(RCC) Why my capture is not enough good ?

Request for feedback
Werny MichaelBrian Puhl
•34 replies•1.9k views
Werny Michael avatar
Werny Michael avatar
hello everyone, last week I created an image that I call my best photo.

https://www.astrobin.com/uz9nkv/

This one only had two positive votes from the judges.
As it is an area that is rarely photographed and I spent 11.5 hours capturing it @F2 , I am a little disappointed.
Could you help me understand the weak points of this photo?

A big thank-you !

Clear Skies !
Respectful Engaging
Andre Vilhena avatar
Andre Vilhena avatar
Hi,

I think the jury is technically demanding and in your photo you can immediately spot small scale artifacts around stars. That's an absolute no go.

Cheers,
André
andrea tasselli avatar
andrea tasselli avatar
Purple sky? Too soft? Star size and colors?
Werny Michael avatar
Werny Michael avatar
Andre Vilhena:
Hi,

I think the jury is technically demanding and in your photo you can immediately spot small scale artifacts around stars. That's an absolute no go.

Cheers,
André

OK thanks
andrea tasselli:
Purple sky? Too soft? Star size and colors?

Thanks, it's SPCC color correction. I think, this color is real
andrea tasselli avatar
andrea tasselli avatar
Werny Michael:
Thanks, it's SPCC color correction. I think, this color is real


Pretty sure last time I've checked the sky there wasn't purple. SPCC or no SPCC. Anyhow to each their own...
Brian Puhl avatar
Brian Puhl avatar
The biggest technical flaw that stands out to me is the background gradient.   One corner of the image, it's blue, one is purple, and the bottom left hovers around green.    Dynamic Background Extraction is a pillar of proper data processing.  Gradients such as yours are usually caused by some sort of light pollution, and they only get exaggerated as your processing continues.   I would find some tutorials on how to improve your DBE.

The image itself is soft, lacking detail.   I don't gain anything by opening it full screen.   There's many reasons for this, I can only speculate.   That would be the number two reason why I wouldn't promote your image.

The stars… well, those are just RASA stars IMO.   They could use a little work, but they rarely are appealing.   In this case, I think they're the least of your issues.
Well Written Helpful Insightful
Die Launische Diva avatar
Die Launische Diva avatar
The background color inconsistency (and not accuracy) is something one can easily observe by having a look at most of your gallery thumbnails so that's what I think is the main issue of your image.
Werny Michael avatar
Werny Michael avatar
Ok I understand the strengths and weaknesses of the image better. I need to improve, I think the acquisitions are good (bortle 3-4).
I try to be as clean as possible so as not to overprocess. This is a great goal for 2024. A very big thank you to everyone for taking the time to respond to me !

:-)

Clear skies !
Respectful Supportive
Christian Koll avatar
Christian Koll avatar
Michael,

I would also agree to what has been said before; the gradient in the image ist possibly the biggest issue (but also very easy to cure: free Graxpert AI).
It also seems you have blurred the image a little bit to reduce noise, hence the rather soft impression?
But I would definitely not go so far as to call this a bad image, quite on the contrary.

With new tools like BlusXterminator or NoiseXterminator the benchmark for top astro-photographs went up even higher, so you'll need to be on the leading edge of software and processing.

Please also keep in mind, getting chosen for awards on Astrobin also involves some luck, there is no guarantee to be picked.

Keep up the good work!
Chris
Helpful Insightful Respectful Engaging Supportive
Werny Michael avatar
Werny Michael avatar
Thank you for this feedback, indeed I am going to test GraXpert on this photo to see if I can improve.

My goal is not specifically to be nominated, but receiving constructive criticism is very enriching.

Thanks again to everyone ! ( And sorry for my english)
Respectful Supportive
Brian Puhl avatar
Brian Puhl avatar
I can't see any reason given your situation to need something like Graxpert.   DBE will work just fine, you just need to hone your skills in.
SemiPro avatar
SemiPro avatar
I think everyone has covered the processing side, but there are special considerations you need to factor in with a RASA:
  • Tube currents are a thing. This is still 200mm of aperture and there is a lot of air sloshing around in there. Make sure you let your RASA acclimate properly to the outside environment or it will affect the sharpness of individual subs. If you are moving your RASA from indoors to outdoors, acclimation can take hours.
  • If you have not already properly routed your cables, then that is a must. You can use something like this. If that is not to your liking, then you can find a way to route them using a 3D printed spider vane to give your stars newtonian-like spikes.


Your sub-exposure lengths are also excruciatingly long for broadband at F/2. You are needlessly blowing out the stars while having them bleed into the surrounding pixels for very minimal gain.



Here are the sub-exposure lengths I use for the 2600MC depending on the light pollution and what focal ratio I am at. The "extra noise" is the total extra noise from stacking.

