Embedded Files
Sunday, Sept. 22, 2024
I guess I haven't exactly posted in a while.
I had a good streak of about one post per month, but life happened and I became busy. After a long hiatus from photography to pursue astrophysics research (more on that coming soon), I decided to get back into the swing of things before band season started ramping up and I inevitably contracted senioritis.
I set up my scope and had arguably one of my smoothest imaging sessions ever! I balanced the scope on my first try, the alignment was near-perfect, the wind was at a minimum, and everything WORKED! Feeling proud of myself, I returned indoors a few hours later and processed my pictures in Siril using my standard workflow. And here, everything went smoothly. I didn't have any weird software troubles.
Everything was going. . . oddly well.
Omega Nebula (M17) - Aug. 24
My First Attempt
My resulting image was, arguably, my best one yet. There were so many stars! The nebula was bright and in high contrast! The sky didn't look like pixellated garbage!
As you can tell from my glut of exclamation marks, I was elated.
But there was a small nagging in the back of my mind, an itch I couldn't scratch. Something was wrong. A fatal flaw with my imaging that I couldn't see, something so horrible that everything would come crashing down.
Then I decided "Oh whatever" and decided to go to sleep since I had school the next morning anyway.
But the thought never left.
Not long after, I had managed to one-up myself again, with a picture of the Eagle Nebula in an ad-hoc salvage mission after the first test for Piggyback Mount v2 went rather poorly (more on this coming soon too). I only got about 30 minutes of total integration time, so I wasn't particularly optimistic when I started to process it, but lo and behold, I could see the Pillars of Creation.
To continue my streak of good luck, I had even managed to screw up the flat frames, but in such a way that the only wrong pixels were in areas I could crop.
I did my process again.
Got an image again.
Really liked it again.
But still, something was off. By this point, I had also been looking at my Omega Nebula picture several times throughout the day, playing around randomly with image settings in Apple Photos during my free time. I knew the stars weren't perfectly round, nor was the nebula particularly defined near the edges, but surely there was something else. On a whim, I decided to search for an astrophoto from someone far more experienced than I was.
And that's when I saw it. The stars.
Eagle Nebula (M16) Sept. 9
The stars were, what I eventually learned was called, bloated. In the original data, they seem fine, like little pinpoints in the sky surrounded by the noisy background of Bortle 6 raw image files. In the final picture, they were big. Too big.
Apparently, as I was processing my image to bring out the nebulosity in M16 and M17, the data from the stars were also being affected. Each star has a small bit of light around it in a halo depending on how bright they are. The halo is hidden most of the time by how much data a picture can really show at once, which is why processing exists in the first place. As I stretched the light curves, the light became more and more obvious, blending with the edges of the star until I got bright blobs that were annoying to see and limited how much of the nebula I could really show.
I already knew of a way to fix it, luckily. Starnet software.
Now, I would say there was a struggle, some obstacle I had to overcome, but this seemed to go well, too.
The software was free and available for download on the web. It was made by an amateur astronomer for amateur astronomers. The documentation was clear, precise, and actually helpful (*cough* CELESTRON *cough*). After a few minor hiccups with getting the download to show up in Siril, I implemented it immediately, following YouTube tutorials from Deep Space Astro and Nebula Photos.
It was, in my opinion, simple to learn and implement. All I needed to do was click a few menu options to add Starnet to Siril and use a few commands to create what's called a "star mask" to separate my background image and my stars. From there, I could process each one individually before adding them back together.
Random screenshot for this blog specifically because I want to keep things interesting and not just a wall of words.
I did it with the Eagle Nebula first, since that was, in my opinion, more in need of "fixing". Here's a before and after:
Before
After
I did the same thing with the Omega Nebula about a week later to similar improvements:
Before
After
Voilà! Case closed! Done! Good! I have learned it!
Right?
Well, no. There's a good reason I say "kinda" and not "completely". While there's certainly a marked improvement in the quality of the images, there are a few glaring issues that appear if you look closer.
First, let's start off with the stars themselves. Starnet improved them significantly, sure, but I may have gone a bit overboard with the star sizes, taking away too many in favor of the nebula at the cost of making the sky look empty. Even individual stars, especially the ones that were already saturated from the raw data, started looking less than stellar (I'll see myself out after that pun). For example, look at the star in the top right. It's downright disgusting
Eeeeeeeeewwwwwwwwww
All those grey splotches aren't supposed to be there. Those are artifacts from star removal. At least, that's what I think.
The pillars were a lot more grainy in the Starnet version
Starnet also has the rather annoying habit of creating some artifacts from where the stars disappear. I struggled to get rid of them, but couldn't find any easy way to do so with the software I had. Upon further research, I learned that there's a tool in Photoshop that can fix this easily.
I don't have photoshop. I own a Mac.
Overall, using Starnet also makes noise in the background a lot more obvious since it makes stretching the light curves favorable. Any imperfections in the data, either from the original pictures or added from processing, become a lot easier to accidentally over-stretch.
Still, I'm pretty happy with these results, and Starnet is now a permanent part of my workflow. It's just another step along the path to getting even better pictures in the future, and I'm sure these issues will go away as I learn better processing techniques. Until then, keep an eye out for more pictures using Starnet!
I promise I'll post more frequently in the future.
Page updated
Google Sites
Report abuse