Richard avatar
Hi All,

So, I ve just finished processing the Pacman nebula, roughly 17 hours worth of data processed in there (mono), and I think I ve hit a ceiling in either the location from where I'm taking pictures from at the FL I'm shooting with, or my imaging processing technique is just bad. In any case, I d really like to know your thoughts on the picture and how you think I could get a better result. Really would appreciate some constructive feedback from you on whether there are any other techniques you could recommend, I also did an HOO of the bubble, at half the total exposure time and detailwise they look very similar, so I m doubting integration time.

heres the link to the Pacman.  https://www.astrobin.com/thz931/

Thanks and look forward to your
andrea tasselli avatar
I'd question the wisdom of imaging at that image scale.
Steven avatar
Probably best to post this in the Requests for constructive critique part of the forum next time. 

Few things I'm noting

- the stars have a very red halo around them. I'm going to guess it's from the Sii data.
Do you edit using starless versions? it might be worth doing that if you aren't doing that already. Creating a starless version of the nebula, so you can pump up saturation and brightness of the nebula, without affecting the stars. And adding the stars back in after.

Or masking out all the stars, and reducing the saturation of those.

- the Oiii data is very green. Now, this is more personal taste, I prefer a slightly more blue look instead of aqua/green. Obviously it's fine if you prefer the green look.. 
I'd throw on a "Selective colour" adjustment layer, go to the "green" tab and reducing the yellows.

- maybe a bit high contrast? it looks great for the center and core of the nebula, but some of the fainter dust around the nebula is almost lost because the blacks are so black. Reducing the contrast or playing with curves might allow you to pull out a little more of that fainter dust.

I'm going to link this onehere as an example. smaller stars, better coloured stars, nice blues, loads of surrounding dust visible.
Obviously, this is depending on your raw data and equipement. And I haven't seen your raw files, but judging from your equipment, I reckon there might be more for you to pull out of your data, especially if you go in to the individual data first and edit the Ha/Oiii/Sii separately first to pull out a lot of data there, before combining them.

Either way, It's all hard to judge without knowing your processing flow and knowing how you edit your data.
Richard avatar
@Steven de VetFirstly, The linked pic for 5 hrs integration is amazing!

Secondly, thanks for the tips!

so my workflow is:
star align 
crop x 3
DBE x 3
multiscale linear transform x 3
histogram Transformation x 3
starnet x 3
rangemask
sharpen
denoise
curves
Stars back in with pixelmath
Pixelmath combination of all 3
Lrgb combination (just L with HA)
curves/saturation/hue
Richard avatar
andrea tasselli:
I'd question the wisdom of imaging at that image scale.

it was mainly for framing that i used a 1x flattener/reducer, do you think its because i am not getting a good enough amount of light in and should increase my "/pixel, currently at 0.86, reducing it by 0,72x is the best i can do, but then i lose the FL. Please explain what you think would be good to do for aquisition.
andrea tasselli avatar
So it is you drizzled the original raws? What is the average FWHM of the raws then? In terms of practicality and knowing the aperture you're working with I'd say your sweet spot should be no higher than 1"/px also considering the QE of your imager.
Uwe Deutermann avatar
Can you tell us what YOU think what is bad with your image? For me it is most important that you like what you created. Once we know we can maybe give advice of those things that you really dislike. And also important: what is your expectation? When I targeted the same object last week I had an inspiration image, knowing though that I will not get there (for several reasons). Which I did not mind since I was happy with my result.
Richard avatar
Uwe Deutermann:
Can you tell us what YOU think what is bad with your image? For me it is most important that you like what you created. Once we know we can maybe give advice of those things that you really dislike. And also important: what is your expectation? When I targeted the same object last week I had an inspiration image, knowing though that I will not get there (for several reasons). Which I did not mind since I was happy with my result

