result of two nights combined

[TiffsAndAstro]

just wondering if anyone can suggest anything for this result?

i can see some blobs from flats not working (maybe?) and a red and a green gradient. this is from two groups stack in dss tried average and median kappa seem to get same or very similar results.

i'd do this in siril but having some issues with siririlc.

 

[Elp]

You need to do a background extraction on the image to process it further. I suspect it will cause a random red green pattern issue though. I have a lot of issues imaging "true" colour in my bortle 7 and get similar to you.

What issues are you having with Siril?

[TiffsAndAstro]

31 minutes ago, Elp said:

You need to do a background extraction on the image to process it further. I suspect it will cause a random red green pattern issue though. I have a lot of issues imaging "true" colour in my bortle 7 and get similar to you.

What issues are you having with Siril?

Sorry yeah I know to do a background extraction (and more,) it's part of my workflow but I never seen a red gradient and a green one on same stack before. Also they seem perpendicular to each other.

Stars in corners look square but maybe a little back focus too far from sensor. Maybe under sampled more likely ? 

Guiding was about 0.20 (1.6") all night (though straight up) not sure if those guiding numbers are ok? My image scale is 2.02" so maybe?

Hfr was under 3 all second session, but those sins were 60 sec first session was 120 sec. Second session sins max score in dss was 600, min score in first session was 1000.

Both sessions same dither setting every 3 or 4 subs 7px in siril.

As for sirilic, it goes through and works fine until it tries to look for .seq files then aborts. I'm wading through the docs for siril but struggling to fix.

Id guess I'd missed a check box, but very few I need to check.

Below is a quick post process including photometric cc and I've over saturated cos I also think "natural colour" isn't great but no filters.

Moon was up and bright though.

Was hoping to see a bigger improvement after two nights combined. Optimistically id say noise was reduced a bit. I probably shouldn't expect too much until I'm a bit more consistent and go 2 to 4 to many more hours.

Edited April 21 by TiffsAndAstro

[Elp]

When you load files in Siril to process there's a drop down box near the bottom right where you input a filename, the drop-down you can choose something like linking to your files, saving to new images or making a sequence. The sequence option writes all the new files to a large sequence file in your default Siril processing folder (so you need a lot of free HDD space), and every operation thereafter also generates a new sequence file so you can go back to a previous option. I personally process to sequences, then delete them once I have my final stack.

[TiffsAndAstro]

11 minutes ago, Elp said:

When you load files in Siril to process there's a drop down box near the bottom right where you input a filename, the drop-down you can choose something like linking to your files, saving to new images or making a sequence. The sequence option writes all the new files to a large sequence file in your default Siril processing folder (so you need a lot of free HDD space), and every operation thereafter also generates a new sequence file so you can go back to a previous option. I personally process to sequences, then delete them once I have my final stack.

Ah yeah, I'm trying sirilic to (effectively ) create a custom script. It seems pretty ace, just struggling a little.

I'll see if it plonked the .seq files in wrong place.

Ty for the help

 

and yes, I massively underestimated storage space. Just for post processing. Never even considered storing multiple sessions.

currently trying to beg/borrow/steal stuff to concoct a cheap Nas or possibly flash drive raid arrays lol.

and then double that for some sort of redundancy

Edited April 21 by TiffsAndAstro

[Vroobel]

27 minutes ago, TiffsAndAstro said:

Was hoping to see a bigger improvement after two nights combined.

 

It's exactly the bigger improvement.

You don't need to use any sophisticated techniques to finish it, it's OK. You can add some photons next time if you want.

How do you deal with noise in the post-processing?

[TiffsAndAstro]

4 minutes ago, Vroobel said:

 

It's exactly the bigger improvement.

You don't need to use any sophisticated techniques to finish it, it's OK. You can add some photons next time if you want.

How do you deal with noise in the post-processing?

That's kind of you to say, but the edge stars are now squarish in the combined stack which to me means it must be worse in whatever respect that suggests.

I don't really handle noise in post except a remove green noise before I photometric cc. It works well from what I can tell.

The general noise reduction feature doesn't really seem to have much effect, but I haven't watch any videos on it yet.

I'm using deep space astro siril YouTube tutorials mainly. His videos are a nice balance between brevity and really in depth.

[Vroobel]

Could you share the master file, please?

[TiffsAndAstro]

1 minute ago, Vroobel said:

Could you share the master file, please?

Sure will be an hour or so though on phone atm.

You want the tiff of the combined stack I took a quick screener of at posted at the top?

It's about 700mb from memory I think.

I "solved" the drizzled starnet issue. I just resample back down to sensor size before star removal. It's not ideal. Obviously none of that except the drizzle will be in the tiff I saved from dss

 

[Vroobel]

Tiff please.

[TiffsAndAstro]

53 minutes ago, Vroobel said:

Tiff please.

tiff ty

2 nights m101 for vroobel.tif

[Vroobel]

Now it looks better and also didn't require so much stretching.

 

 

Edited April 21 by Vroobel

[Elp]

I wouldn't worry too much about NR at this point.