While I can respect not wanting to store and stack hundreds of frames, you should consider shortening your exposure time to at least 60 seconds.

As for the actual picture, you can compare yours with this one: https://www.astrobin.com/nmu3h4/B/?q=&camera=
Well Written Helpful Insightful Engaging
Werny Michael avatar
Werny Michael avatar
I think everyone has covered the processing side, but there are special considerations you need to factor in with a RASA:
  • Tube currents are a thing. This is still 200mm of aperture and there is a lot of air sloshing around in there. Make sure you let your RASA acclimate properly to the outside environment or it will affect the sharpness of individual subs. If you are moving your RASA from indoors to outdoors, acclimation can take hours.
  • If you have not already properly routed your cables, then that is a must. You can use something like this. If that is not to your liking, then you can find a way to route them using a 3D printed spider vane to give your stars newtonian-like spikes.


Your sub-exposure lengths are also excruciatingly long for broadband at F/2. You are needlessly blowing out the stars while having them bleed into the surrounding pixels for very minimal gain.



Here are the sub-exposure lengths I use for the 2600MC depending on the light pollution and what focal ratio I am at. The "extra noise" is the total extra noise from stacking.

While I can respect not wanting to store and stack hundreds of frames, you should consider shortening your exposure time to at least 60 seconds.

As for the actual picture, you can compare yours with this one: https://www.astrobin.com/nmu3h4/B/?q=&camera=

Very interesting !  Thanks
Werny Michael avatar
Werny Michael avatar
I tried to apply your advice. Do you find this better? 

https://www.astrobin.com/uz9nkv/B/

Thanks at all !
Shinpah avatar
Shinpah avatar
Your image is littered with pink/purple squares.

I've seen colored squares show up in images that have had starxterminator run on them without selecting "large tile overlap". I've explicitly seen these exact colored purple squares on one other person's image - they claim deepsnr generated the squares. These kind of tile artifacts seem to be the hallmark of neural net denoising and star removal processes.

You can also find this kind of color mottling when using MMT in pixinsight to denoise the chrominance of the image - although this tends to present as circles and ovals.
Well Written Helpful Insightful Concise
kurt ragsdale avatar
kurt ragsdale avatar
It is hard to compete with photos produced on here with most equipment one can afford on a certain budget. I mean your going up against photos made by observatory class instruments on here. The photo is quite good I wish I could get a photo that good but been at it less than 6 months and heavily light polluted skies unless I travel to the mountains around here. I noticed the stars on those Rasa telescopes aren't quite as good as on refractors and other telescopes. I noticed that Celestron is coming out with a 6 inch version of the Rasa that is a smart telescope with camera and all built in it the only drawback is I think other telescopes like premium refractors and planewave telescopes produce better background stars and look better overall if done right.
SemiPro avatar
SemiPro avatar
Werny Michael:
I tried to apply your advice. Do you find this better? 

https://www.astrobin.com/uz9nkv/B/

Thanks at all !

This is looking much better!
Brian Puhl avatar
Brian Puhl avatar
kurt ragsdale:
It is hard to compete with photos produced on here with most equipment one can afford on a certain budget. I mean your going up against photos made by observatory class instruments on here.



I'm not really agreeing with this at all.  His equipment is perfectly capable of being competitive.   He just needs to work on his processing a little.
Werny Michael avatar
Werny Michael avatar
Your image is littered with pink/purple squares.

I've seen colored squares show up in images that have had starxterminator run on them without selecting "large tile overlap". I've explicitly seen these exact colored purple squares on one other person's image - they claim deepsnr generated the squares. These kind of tile artifacts seem to be the hallmark of neural net denoising and star removal processes.

You can also find this kind of color mottling when using MMT in pixinsight to denoise the chrominance of the image - although this tends to present as circles and ovals.

It's true, purple squares on my picture :/
kurt ragsdale:
It is hard to compete with photos produced on here with most equipment one can afford on a certain budget. I mean your going up against photos made by observatory class instruments on here. The photo is quite good I wish I could get a photo that good but been at it less than 6 months and heavily light polluted skies unless I travel to the mountains around here. I noticed the stars on those Rasa telescopes aren't quite as good as on refractors and other telescopes. I noticed that Celestron is coming out with a 6 inch version of the Rasa that is a smart telescope with camera and all built in it the only drawback is I think other telescopes like premium refractors and planewave telescopes produce better background stars and look better overall if done right.

Thank you for your encouragement, but I can make better use of my captures. Not for a picture of the day, but for better results.
Brian Puhl:
kurt ragsdale:
It is hard to compete with photos produced on here with most equipment one can afford on a certain budget. I mean your going up against photos made by observatory class instruments on here.



I'm not really agreeing with this at all.  His equipment is perfectly capable of being competitive.   He just needs to work on his processing a little.

