposures. A shorter set to extract star images from (under any skies) and a set of longer exposures (preferably under darker conditions) to get the faint nubula details making for a higher dynamic range overall.

Sounds wise. I’m at a point of trying to find a medium between time and quality so want to simplify the process but not at the sacrifice of the results. Tough; indeed. Onward and upward always though.

=16pxI am getting about 3-5 on guiding. Should I try focusing in on my guide scope to hone this in or is that just the HEQ5 limitation?
My pacman was also heavily cropped.

When you say 3-5 on guiding, is that your RMS error? If so that’s a little high. Less than 1 is kind of the unofficial standard but it depends on your setup, focal length, pixel size, etc but the lower the better.

Yungshih Lee:
A lot of things can make stars bloat. Focus, dew, seeing, clouds, guiding, quality of lenses, quality of filters, collimation, you name it. I would suggest first to look if it's a one-time issue or consistent in your images. If it only happens once in a while, there is nothing to worry about. It happens to all of us. Seeing is the most common issue, and sometimes there are high clouds too thin to tell in the dark but affect the image. If you shoot in a high-humidity environment, a dew-removing device is a must. If you consistently get big stars, I would suggest starting by examining the imaging train. The quality of the telescope lens is unfortunately an important factor as well. After all, there is a reason why top-tier refractors are so expensive.

Exposure is not so much a factor, in my opinion. You can get over-saturated stars by overexposing a frame but it doesn't necessarily make stars a lot bigger, at least not to a very noticeable extent. But that may also depend on your camera, I suppose.

And finally, there are ways to make stars smaller in post-processing. Many solutions are available in various software. But the most effective way is the new Bill Blanshan's method in PixInsight. It really works like magic! If you use PixInsight, you can watch this video to learn more: https://www.youtube.com/watch?v=rM3-yAcAbZc&t=0s

Yungshih Lee, thanks for bringing this set of star reduction techniques to our attention, very useful, I also liked the additional explanations that Bill gave on his Pixelmath expressions.

Blaine Gibby:
In my opinion 600s is too much for the 183's well capacity. As the pixels fill up with photons, they begin to spill over onto neighboring pixels, resulting in bloated, blown out stars. I'd try shorter exposures, 180 or maybe 300s.

Could also be your guiding. What kind of numbers are you getting? The fine details are fuzzy in the packman suggesting a lot of noise reduction was used. Bad guiding can give you poor signal to noise ratios after stacking. Shorter sub exposures can also compensate somewhat for poor guiding.

You can only reduce stars so much with processing before it becomes noticeable. Check out StarXterminator photoshop plugin.

I am getting about 3-5 on guiding. Should I try focusing in on my guide scope to hone this in or is that just the HEQ5 limitation?

My pacman was also heavily cropped.

*The HEQ5 Pro can support guiding down to about 0.5 arc secs RMS in my experience (with the right seeing conditions and guide set-up), so I would look elsewhere.

Blaine Gibby:
=16pxI am getting about 3-5 on guiding. Should I try focusing in on my guide scope to hone this in or is that just the HEQ5 limitation?
My pacman was also heavily cropped.

When you say 3-5 on guiding, is that your RMS error? If so that’s a little high. Less than 1 is kind of the unofficial standard but it depends on your setup, focal length, pixel size, etc but the lower the better.

Yes. What can I do to hone this in?

Andy Wray:
Blaine Gibby:
In my opinion 600s is too much for the 183's well capacity. As the pixels fill up with photons, they begin to spill over onto neighboring pixels, resulting in bloated, blown out stars. I'd try shorter exposures, 180 or maybe 300s.

Could also be your guiding. What kind of numbers are you getting? The fine details are fuzzy in the packman suggesting a lot of noise reduction was used. Bad guiding can give you poor signal to noise ratios after stacking. Shorter sub exposures can also compensate somewhat for poor guiding.

You can only reduce stars so much with processing before it becomes noticeable. Check out StarXterminator photoshop plugin.

I am getting about 3-5 on guiding. Should I try focusing in on my guide scope to hone this in or is that just the HEQ5 limitation?

My pacman was also heavily cropped.

*The HEQ5 Pro can support guiding down to about 0.5 arc secs RMS in my experience (with the right seeing conditions and guide set-up), so I would look elsewhere.

What are some ways to reduce guiding error?

Question, did you by chance process in Pixinsight then another program? I had this happen to me when I had begun in Pixinsight and had separated the stars and then brought both the starless star mask into Photoshop but had accidentally saved the star mask as 32 bit. Also, if you stretched the Oiii channel heavily with the stars still in, it can do this as well. (Another personal experience!)

Jared Holloway:
Question, did you by chance process in Pixinsight then another program? I had this happen to me when I had begun in Pixinsight and had separated the stars and then brought both the starless star mask into Photoshop but had accidentally saved the star mask as 32 bit. Also, if you stretched the Oiii channel heavily with the stars still in, it can do this as well. (Another personal experience!)

I use APP then photoshop.

I will see what I can adjust in my acquisition and processing to improve.

Related discussions