How to determine the quality of a picture we have processed? - Forums

Topic starter

Jan 20, 2022

Hi guys! I had some problems with processing techniques:
(1). How to judge a picture which after-processing?
(2). The processing techniques have not been updated for a long time, how can I learn more?
(3). I don't know the advantages and disadvantages of the Deep Sky picture after I process it, how can I know it?
Thanks for the response!

Steven

Jan 20, 2022

1)
do you like the picture? yes? no? does it matter? the only thing that matters is wether or not you are happy with it.
You can always start from scratch 1-2 days later, re-edit it. And then put picture 1 and 2 next to each other and decide which you like best. That is probably the best way to determine it.

One thing you could do is use the search function on astrobin, look for your target. See what kind of equipment people are using.. Find images of people that used similar equipement, see what filters they've been using, how much integration time they had, what programs they used. etc etc. And be realistic. If someone is using 100.000 dollars worth of equipement and integrated 100 hours into a final image, don't expect to get the same if you have other equipement or only a few hours of data.

2)
YouTube is always a great way to learn more and new things. I often process images from scratch, watching YouTubers who go through the entire process with a full length photoshop or pixinsight video. And I just follow along to see what I end up with, and what things work for me, what thing don't. Maybe it turns out the way I like it, maybe it doesn't. But I usually do learn some new techniques.

3)
see 1 and 2.. reprocess it, follow along with a YouTube tutorial, see what works, see what doesn't work.
You could always use the "Request of constructive critique" part on the forum here. Post your images there to get some feedback.
If you do, it usually helps to say whatever it is you're not happy about, or what you'd like to improve.

What others sometimes do, is they make the RAW stacked data available to download, and let other people have a crack at editing the data, usually giving very different and interesting results. - for example: this post

Helpful Supportive

Topic starter

Jan 20, 2022

1)
do you like the picture? yes? no? does it matter? the only thing that matters is wether or not you are happy with it.
You can always start from scratch 1-2 days later, re-edit it. And then put picture 1 and 2 next to each other and decide which you like best. That is probably the best way to determine it.

One thing you could do is use the search function on astrobin, look for your target. See what kind of equipment people are using.. Find images of people that used similar equipement, see what filters they've been using, how much integration time they had, what programs they used. etc etc. And be realistic. If someone is using 100.000 dollars worth of equipement and integrated 100 hours into a final image, don't expect to get the same if you have other equipement or only a few hours of data.

2)
YouTube is always a great way to learn more and new things. I often process images from scratch, watching YouTubers who go through the entire process with a full length photoshop or pixinsight video. And I just follow along to see what I end up with, and what things work for me, what thing don't. Maybe it turns out the way I like it, maybe it doesn't. But I usually do learn some new techniques.

3)
see 1 and 2.. reprocess it, follow along with a YouTube tutorial, see what works, see what doesn't work.
You could always use the "Request of constructive critique" part on the forum here. Post your images there to get some feedback.
If you do, it usually helps to say whatever it is you're not happy about, or what you'd like to improve.

What others sometimes do, is they make the RAW stacked data available to download, and let other people have a crack at editing the data, usually giving very different and interesting results. - for example: this post

  Thank you very much, Steven!
I wasn't in the habit of looking at Astrobin before processing images, but I will search in there at the next time! “And be realistic. If someone is using 100.000 dollars worth of equipement and integrated 100 hours into a final image, don't expect to get the same if you have other equipement or only a few hours of data.“I think the same as you! I think I should use all the capabilities of the equipment I have now, and practise my processing techniques. When I cannot take better photos due to equipment, I will change the equipment(if i can afford it)...
  At the same time, thank you for providing a way for me to learn processing techniques.

Wish you can take the picture that makes you satisfied! Thank you for your advice.

Respectful Supportive

Sean van Drogen

Jan 20, 2022

On point 1 I 100% agree with Steven, most important part is if you are happy with your picture, and indeed be realistic what you can can get with your gear and time.

for point 2 I dont do well with youtube I prefer the various forums and books. I only use PI so the main books for this is what I have.
but if you have a real good data set come back to it after you learn something new and reprocesses to see if you like the result better.

CCDMike

Jan 20, 2022

Sean van Drogen:
On point 1 I 100% agree with Steven, most important part is if you are happy with your picture, and indeed be realistic what you can can get with your gear and time.

for point 2 I dont do well with youtube I prefer the various forums and books. I only use PI so the main books for this is what I have.
but if you have a real good data set come back to it after you learn something new and reprocesses to see if you like the result better.

Hi Sean,
which books do you recommend most for an PI beginner?

CS
Mike

CCDMike

Jan 20, 2022 · Edited: Jan 20, 2022

I think there is no "standard" out there what is well done and what is crap. It is all about personal feeling.
Some prefer a black background where others (as me) think that it looks unnatural. Talking about colours is another topic. Especially in NB you can do wierd things that look very good.
My personall view is, no matter if it's a nebula or a galaxy, the star colour should be well balanced. It's boring if you just have white stars as well as crazy red, pink and dark blue ones.

