"Busy" looking images?

[STEVE333]

I used a different technique to reintroduce the stars and like the results better. The stars look more "natural", especially over the nebula. So many options on how to do processing!

East_Veil_Stars_2 by STEPHEN KING, on Flickr

Cheers,

Steve

[Larry 1969]

I also used for 5 for a long time. 1 does not work for me - at all. I tried 2 and lo and behold the image was vastly improved. Well, as I type this I also just realized that I made the switch from 5 to 2 when I stopped using the Canon 70D and shifted to the ASI294MC cooled. I recall trying various settings and 2 seems to be perfect for the 294. Unfortunately, StarTools doesn't explain a lot of the underlying math involved, and the limited help is mostly useless. I started out doing what many tutorials say to do, but when I get an image that is not satisfactory - or what I think it should look like then I begin to tinker with settings. But it's mostly guess work. I wish there were better explanations for each of the "steps" i.e. contrast, demosaic, life, etc do and what changing the values really does. For instance what does the difference in structure size really mean and when should I apply one of the three choices?

Interesting... I too just left a DSLR (Canon 60D) and switched up to a 071 MC Pro... I'll have to fool with the dark anomaly setting...
I find StarTools pretty complicated, but I have literally ZERO background in any image processing. I was new to astronomy and photography when I took up astro-photography... It's been a very steep, yet enjoyable learning curve for sure! Mostly because of the generous help from the fine folks on this forum and some others!

I don't think Pixinsight would be any "easier" to use either. Not only do you have to pay up for the software, you have to pay to attend classes on how to use it... Well, you don't HAVE to, but the fact that people are willing to pay to learn it tells me it's probably not any "easier" than StarTools...

Larry

[STEVE333]

I find StarTools pretty complicated, but I have literally ZERO background in any image processing. I was new to astronomy and photography when I took up astro-photography... It's been a very steep, yet enjoyable learning curve for sure! Mostly because of the generous help from the fine folks on this forum and some others!

I don't think Pixinsight would be any "easier" to use either. Not only do you have to pay up for the software, you have to pay to attend classes on how to use it... Well, you don't HAVE to, but the fact that people are willing to pay to learn it tells me it's probably not any "easier" than StarTools...

Larry

Hi Larry - I feel your "pain" learning how to process DSO images. I haven't tried StarTools, but, have seen many beautiful images processed with it.

I started off with Gimp then moved up to Photoshop CS2. I found many helpful tutorials for Photoshop which got me started. After another year or so I finally moved to PixInsight. I did have to pay the one-time purchase price, but, all subsequent upgrades have been free. I was happy to learn that there are many free video tutorials for PixInsight too. I've never taken any classes because the tutorials seemed to answer my questions. Of course there is no substitute for "playing" with the software to get a feel for it.

Whichever way you go I wish you the best of success. You sound willing to take on the processing challenge which is at least half of the battle.

Look forward to your next image(s).

Steve

[Larry 1969]

Hi Larry,

Depending on whether you are a purist, relying on "made up" data may be objectionable. As a developer I would leave that decision to the user (there is a guide for using StarNet++ with StarTools here if you want for example), however when publishing an image, I, for one, would certainly want to know about any treatment that completely made up data. Sadly (fortunately?) I can usually pick such a treatment quite easily, but others may not.

Thanks again Ivo!
This is what I think I like about StarTools!
I believe my images are scientifically correct whether I would prefer them to appear a certain way or not.
Some astro-photographers are after the "best possible image" while I'm after the "best I can do with the real data I alone captured"...
That's what makes it "real" for me! Anyone can look up Hubble images but these photons actually hit my camera's sensor!

Larry

[Larry 1969]

Whichever way you go I wish you the best of success. You sound willing to take on the processing challenge which is at least half of the battle.

Look forward to your next image(s).

Steve

Thanks Steve! Certainly not knocking Pixinsight. It appears to be the Cadillac of software as there are a TON of users.
I'm starting to get used to StarTools and I'm going to stick with it. Plus Ivo supports this forum to some extent.

Practice, practice and practice.....

Larry

[Larry 1969]

I used a different technique to reintroduce the stars and like the results better. The stars look more "natural", especially over the nebula. So many options on how to do processing!

East_Veil_Stars_2 by STEPHEN KING, on Flickr

Cheers,

Steve

I do like what you were able to do with my image....
Thanks again Steve!

