12 replies to this topic

[Adubs Astro]

Hello! 

 

I am under a month old in astrophotography and learning as I go. 

 

Gear:

• Sony a6400
• Sigma F2.8 24-70 mm Lens
• Kase Clip-In Night Filter (BH #KA1121040009)
• [pro]master alimunum tripod

Settings:

• 1,000 lights
• 1.6s Exposure
• IOS 800
• F2.8
• 70mm
• Bortle 6 Skies (I understand it's high)

Software:

• Siril

Steps used in Siril:

1. OSC_Preprocessing
2. Photometric Color Calibration using Orion Nebula

 With 100 bias, dark, and flat calibration frames, I've been able to produce the following image. 

 

 

After I attempt background extraction, I get this. 

 

 

So, a couple of questions. 

 

1. Is all that haze and or gradient that I am seeing light pollution? 
2. Am I on the right track?
3. Is there a way to remove that haze without disturbing real data? 
4. If I had a Sky Watcher Star Adventurer GTI or better, would longer exposures only intensify this gradient?  
5. What can I do to improve upon this?
6. Should I have left the lens at 24mm instead of 70mm, and would it have given better results?

Please feel free to add or point me in the right direction. :-)

 

Looking forward!

Edited by Adubs Astro, 07 December 2024 - 07:29 AM.

[Infomastr]

The gradient appears due to lack of calibration frames.  You need calibration frames (bias, flat, dark) to correct your light frames before stacking.  Since you are using Siril you should check out this and this (particularly the portion explaining how flat frames correct light frames). 

 

[Tapio]

To get better:

- get a tracker mount so you can take longer and less frames - helps with processing time and in general.

- learn to take calibration frames, flats and bias at least, darks are sometimes optional

 

Light pollution filters are becoming thing of the past because of transition of led lighting.

[Adubs Astro]

The gradient appears due to lack of calibration frames.  You need calibration frames (bias, flat, dark) to correct your light frames before stacking.  Since you are using Siril you should check out this and this (particularly the portion explaining how flat frames correct light frames). 

I have 100 bases, darks and flats. I apologize, ill edit my post so this is known. 

 

I realize I did most of the basic steps, but it looks like more are provided at the links you provided. I will read them over. Thank you!

Edited by Adubs Astro, 07 December 2024 - 07:43 AM.

[Adubs Astro]

To get better:

- get a tracker mount so you can take longer and less frames - helps with processing time and in general.

- learn to take calibration frames, flats and bias at least, darks are sometimes optional

 

Light pollution filters are becoming thing of the past because of transition of led lighting.

Updated post to reflect calibration frames where taken and used. 

 

Tracker is on my Christmas wishlist. 

 

I wondered about the transition to LED lighting and how that affects things. Thank you for this thought! :-) 

Edited by Adubs Astro, 07 December 2024 - 07:39 AM.

[Adubs Astro]

As an update, I'm willing to upload my project folder to Dropbox if someone is willing to look at the files. :-) 

[Infomastr]

I have 100 bases, darks and flats. I apologize, ill edit my post so this is known. 

 

I realize I did most of the basic steps, but it looks like more are provided at the links you provided. I will read them over. Thank you!

How did you shoot flats? 

 

Another list of things to look at would be your camera settings.  You want RAW images, manual focusing, no noise reduction, etc.  A number of things that modern cameras do automatically to improve terrestrial imagery are detrimental to astrophotography.

 

If there is actual gradient in the images from light pollution (rather than issues from lack of or improper calibration), you will probably have more success in removing it by doing per-sub background extraction.  There is another Siril tutorial on how this works.

[Adubs Astro]

How did you shoot flats? 

 

Another list of things to look at would be your camera settings.  You want RAW images, manual focusing, no noise reduction, etc.  A number of things that modern cameras do automatically to improve terrestrial imagery are detrimental to astrophotography.

 

If there is actual gradient in the images from light pollution (rather than issues from lack of or improper calibration), you will probably have more success in removing it by doing per-sub background extraction.  There is another Siril tutorial on how this works.

** Edit: I just found that you can't edit a post after a certain amount of time has elapsed; that's unfortunate. **

 

I will again update my starting post to reflect these things. Thank you for pointing them out!

 

For flats, I used a doubled-up white t-shirt and a light panel over the lens, then used A mode so the camera chose the correct settings for exposure.

 

I used RAW output, manual focusing, and turned off all noise reduction or automated trickery. I found ONE source online for a6400 settings I used, but I made sure to go one by one and shut it off. If it had anything to do with image "clean up," I shut it off.

 

Thank you! I was thinking that manually processing these from start to finish may give better results; after all, the computer in our heads is fairly good if it has the right algorithms. :-)

Edited by Adubs Astro, 08 December 2024 - 07:46 PM.

[Adubs Astro]

I've been working with the data quite a bit using the information and guidance I've received so far. It seems like the Siril guide could use some TLC. By focusing on my main interest, I managed to produce the image below. I'm more than willing to share my data with someone more experienced to gain help and a better understanding if anyone is interested. On another note, I received a Sky-Watcher Star Adventurer GTi for Christmas!

 

[Ranger Tim]

Massive improvement. Long way to go though. Keep working on your exposure to prevent stars from “blooming.” Try it without the night filter. You don’t need more than about 30 calibration frames of each type. When you start tracking try a Lower ISO and longer subs. Watch for blurred stars to find the right length.

[vidrazor]

 With 100 bias, dark, and flat calibration frames, I've been able to produce the following image.

After I attempt background extraction, I get this.

So, a couple of questions.

It seems like the Siril guide could use some TLC.

I'd like to add a few notes here:

 

A) Don't use that drop-in filter, it won't do much, and can actually be making things worse. Those filters are essentially useless on modern full spectrum light pollution, and all they typically do it mess up your color, and even luminosity.