Find your own way and always ask some experienced people here on AB if you need a honest opinion

Stay safe!
Mike

andrea tasselli

Frequent contributor

Jan 20, 2022

which books do you recommend most for an PI beginner?

Not Sean but I'd go for "Inside Pixinsight" 2nd edition

Die Launische Diva

Jan 20, 2022 · Edited: Jan 20, 2022

The IOTD process Guidelines offer some good advice regarding the qualities a good astrophoto should have.

Regarding (1), I advise to step back from the image you process and come back after a while in order to assess it. The longer you stay away, the better. This breaks the sentimental connection between you and your image and will let you to be a more objective judge of your own work. Detachment from your effort is beneficial since you will focus on judging the result and not in celebrating on your effort or punishing yourself for the small mistakes you have done (and probably you are the only one noticing them). You want to avoid thoughts like these: "$10000 equipment, 100 hours of exposure under sub-zero temperatures, 100 hours spent on processing using the latest 32-core desktop computer... It MUST be a good image" and start thinking like this: " Does my image respect the object depicted and the data available? Does my image fulfill my intentions? Are my intentions effortlessly communicated to my audience?"

Regarding (2) I don't have a good tip, but I think it is better for anyone to first have settle on a very simple processing workflow which produces decent results, and then refine it, if it is necessary to do so. Having a solid knowledge of the available software tools helps a lot. If you have questions on the tools you are using, better ask in the proper places like the dedicated fora of the software you are using, be it PixInsight, APP, or Siril.

Lastly, for (3) you may assess your image against the IOTD Guidelines I mentioned earlier, even if you don't care so much about the IOTD.

Regarding the ever-lasting debate on the positive correlation between equipment price vs. quality, I tend to have a different opinion: Expensive equipment buy us repeatability and convenience, not processing skills and a good taste. Money does not directly result in better images. An expensive Leica camera won't turn me into a Henri Cartier-Bresson, and if you drop me in the Grand Canyon with a large-format camera you won't turn me into a modern Ansel Adams

I hope my advice is helpful and reasonable. And after browsing your image gallery, I have to say that your work seems to be already very good!

Helpful Insightful Respectful Engaging

Sean van Drogen

Jan 20, 2022

andrea tasselli:
which books do you recommend most for an PI beginner?

Not Sean but I'd go for "Inside Pixinsight" 2nd edition

mastering pixinsight as well

Jeffrey K Lovelace

Jan 20, 2022

Great topic! I think you're way ahead on the following issues, but it's worth adding a few basics for others new to this practice.

First, an often overlooked element of good editing is proper monitor calibration. I see a lot of images with an obvious color bias, or which are generally too bright with blown out highlights or too dim; and my guess is that their monitor isn't calibrated. I use datacolor's SpyderX products. It's hard to tell if your image is any good if your monitor is way our of calibration.

Here's another basic, general guideline that I often see violated: Highlights should never be so bright that you lose detail, and the dark regions should never be completely black. Also you know that you have pushed color saturation way too far when you have lost detail due to over saturation. If you can still see fine detail in your highlights, darker areas, and most color saturated areas, you are good.

Another useful technique is to compare your image to others of the same target that have earned top pick or IOTD. Only reference awarded images from the past three years as everything has advanced quite far recently. Don't try to match/copy others, but if you like your approach side by side with theirs, then you are good.

Finally, look at detail in your image down to the pixel level. Late in your editing process scan through your entire image at 400%+ magnification. look for artifacts from star removal or over sharpening, or not enough/too much noise reduction, etc. Mostly look at your stars: are they crunchy/chunky, too fuzzy, oblong/misshapen; surrounded by dark rings or color blotches, etc.? Then research or invent ways to fix these issues. If you think that others won't look at your image on that level, you would be way wrong. The IOTD staff's job is to look at images at this level, for instance. If you like your image at 5x, then you are good.

I could go on, but as I have already achieved TLDNR it's probably a good time to stop. Hope this helps (lol).

-Jeff

Helpful Engaging Supportive

Topic starter

Jan 21, 2022

Sean van Drogen:
On point 1 I 100% agree with Steven, most important part is if you are happy with your picture, and indeed be realistic what you can can get with your gear and time.

for point 2 I dont do well with youtube I prefer the various forums and books. I only use PI so the main books for this is what I have.
but if you have a real good data set come back to it after you learn something new and reprocesses to see if you like the result better.

Hi Drogen
I've been reading 《MasteringPixInsight》recently. This book is superb! It's gonna be great for my post-processin.

Topic starter

Jan 21, 2022

I think there is no "standard" out there what is well done and what is crap. It is all about personal feeling.
Some prefer a black background where others (as me) think that it looks unnatural. Talking about colours is another topic. Especially in NB you can do wierd things that look very good.
My personall view is, no matter if it's a nebula or a galaxy, the star colour should be well balanced. It's boring if you just have white stars as well as crazy red, pink and dark blue ones.

Find your own way and always ask some experienced people here on AB if you need a honest opinion

Stay safe!
Mike

Yes Mike, I agree with you!
As long as it fits the science and satisfies by myself.

Topic starter

Jan 21, 2022

Die Launische Diva:
The IOTD process Guidelines offer some good advice regarding the qualities a good astrophoto should have.