Larry

[startoolsastro]

I usually then do another auto dev with a slice of ROI and keep but I found that to look over developed so I did do a manual develop.
If I were to do another auto dev, what would you suggest for my ROI? I just used a slice of the nebulosity...

That should work indeed; whatever is a good representation of the dynamic range problem that AutoDev needs to solve for. "Ignore Fine Detail <" should help if your image has a lot of noise grain.

Unfortunately, StarTools doesn't explain a lot of the underlying math involved

If I told you it specifies the kernel size of a Gaussian filter, would that help? I personally, however, don't think it is very useful to explain the basics of convolution filters every time one is used, as that would get old very quickly (and lose a lot of people)!

limited help is mostly useless.

I started out doing what many tutorials say to do, but when I get an image that is not satisfactory - or what I think it should look like then I begin to tinker with settings. But it's mostly guess work. I wish there were better explanations for each of the "steps" i.e. contrast, demosaic, life, etc do and what changing the values really does. For instance what does the difference in structure size really mean and when should I apply one of the three choices?

I think I know where you're coming from, however there are, for starters, no less than two manuals (if you're English speaking that is), offline in-app help (per per module and parameter), online in-app help (per module), website with documentation per module, and (very) detailed user notes (aka "Guy's User Notes" on the forums). All that said, "basic" knowledge about what a module or algorithm does, is however assumed.

E.g. if something is titled "Wavelet Sharpening", it is assumed that the user is across how a wavelet transform works and how the parameters roughly relate to the variables that go into it - this aspect will not be in the StarTools documentation. The "basic" mathematical principles behind that can be found on, for example, Wikipedia. Note I keep using "Basic" between quotes here because, more often than not, it is not exactly light matter to digest. And consequently - more often than not - users are not interested in learning the exact ins and outs of an algorithm; for many it is often enough to visually see what a parameter does through experimentation.

Of course, a more fundamental understanding of what it is going on, and how things fit in a valid signal processing workflow, is greatly beneficial. StarTools in that sense is a somewhat problematic subject, as it innovates on basic/expected/familiar "textbook" signal processing as - for example - found in PI. E.g. you're all studied up, and you're finally across what you can do and can't do in your signal path, how it all works mathematically, how deconvolution or color balancing can only be performed on linear data etc., and then StarTools pulls the rug from under you again by somehow committing every sin under the sun with mathematically correct results. Everything you learned is seemingly wrong/outdated again. Compared to traditional applications, ST's signal evolution Tracking/3D processing thing is an extra "head screw" layer on top (but absolutely worth getting your head around if you are so inclined!)

Hope this make sense/help!

[Star Dad]

"If I told you it specifies the kernel size of a Gaussian filter, would that help? I personally, however, don't think it is very useful to explain the basics of convolution filters every time one is used, as that would get old very quickly (and lose a lot of people)! "

Well, I do happen to understand what a Gaussian filter is - at least theoretically. But what is the "structure size" that we are supposed to select? Ie what is small vs what is large? Are we talking a small nebula vs a very large one? Does it relate to the density of the object - ie a "thin", faint nebula vs a brilliant, condensed one - ie Veil VS Ring? For me that is what is lacking. I don't understand the meaning of the word structure as used in this context.

I will have to re-read the docs. I read them some time ago and maybe just glossed over the details while trying to get a feel for the software.

BTW I am in no way denigrating StarTools. I LOVE StarTools. It has really improved my imaging by an immeasurable amount. I recommend it to all my AP friends.

[Larry 1969]

BTW I am in no way denigrating StarTools. I LOVE StarTools. It has really improved my imaging by an immeasurable amount. I recommend it to all my AP friends.

Same here! I tell everyone about this software!
I have a Facebook page dedicated to my AP and give StarTools credit every time.

Larry

[startoolsastro]

Thank you [mention]Star Dad[/mention] and [mention]Larry 1969[/mention]. Nothing is perfect, so if there are areas ST can/should improve in, please do let me know. That goes for the documentation as well (though with my former comments/post in mind).

Star Dad, I think you may really appreciate Guy's User Notes on the forum (for example these notes on the Sharp module).

Clear skies!