[TiffsAndAstro]

1 hour ago, Elp said:

I wouldn't worry too much about NR at this point.

its a dslr, its going to have lots of noise. i really dislike the obvious red stuff, but hopefully more total integration time will improve it.

also a camera upgrade.

just concerning to see two different coloured gradients across my first attempt at stacking across multiple sessions and wondered if i did something wrong.

Edited April 21 by TiffsAndAstro

[TiffsAndAstro]

5 hours ago, Vroobel said:

Now it looks better and also didn't require so much stretching.

 

 

that looks pretty amazing tbh

maybe ill try and get pixinsight if/when i ever slightly approach your skill level with gear and processing.

Edited April 21 by TiffsAndAstro

[Elp]

10 hours ago, Elp said:

I wouldn't worry too much about NR at this point.

It was in reply to the example processing. All the galaxy detail has been blurred out via using NR on it. It can be processed starless and use NR on the star only layer, or mask the galaxy and don't use NR on it.

You need many hours of total time, short sessions will result in noise, but the final result even when using a  cooled astro camera will require NR, the noise pattern is simply finer.

Many people have been imaging with dslr bodies for years and get good results.

Edited April 22 by Elp

[TiffsAndAstro]

19 minutes ago, Elp said:

It was in reply to the example processing. All the galaxy detail has been blurred out via using NR on it. It can be processed starless and use NR on the star only layer, or mask the galaxy and don't use NR on it.

You need many hours of total time, short sessions will result in noise, but the final result even when using a  cooled astro camera will require NR, the noise pattern is simply finer.

Many people have been imaging with dslr bodies for years and get good results.

I don't quite follow, sorry.

When you say "All the galaxy detail has been blurred out via using NR on it'" do you meany crap quick post process or vrobells quite amazing result from same data?

I didn't (as far as I know lol) use any nr, but I'm assuming vrobel did. And I can see way more detail in his image. Way way more.

Vrobel image made me sad and happy. Sad my processing skills are rubbish in comparison , but happy that someone with skills can make something that (to me) looks so incredibly better with my same limited data.

I think it shows my data acquisition seems ok, but I need to work a lot more on my processing skills.

******************

ok, looking at both images again, and i can see how vrobel's is more blurred (in a way) but i think its a much, much better image than mine by quite a long way. My version looks more grainy and granular and far worse.  I can tell my own post processing skills are going to need a lot of time and practice to improve, and, even if they do,  my aesthetic sense of what is attractive on the eye may need even more work

Edited April 22 by TiffsAndAstro

[alacant]

5 minutes ago, TiffsAndAstro said:

I don't quite follow,

May I help? If you compare the galaxies, you see this:

[Elp]

6 minutes ago, TiffsAndAstro said:

but I'm assuming vrobel did. And I can see way more detail in his image. Way way more.

It's not an attack on his processing, @Vroobel likely did it quick to show the background cleanup that can be achieved if you know how and that showed an excellent difference in this regard. @alacant has highlighted the difference above that I mentioned about application of NR.

[Elp]

I always say something like AP is around 20pc image acquisition, and the remainder is post processing skill. You can have the best data in the world but you won't see much with poor PP skills, average data can look pretty good with good PP skills.

[TiffsAndAstro]

13 minutes ago, alacant said:

May I help? If you compare the galaxies, you see this:

maybe i can put each on a different layer in gimp and get the best of both worlds (whorls ) ?

i think i have enough to be going on with, collecting my images and then processing them (badly) without even really having thought much about the aesthetics. I showed a friend without saying which was which and they said mine had more detail and vrobels is more blurry.

ill carry on adding to my m101 project in the hope that by the time i get maybe 20 hours of data my processes skills will have improved a bit more by then.

[TiffsAndAstro]

15 minutes ago, Elp said:

I always say something like AP is around 20pc image acquisition, and the remainder is post processing skill. You can have the best data in the world but you won't see much with poor PP skills, average data can look pretty good with good PP skills.

i'd have preferred it if you'd said the oppopsite as while both are not great, i think my aquisition skills are ok, and my processing skills are bad.

[TiffsAndAstro]

19 minutes ago, Elp said:

It's not an attack on his processing, @Vroobel likely did it quick to show the background cleanup that can be achieved if you know how and that showed an excellent difference in this regard. @alacant has highlighted the difference above that I mentioned about application of NR.

i didn't read it as an attack and i hope vrobel didn't either. i think its another interesting and challenging part of this pass time. 

again, i really appreciate everyone's contribution to this thread (again) as even some quick comments i can find really useful, if only to challenge my 'default' thinking on this subject. 

[TiffsAndAstro]

vaguely related question, if i got an uncooled osc (im thinking an imx585 based sensor) would this data i already have, taken by my dslr, still be of use in an ongoing project, or would i likely need to start from scratch? 

[Vroobel]

Well, indeed, I was focused on the noise reduction too much. I agree that the main object is blurred. I used a GeneralizedHiperbolicStretch which I still learn. 

BTW, I operate on the layers, process mostly the starless and use masks. I have to look deeper into the noise related to the stars layer.