I am a little disappointed with what i m getting as a result, I seem to always be at odds between sharpness and noise. I certainly dont want my pictures to look artificial but i would like smooth pictures.  Maybe my noise reduction technique is bad or my sharpening skills are. But i ve integrated a significant amount of data, so I m also wondering whether i would get a better signal to noise by integrating more data, or whether I should be looking at changing settings when i capture (gain/ exposure time per frame). Where would i benefit the most?
I shoot at unity gain for the ASI1600gt across all filters - 139, with this picture at -10 but now as its getting colder i can go to -20.
the Fwhm according to DSS is 3.54 on my best frame and i exclude everything after 4.0
My guiding is giving me an average readout of 0.43 considering I am using an OAG the FWHM of the guidescope is also the same sitting at an average of 4.0.
With narrowband imaging I had hoped to achieve better results, when i look at my OSC result for 3 hours integration I m getting a better picture IMO, sure its got its flaws but the signal to noise is much more satisfying. Orion is also much lower to the horizon in comparison to the pacman but the pixel size is identicle allbeit a larger FOV. Maybe i should spend an evening on it with my OSC and see what i can get.

M42 revisited 1 year on

https://www.astrobin.com/zftwd0/
andrea tasselli avatar
You can't compare the two. First because M42 is MUCH brighter than NGC281 and secondly because the ASI2600 is much more efficient than the ASI1600.
Steven avatar
Would have to agree with Andrea here.
Orion is much brighter than the Pacman Nebula. also about 8000 lightyears closer to the earth, making it a very different target to shoot and giving very different results.. If you want a better comparison, maybe you should shoot targets that are brighter and/or closer.

And yes, the 1600GT and the 2600 are 2 different beasts. and it's difficult to look at 2 images and compare the 2.
Lower read noise, 12bit vs 16bit depth, larger dynamic range. creating better noise performance and better depth already. 

And mono is more challenging to edit. finding a balance between the layers, colours, luminance layers, stars, etc can be very frustrating. And trying out all the different palettes, SHO, HOO, OSH, etc..  But with my colour RGB camera, I can just remove the stars, do some tweaks to contrast, curves, some other bits and bobs, plop the stars back in and "hey presto" a final image..

I'm not a pixinsight user, so I can't say much about your workflow. the only thing I can suggest is to look at an absolute crap tonne of videos of others processing their narrowband data in pixinsight and photoshop on YouTube. There are so many tricks and different ways of doing things, there isn't just 1 method that will work for all narrowband images… Often a trick that works for the data of nebula "A" won't work for the data of nebula "B".
Richard avatar
andrea tasselli:
You can't compare the two. First because M42 is MUCH brighter than NGC281 and secondly because the ASI2600 is much more efficient than the ASI1600.

I get that there are obvious differences in the cameras, brightness in the objects, and different bit depth, I own the two cameras and have taken both pictures. That wasnt quite what i was getting at when showing them and should have perhaps taken a different example. like The one Steven provided. difference being integration time ,location and processing technique.  https://www.astrobin.com/me578g/B/

Coming back to my initial question, whether or not my location, my integration time or my technique is the thing that is limiting my picture.
I ll make the stacked data available and move it over to the other forum, and see if others can make something better out of the data and hopefully be able to learn something in the process, in parallel i ll use my 2600mc on the same target and see what the outcome is. regardless of the difference in cameras I find it hard to believe that i cant do better with a mono camera over the 2600mc, disappointing.
Uwe Deutermann avatar
Richard:
I am a little disappointed with what i m getting as a result, I seem to always be at odds between sharpness and noise. I certainly dont want my pictures to look artificial but i would like smooth pictures.  Maybe my noise reduction technique is bad or my sharpening skills are. But i ve integrated a significant amount of data, so I m also wondering whether i would get a better signal to noise by integrating more data, or whether I should be looking at changing settings when i capture (gain/ exposure time per frame). Where would i benefit the most?
I shoot at unity gain for the ASI1600gt across all filters - 139, with this picture at -10 but now as its getting colder i can go to -20.
the Fwhm according to DSS is 3.54 on my best frame and i exclude everything after 4.0
My guiding is giving me an average readout of 0.43 considering I am using an OAG the FWHM of the guidescope is also the same sitting at an average of 4.0.
With narrowband imaging I had hoped to achieve better results, when i look at my OSC result for 3 hours integration I m getting a better picture IMO, sure its got its flaws but the signal to noise is much more satisfying. Orion is also much lower to the horizon in comparison to the pacman but the pixel size is identicle allbeit a larger FOV. Maybe i should spend an evening on it with my OSC and see what i can get.