[Larry 1969]

Hi Larry,

If you're using StarTools, the Life module gives you a few ways of pushing back busy star fields and re-focus the viewer's attention on the larger scale structures. Try the Isolate preset (w/o mask), or try the a tweaked version with Parameter [Detail Preservation] set to [Min Distance to 1/2 Unity]. You can even run two iterations (a combination of the two ) if you want;



Note that the above techniques do not use star masks at all.

Thanks for the suggestion Ivo, however, if I clear the mask in the "Life" module I get a message stating "Life cannot function without a mask"...
Is there some sort of workaround for that?

Larry

[bobharmony]

Hi Larry,

If you're using StarTools, the Life module gives you a few ways of pushing back busy star fields and re-focus the viewer's attention on the larger scale structures. Try the Isolate preset (w/o mask), or try the a tweaked version with Parameter [Detail Preservation] set to [Min Distance to 1/2 Unity]. You can even run two iterations (a combination of the two ) if you want;



Note that the above techniques do not use star masks at all.

Thanks for the suggestion Ivo, however, if I clear the mask in the "Life" module I get a message stating "Life cannot function without a mask"...
Is there some sort of workaround for that?

Larry

Larry - I believe the workaround is to use a "Fill" mask for the Life module. It is a mask with every pixel selected.

Bob

[SkyHiker]

The original post has images where the stars have horizontal and vertical artifacts and most reprocessed images do too except Ivo's. Does StarTools somehow fix all non round stars, or did Ivo use Larry's original image as a starting point?

Working in Gimp, using star masks inevitably introduces artifacts while Starnet++ does a much better job. It gives me a lot of freedom in how I mix the stars back in and generally gives me much better results than when I don't use it. My imaging time is usually less than 2.5 hours so my images are rather noisy and sensitive to improvements. I spend a lot of time editing and enjoy doing the pixel math.

[Star Dad]

StarTools can round stars. If tracking is not quite perfect I get tiny ellipses on occasion. I then use the round star tool and it certainly makes them look a lot better.

[Larry 1969]

Larry - I believe the workaround is to use a "Fill" mask for the Life module. It is a mask with every pixel selected.

Bob

Thanks Bob!
That's what I've been doing up until now but Ivo stated "Try the Isolate module (w/o mask)" and that made me try to remove the mask all together which resulted in the message...

I set up a new imaging PC last night and tried it out on The Cocoon Nebula. In between the clouds I was able to get almost 3.5 hours on target. I could certainly use more TOT but, during processing, I did remove the mask and made a star mask then ran Isolate in the Life module and my result was far less "busy" looking.

[Larry 1969]

Larry, those are stunning images... The Veil in particular.

I took the liberty of throwing your image of The Veil into Photoshop to see what I could do with the stars. I put a feathered mask around The Veil and adjusted the histogram to dampen the brightness of the surrounding area.

(I'm a solar imager and haven't shot a DSO since moving to Las Vegas ten years ago.) LOL

Thanks Lowjiber! I do like that but, I'm beginning to accept the fact that those stars are really there and they just make for a busy image...
I guess I'm more of a purist than an artist...

Larry

[Gordon]

I took a crack at it using the 'star reduce' function in star tools.

[Larry 1969]

I took a crack at it using the 'star reduce' function in star tools.

Thanks Gordon!
That is pretty cool except for the ringing or little halos around the stars.
Where exactly is the "star reduce" function?

Larry

[SkyHiker]

I'm beginning to accept the fact that those stars are really there and they just make for a busy image...
I guess I'm more of a purist than an artist...

I don't think it has anything to do with purity, the image is made busy by the processing that you do on it to bring out the nebulosity. The blow up is an artifact from the atmosphere and technology so why not use technology to suppress them.

Your starmask-processed image shows blurring artifacts. You could try blurring it less but how about giving Starnet++ a shot then decide what you like best?

BTW you have very nice material and I'm nit picking a bit.

[Larry 1969]

I don't think it has anything to do with purity, the image is made busy by the processing that you do on it to bring out the nebulosity. The blow up is an artifact from the atmosphere and technology so why not use technology to suppress them.

Your starmask-processed image shows blurring artifacts. You could try blurring it less but how about giving Starnet++ a shot then decide what you like best?

BTW you have very nice material and I'm nit picking a bit.

Thanks Henk!
I do appreciate the feedback!
I will look into this "Starnet++" and give it a shot!

Larry