I agree with you, I need to improve! Thanks



I think my unit captures are too long and "eat" the color.
Shinpah avatar
Shinpah avatar
I don't think it's likely that too long subexposures is the root of your problem. This can only cause issues with the bright end of images and you're seeing the problem in the darker areas.

I looked back at a few of your other images and I only see the same issue in the witch head image from 6 days ago. I think it's likely this is either a specific processing step you're taking (like the denoising/star removal I referred to) that you started recently.

Potentially I could see this issue also happening if you're exporting to a lossy format and using too high compression and this issue only appears after exporting the image. I don't think this is likely though.
Helpful Insightful Respectful Concise Supportive
Brian Puhl avatar
Brian Puhl avatar
I wonder, are you using starless processing techniques?
Well Written Respectful
Dave Rust avatar
Dave Rust avatar
Is winning judges approval a priority? With thousands of images being submitted, I don’t ever expect to be picked. Better to shoot for your own gratification. 

I have been tagged as a contender once in 3 years, but did not win. 

I think your image looks terrific for 11 hours. Any images of that target that are more detailed or colorful are typically double the number of subs. I rarely have the patience to stay on a single target that long! That being said, I am slowly moving in that direction. 

Keep on doing what you’re doing.
Helpful Insightful Respectful Engaging Supportive
Werny Michael avatar
Werny Michael avatar
I don't think it's likely that too long subexposures is the root of your problem. This can only cause issues with the bright end of images and you're seeing the problem in the darker areas.

I looked back at a few of your other images and I only see the same issue in the witch head image from 6 days ago. I think it's likely this is either a specific processing step you're taking (like the denoising/star removal I referred to) that you started recently.

Potentially I could see this issue also happening if you're exporting to a lossy format and using too high compression and this issue only appears after exporting the image. I don't think this is likely though.

Ok thanks, I'm going to look at this side
Brian Puhl:
I wonder, are you using starless processing techniques?

Yes, for star reduction.
Dave Rust:
Is winning judges approval a priority? With thousands of images being submitted, I don’t ever expect to be picked. Better to shoot for your own gratification. 

I have been tagged as a contender once in 3 years, but did not win. 

I think your image looks terrific for 11 hours. Any images of that target that are more detailed or colorful are typically double the number of subs. I rarely have the patience to stay on a single target that long! That being said, I am slowly moving in that direction. 

Keep on doing what you’re doing.

Thank you for your message, no I am not looking for approval from the judges... I am trying to improve my photos for this year of 2024. Your photos are very beautiful! Thanks
Pedro A. Sampaio avatar
Pedro A. Sampaio avatar
If I may chime in….

How much did you adjust with curves? Specifically, how much increase in saturation did you do? Specially for those kind of targets, we want to "bring the colors out" by increasing the saturation, sometimes by a lot. But by increasing saturation, you also increase the difference in the colors in each part of the image. If your background is not properly neutral (and your image properly color corrected), any small difference in color will become by the end of your process. I know I've struggled quited a few times with this issue - and it's sometimes very frustrating because you only notice at the end. Some very meticulous curves adjustments can sometimes alleviate the issue,  but the ideal is to not have it at alll.

In regards to star removal techniques, that's an interesting topic as well. After the arrival os BlurX and GHS, I've found myself using less and less star removal (be it Starnet ou StarX). I think that star removal more often than not just introduces more artifacts than it helps overall. And with BlurX I can do proper deconvolution and star reduction at the linear state with minimal to no artifacts, and with GHS and proper curve adjustments, I can stretch the image while protecting the highlights (i.e. stars are not blown out).

In the end, I find that it unless I'm aiming for a starless image (which I never did until today), the most I use StarX during the linear process is to "see" what kind of data I have before stretching - i.e. STF tends to blow up stars, and sometimes having a separate image without stars is useful to see what is nebulosity, etc. I still just work on the regular image though.
Helpful Insightful Respectful Engaging
Die Launische Diva avatar
Die Launische Diva avatar
Pedro A. Sampaio:
But by increasing saturation, you also increase the difference in the colors in each part of the image. If your background is not properly neutral (and your image properly color corrected), any small difference in color will become by the end of your process. I know I've struggled quited a few times with this issue - and it's sometimes very frustrating because you only notice at the end.

^^^This!

And that specific part of the sky is extremely complex thus requiring extra care with gradient modelling:

https://astrob.in/h053kp/B/

https://astrob.in/fiyi3z/0/

And a proper gradient modelling always starts with a very good flat correction...

It is not just a matter of expensive equipment.
Related discussions
RCC NGC2244 - Request for Critique
Hello - I'm new to astrophotography, and I'd appreciate any critique you'd like to give on this image of NGC2244. Thanks in advance for any help you can provide!
Beginner seeking critique on astrophotography images; similar request for feedback.
Jan 23, 2022