Please do not be too disappointed, I looked at it now in full resolution and there is a lot in there that you can be happy about! And I liked your answer what you want to achieve! I copied your full resolution image into my viewer and I am actually not as critical about the noise, it looks fairly good. And ... a complete noiseless image is sometimes looking very artificial, and there are quite a few here on Astrobin that do not favor this at all. 
Also good information about your setup and what you are using. And since you have the VERY same telescope I might be of help, although ... I am not the TP or IOTD gatherer like others here. 

I used with the ASI1600 usually a gain of 200 with offset 30, I liked that most. Some people go even up to 300/50. I tried that once and was not SUCH a fan of it, but maybe I did not give it enough time (all for narrow band that is). Being cooler is always better in general, so go for it when it is getting colder 😊. 
My FWHM is usually pretty close to yours, I know ... people say that this is not good enough, but one can do something about it with processing, hence I think that this is ok. 

With my recent image I actually did almost the same that you did (same target: https://www.astrobin.com/cst4xu/), comparing with a wonderful extraordinary image (see version C and https://www.astrobin.com/l8tmba/), but I knew that this would be impossible to achieve. So what is the main difference? Definitely sky conditions! Maybe more FL, but then the FoV is too small. Definitely processing skills, the top notch imagers here on Astrobin have many many years of experience. I can see that for myself, I tried more and more and more and I believe I am getting better. 

Coming back to your image! What did I not like (and believe me, I have the same list for my image)???? And please don't forget, this is a personal opinion and not a general one, others might see it much different.
The first thing that popped up were the stars: they do have a red ring around them, I think all of them. This gives the whole image a red "noise", beside that the stars do not look natural. The size of the stars are a tick too large, I would try to reduce them (there are multiple techniques in processing, and no ... star reduction on the final image is not the best answer, I usually combine star images with starless  images a lot). And then maybe the most subjective opinion: I am not a fan of exaggerated color, I would have tamed it definitely down. Also ... a lot of people do not like any green in the images (personally I like it sometimes, so I am not a 100% opponent of it), so the inner core is "not blue enough". Maybe I would play around with some sharpening tools, that would make the filaments more detailed, but not overdoing it, if your data do not give it, then do not do it.

Ok, sorry for the long response, but I thought your reply was very honest and I read your (not necessarily needed) frustration, your image is not as bad as I think you think about it. And not every image will be a TP or at least a nomination (I did not get one either 😊, and I truly believe that it deserved one 😊, of course 😊😊). 

So please PM me if you have some question about some of the techniques that can be used, I will try to give you my best answer that I have (not the best of all of the people here, but at least an ok one I believe). Keep on going Richard!

Uwe
Matthew Proulx avatar
My .2c
For targets that are 100% emission nebula, I would put the most integration time into the Ha and make a lumainance master predominently (if not all) Ha.

I think theres lots of data already there to work with.
Richard avatar
Uwe Deutermann:
Richard:
I am a little disappointed with what i m getting as a result, I seem to always be at odds between sharpness and noise. I certainly dont want my pictures to look artificial but i would like smooth pictures.  Maybe my noise reduction technique is bad or my sharpening skills are. But i ve integrated a significant amount of data, so I m also wondering whether i would get a better signal to noise by integrating more data, or whether I should be looking at changing settings when i capture (gain/ exposure time per frame). Where would i benefit the most?
I shoot at unity gain for the ASI1600gt across all filters - 139, with this picture at -10 but now as its getting colder i can go to -20.
the Fwhm according to DSS is 3.54 on my best frame and i exclude everything after 4.0
My guiding is giving me an average readout of 0.43 considering I am using an OAG the FWHM of the guidescope is also the same sitting at an average of 4.0.
With narrowband imaging I had hoped to achieve better results, when i look at my OSC result for 3 hours integration I m getting a better picture IMO, sure its got its flaws but the signal to noise is much more satisfying. Orion is also much lower to the horizon in comparison to the pacman but the pixel size is identicle allbeit a larger FOV. Maybe i should spend an evening on it with my OSC and see what i can get.