Regarding (1), I advise to step back from the image you process and come back after a while in order to assess it. The longer you stay away, the better. This breaks the sentimental connection between you and your image and will let you to be a more objective judge of your own work. Detachment from your effort is beneficial since you will focus on judging the result and not in celebrating on your effort or punishing yourself for the small mistakes you have done (and probably you are the only one noticing them). You want to avoid thoughts like these: "$10000 equipment, 100 hours of exposure under sub-zero temperatures, 100 hours spent on processing using the latest 32-core desktop computer... It MUST be a good image" and start thinking like this: " Does my image respect the object depicted and the data available? Does my image fulfill my intentions? Are my intentions effortlessly communicated to my audience?"

Regarding (2) I don't have a good tip, but I think it is better for anyone to first have settle on a very simple processing workflow which produces decent results, and then refine it, if it is necessary to do so. Having a solid knowledge of the available software tools helps a lot. If you have questions on the tools you are using, better ask in the proper places like the dedicated fora of the software you are using, be it PixInsight, APP, or Siril.

Lastly, for (3) you may assess your image against the IOTD Guidelines I mentioned earlier, even if you don't care so much about the IOTD.

Regarding the ever-lasting debate on the positive correlation between equipment price vs. quality, I tend to have a different opinion: Expensive equipment buy us repeatability and convenience, not processing skills and a good taste. Money does not directly result in better images. An expensive Leica camera won't turn me into a Henri Cartier-Bresson, and if you drop me in the Grand Canyon with a large-format camera you won't turn me into a modern Ansel Adams

I hope my advice is helpful and reasonable. And after browsing your image gallery, I have to say that your work seems to be already very good!

Thank you for your detailed reply, Diva!
Practice processing techniques to match the quality of the images to the time and equipment. At present, all my PI questions can be answered in the 《Mastering pixlnsight》. I think IOTD it's what I'm after, but I think I shouldn't lose my style by pursuing it too much.About 'equipment price vs. quality',I couldn't agree more! And thank you for saying so~~

Topic starter

Jan 21, 2022

Jeffrey K Lovelace:
Great topic! I think you're way ahead on the following issues, but it's worth adding a few basics for others new to this practice.

First, an often overlooked element of good editing is proper monitor calibration. I see a lot of images with an obvious color bias, or which are generally too bright with blown out highlights or too dim; and my guess is that their monitor isn't calibrated. I use datacolor's SpyderX products. It's hard to tell if your image is any good if your monitor is way our of calibration.

Here's another basic, general guideline that I often see violated: Highlights should never be so bright that you lose detail, and the dark regions should never be completely black. Also you know that you have pushed color saturation way too far when you have lost detail due to over saturation. If you can still see fine detail in your highlights, darker areas, and most color saturated areas, you are good.

Another useful technique is to compare your image to others of the same target that have earned top pick or IOTD. Only reference awarded images from the past three years as everything has advanced quite far recently. Don't try to match/copy others, but if you like your approach side by side with theirs, then you are good.

Finally, look at detail in your image down to the pixel level. Late in your editing process scan through your entire image at 400%+ magnification. look for artifacts from star removal or over sharpening, or not enough/too much noise reduction, etc. Mostly look at your stars: are they crunchy/chunky, too fuzzy, oblong/misshapen; surrounded by dark rings or color blotches, etc.? Then research or invent ways to fix these issues. If you think that others won't look at your image on that level, you would be way wrong. The IOTD staff's job is to look at images at this level, for instance. If you like your image at 5x, then you are good.

I could go on, but as I have already achieved TLDNR it's probably a good time to stop. Hope this helps (lol).

-Jeff

Thank you for your detailed answer！
I use Mac to do the process, I think Mac's screen colour is more accurate. I usually search for IOTD/TPN images when I process images! It really is a shortcut. You mentioned paying attention to detail in the image, I really didn't pay much attention to the background and stars before, just to see if there is a big problem, but you're right, the detail of the image is also important. I'll pay attention to that when I work with the graph!

Related discussions

SkyWatcher 150PDS Focuser tilt

Hi guys! I got myself a Sky-Watcher Newtonian 150PDS recently and i've been struggling with collimation for at least two months now. I am inexperienced woth newtonians so i've been reading articles and watching videos for collimation, so i...

Not relevant — focuses on collimation, not image processing techniques.

Oct 9, 2023

Any target and acquisition suggestions for areas with poor seeing?

Hi, all. I was hoping to pick folks brains about ideas for targets and perhaps acquisition techniques for areas with routinely poor seeing. For context, I live in the San Francisco Bay Area on the SF peninsula. My place of residence and where I do 90...

Not relevant — addresses acquisition in poor seeing, not processing.

Oct 23, 2023

WBPP not giving me my total integration time or subs used for my file, when splitting my DSLR OSC data into RGB channels to integrate

Question for any long time Pixinsight users. I've been using the app for a few months now. I have not been doing my initial stacking in WBPP, I use Siril. One reason for this is I don't guide, I have an older LX200 mount that PHD2 just can...

Relevant — discusses PixInsight processing workflow and image file handling.

Sep 1, 2023