 

B) In my experience using photographic lenses for astrophotography, some (but not all) lenses exhibit that circular pattern you see in your image, that I have outlined below. It's essentially a product of internal reflections in the lens, which don't normally show up in the terrestrial conditions, both day and night shooting, that the lenses were designed for, but they can show up when stacking and stretching. There's no easy way to eliminate them, but they can sometimes be eliminated using the AI process in GraXpert, along with other processes.

C) Have you downloaded and installed Starnet, and integrated it into Siril? Separating your target from the stars will allow you greater processing freedom to address issues like the gradients you're getting.

D) In addition to Starnet, install GraXpert, and also look into the Cosmic Clarity suite of programs. All of these programs are free, and merely cost you the time spent to learn ow to use them, which brings me to...

E) If you haven't run across them yet, check out Rich's set of tutorials on Siril (and other programs) on his YouTube page Deep Space Astro. He has one tutorial in particular for beginners that gets you through the fundamental phases of processing data in Siril, including the integration of Starnet into Siril. You should watch that video to help you get up to speed, then watch the other Siril videos he has to help you better understand the other tools in the program, especially generalized hyperbolic stretching which, combined with Starnet, will greatly improve the quality of what you can get from your data.

So try running your data through GraXpert's AI process, both with and without stars, and see if it will help you eliminate the issues you're experiencing. GraXpert now also does noise reduction and deconvolution (sharpening), but your immediate concern is effective gradient removal. There's no guarantee the AI process will eliminate it, but it may mitigate it, and running the RBF process (Siril's RBF gradient reduction code comes from GraXpert) may finish cleaning this up. It's a hit-n-miss affair, but worth a try.

Attached Thumbnails

• 

[Adubs Astro]

I'd like to add a few notes here:

 

A) Don't use that drop-in filter, it won't do much, and can actually be making things worse. Those filters are essentially useless on modern full spectrum light pollution, and all they typically do it mess up your color, and even luminosity.

 

B) In my experience using photographic lenses for astrophotography, some (but not all) lenses exhibit that circular pattern you see in your image, that I have outlined below. It's essentially a product of internal reflections in the lens, which don't normally show up in the terrestrial conditions, both day and night shooting, that the lenses were designed for, but they can show up when stacking and stretching. There's no easy way to eliminate them, but they can sometimes be eliminated using the AI process in GraXpert, along with other processes.

C) Have you downloaded and installed Starnet, and integrated it into Siril? Separating your target from the stars will allow you greater processing freedom to address issues like the gradients you're getting.

D) In addition to Starnet, install GraXpert, and also look into the Cosmic Clarity suite of programs. All of these programs are free, and merely cost you the time spent to learn ow to use them, which brings me to...

E) If you haven't run across them yet, check out Rich's set of tutorials on Siril (and other programs) on his YouTube page Deep Space Astro. He has one tutorial in particular for beginners that gets you through the fundamental phases of processing data in Siril, including the integration of Starnet into Siril. You should watch that video to help you get up to speed, then watch the other Siril videos he has to help you better understand the other tools in the program, especially generalized hyperbolic stretching which, combined with Starnet, will greatly improve the quality of what you can get from your data.

So try running your data through GraXpert's AI process, both with and without stars, and see if it will help you eliminate the issues you're experiencing. GraXpert now also does noise reduction and deconvolution (sharpening), but your immediate concern is effective gradient removal. There's no guarantee the AI process will eliminate it, but it may mitigate it, and running the RBF process (Siril's RBF gradient reduction code comes from GraXpert) may finish cleaning this up. It's a hit-n-miss affair, but worth a try.

Thanks for the reply!

 

A) Duly noted. You're not the first to mention this, so for my next run, I'll just remove it. #TheMoreYouKnow! :-)

 

B) Ah, I see. Good to know. I realize this is also a zoom lens and not a prime, so it's not the most ideal situation, but it's the best I had to get started with. I've slowly been looking for prime lenses to upgrade to, but it's slow going as I have the Sony A6400 and not your typical mirrorless camera that's commonly used. I'm not sure what to look for or avoid. GraXpert has been mentioned, which I'm unfamiliar with, so it's in my plans to install and watch some tutorials to start understanding that software.

 

C) Yes, I have downloaded and integrated StarNet into Siril. However, doing anything that removes or greatly cleans up that issue also affects the nebula that's showing. It does help, but not to the degree I wish it did.

 

D) See B for GraXpert. Cosmic Clarity is new to me; I'll look into it! Can't beat free... most of the time, lol.

 

E) I haven't come across his channel yet. I've bookmarked it now, though, and plan on watching some videos. Thanks! I'm fairly confident with Siril. At this point, it's just a matter of coming across the knowledge and then having the time to sit down, learn it, and use it—that's been the main issue. Luckily (or unluckily lol), I live in MI, and clear skies and good weather are rare this time of year.

 

Thank you for the plan of action! I definitely plan on going over all this. Another thing I've done—though I'm not sure if it's the right way to go—is crop the image down to what I actually want to see before I start processing everything.

[Meteorseeker]

I've been working with the data quite a bit using the information and guidance I've received so far. It seems like the Siril guide could use some TLC. By focusing on my main interest, I managed to produce the image below. I'm more than willing to share my data with someone more experienced to gain help and a better understanding if anyone is interested. On another note, I received a Sky-Watcher Star Adventurer GTi for Christmas!

 

result_1600s_PostStarNet-481kb.jpeg

For what my two cents are worth, I think you did a great job. You have three of the main targets there and the sky's the limit. Keep at it, listen to the suggestions and you'll be amazed how far you will come in a short time. Nicely done. 

 

Jaspn