Please do not be too disappointed, I looked at it now in full resolution and there is a lot in there that you can be happy about! And I liked your answer what you want to achieve! I copied your full resolution image into my viewer and I am actually not as critical about the noise, it looks fairly good. And ... a complete noiseless image is sometimes looking very artificial, and there are quite a few here on Astrobin that do not favor this at all. 
Also good information about your setup and what you are using. And since you have the VERY same telescope I might be of help, although ... I am not the TP or IOTD gatherer like others here. 

I used with the ASI1600 usually a gain of 200 with offset 30, I liked that most. Some people go even up to 300/50. I tried that once and was not SUCH a fan of it, but maybe I did not give it enough time (all for narrow band that is). Being cooler is always better in general, so go for it when it is getting colder 😊. 
My FWHM is usually pretty close to yours, I know ... people say that this is not good enough, but one can do something about it with processing, hence I think that this is ok. 

With my recent image I actually did almost the same that you did (same target: https://www.astrobin.com/cst4xu/), comparing with a wonderful extraordinary image (see version C and https://www.astrobin.com/l8tmba/), but I knew that this would be impossible to achieve. So what is the main difference? Definitely sky conditions! Maybe more FL, but then the FoV is too small. Definitely processing skills, the top notch imagers here on Astrobin have many many years of experience. I can see that for myself, I tried more and more and more and I believe I am getting better. 

Coming back to your image! What did I not like (and believe me, I have the same list for my image)???? And please don't forget, this is a personal opinion and not a general one, others might see it much different.
The first thing that popped up were the stars: they do have a red ring around them, I think all of them. This gives the whole image a red "noise", beside that the stars do not look natural. The size of the stars are a tick too large, I would try to reduce them (there are multiple techniques in processing, and no ... star reduction on the final image is not the best answer, I usually combine star images with starless  images a lot). And then maybe the most subjective opinion: I am not a fan of exaggerated color, I would have tamed it definitely down. Also ... a lot of people do not like any green in the images (personally I like it sometimes, so I am not a 100% opponent of it), so the inner core is "not blue enough". Maybe I would play around with some sharpening tools, that would make the filaments more detailed, but not overdoing it, if your data do not give it, then do not do it.

Ok, sorry for the long response, but I thought your reply was very honest and I read your (not necessarily needed) frustration, your image is not as bad as I think you think about it. And not every image will be a TP or at least a nomination (I did not get one either 😊, and I truly believe that it deserved one 😊, of course 😊😊). 

So please PM me if you have some question about some of the techniques that can be used, I will try to give you my best answer that I have (not the best of all of the people here, but at least an ok one I believe). Keep on going Richard!

Uwe

Thanks Uwe, your feedback means a lot to me, and its all good! Nothing, that I would disagree with!!! And its important to stay honest! I ve posted a new version taking your feedback on board, having more focus around the star colour, they ended up a little diamond like, that was my bad I should have done it a little differently, but i m now satisfied. I went with a more traditional palette for the colors as well this time, let me know what you think.
Uwe Deutermann avatar
Just my view on this image, but it looks for me already much cleaner! Nice!
Related discussions
Frustration with Observatory and HOA
I have plenty of room for a small observatory in my backyard but have been frustrated at every angle by the HOA!! My wife in annoyance said "they've let people build raised gardens, trellises and benches but not a little observatory. That...
Potentially relevant; author may relate to location/setup constraints affecting imaging quality.
Jul 9, 2023
How to process dark nebulas
Hey guys I need help. I'm doing a 10 hour project with my skywatcher 150/750mm and my dslr shooting at a dark nebula like the snake nebula. I basically don't know how to process this type of dark nebulas, I'm not sure if it requires some ...
Author processed extensive nebula data; may share relevant processing techniques.
Jul 14, 2020