#1  
Old 22-09-2018, 03:09 PM
Gavin1234
Registered User

Gavin1234 is offline
 
Join Date: Mar 2018
Location: Sydney
Posts: 227
Not improving

Had a brief window of cloudless skies over the past two nights so I got some more data on my two current targets; horsehead and helix.

I’ve now got about 150 x 5 min subs of horsehead and about 380 x 5 min subs of helix. I’ve integrated the horsehead ones and it went down to about 100 frames during pre processing. Have I just stuffed up the post processing?

I’m still working on integrating the helix ones but I’m a little disappointed at the lack of improvement in my image with the amount of data I’ve been able to add. Truth is that if I was looking at them for the first time I’m not sure I could pick which one was first (with minimal data) and which one had all 100 subs. I know my colour balance is not the same but what do you guys think?

The first one with the very orange flame nebula is my first one with only about 20 subs. The middle one has a lot more subs and the final one with the blue stars and white flame nebula is my latest version with about 100 x 5 min subs used.
Attached Thumbnails
Click for full-size image (ED0A1642-B73E-4725-85C0-4A3C607D085C.jpg)
195.3 KB80 views
Click for full-size image (9FAC12CC-9E3A-45AA-8C84-43E28ADC9E6A.jpg)
201.0 KB74 views
Click for full-size image (6479584D-DC31-4501-B47A-362738E63A3F.jpg)
201.1 KB95 views
Reply With Quote
  #2  
Old 22-09-2018, 04:17 PM
glend (Glen)
Registered User

glend is offline
 
Join Date: Jun 2013
Location: Lake Macquarie
Posts: 7,033
Hard to know how to help without more detail. Your big stars are really burning up the images. What software are you using, what camera, scope, etc? Did you grade your subs? Using DSS and picking the best %?
Describe your workflow.
Reply With Quote
  #3  
Old 22-09-2018, 05:59 PM
The_bluester's Avatar
The_bluester (Paul)
Registered User

The_bluester is offline
 
Join Date: Feb 2011
Location: Kilmore, Australia
Posts: 3,342
I think there is more to be had in the processing (Which I am a long, long way from expert at) but there is less noise and more detail in the last image than in the other two so the extra subs would look to be worthwhile.
Reply With Quote
  #4  
Old 22-09-2018, 06:11 PM
Gavin1234
Registered User

Gavin1234 is offline
 
Join Date: Mar 2018
Location: Sydney
Posts: 227
Thanks Glen, so I used a zwo071mc main camera cooled to -20. Sw Esprit100ED scope, CGX Mount. PHD2 for guiding with decent alignment. Sequence generator pro for exposures.

50 flats, 50 darks, 250 bias.

So using PixInsight I used blink to take out all the crap then I used subframe selector and took out the worst remainders based on FWHM then did the same again using SNRWeight. By the end of the time I stack the lights I’m left with 2/3rds of the originals subs i.e. 150 Dow to 100.


I’ve calibrated the darks and bias, then created masters before using them to calibrate the flats and then made a master flat. Then I calibrated the lights using the darks, bias and flat masters. Then I debayered them, then I aligned them, then I integrated them.

I usually do a small crop to start post processing, followed by an automatic background extraction, colour calibration, noise reduction using Atrous waveletTransform, Then I stretch with histogram transformation, then a curves transformation. That’s basically my workflow, but there are slight differences with each version for example in the first image I used tgvdenoise instead.

Thanks again for taking the time to help. I know none of these images are finished it’s just that I expected to see a much bigger difference at this stage in return for all the extra hours gathering data between the first image and the third.
Reply With Quote
  #5  
Old 22-09-2018, 06:13 PM
Gavin1234
Registered User

Gavin1234 is offline
 
Join Date: Mar 2018
Location: Sydney
Posts: 227
Thanks Paul, yeah I hope I can get more out of it with PixInsight (which I’m very new to). I’ll keep reading up on it but If anyone has any suggestions as to other PixInsight tools/processes I should run on these I’d love to hear them.

Thanks again
Reply With Quote
  #6  
Old 22-09-2018, 07:24 PM
Imme (Jon)
Registered User

Imme is offline
 
Join Date: Aug 2017
Location: Kyneton
Posts: 840
I disagree..... I think it is getting better!
I think the clarity is increasing and noise is reducing
Reply With Quote
  #7  
Old 22-09-2018, 07:39 PM
xelasnave's Avatar
xelasnave
Gravity does not Suck

xelasnave is offline
 
Join Date: Mar 2005
Location: Tabulam
Posts: 16,866
There is a great improvement in my view.
Reprocess until you are happy.
alex
Reply With Quote
  #8  
Old 24-09-2018, 11:10 AM
sil's Avatar
sil (Steve)
Not even a speck of dust

sil is offline
 
Join Date: Jun 2012
Location: Canberra
Posts: 1,474
you're headed in the right direction, but your workflow is still uniform to the frame, look into tutorials on enhancing nebulosity. I recall an awesome video by a german i think called Gerard? Which results in giving you a method to reduce those stars and bringing out their colours as well as being able to bring the nebulosity forward more (brighter/more contrast) without touching the stars. You workflow is never set or good, there is always something to improve.

I think youre third pic is a different colour profile too, i'm seeing artifacts typical of that. try making sure to convert to srgb before you send online.
Reply With Quote
  #9  
Old 24-09-2018, 12:18 PM
RickS's Avatar
RickS (Rick)
PI cult recruiter

RickS is offline
 
Join Date: Apr 2010
Location: Brisbane
Posts: 10,584
Quote:
Originally Posted by Gavin1234 View Post
I’m a little disappointed at the lack of improvement in my image with the amount of data I’ve been able to add.
Like others, I can see an improvement. You'll notice this more in the faint areas. The bright stuff gets to a high SNR with a small number of subs and then doesn't improve much more as you add data.

Quote:
Originally Posted by Gavin1234 View Post
So using PixInsight I used blink to take out all the crap then I used subframe selector and took out the worst remainders based on FWHM then did the same again using SNRWeight. By the end of the time I stack the lights I’m left with 2/3rds of the originals subs i.e. 150 Dow to 100.
You don't need to filter out the low SNRWeight subs unless they are visibly cloud affected or otherwise damaged. By default, ImageIntegration will do noise weighting and give low weights to these subs but you'll still get some benefit from them.

Are you optimising the rejection settings when you integrate or just using the defaults? That might be a fairly simple way to get some more out of your data.

Quote:
Originally Posted by sil View Post
I recall an awesome video by a german i think called Gerard?
That's Gerald Wechselberger, aka "oldwexi". He has some good video tutorials on PixelMath and other topics: http://www.werbeagentur.org/oldwexi/...ixInsight.html

Cheers,
Rick.
Reply With Quote
  #10  
Old 25-09-2018, 08:33 AM
rcheshire's Avatar
rcheshire (Rowland)
Registered User

rcheshire is offline
 
Join Date: Apr 2010
Location: Geelong
Posts: 2,617
IMHO #2 is the right balance, showing lots of improvement and clean calibration. The others are a little overdone. The presence of the brighter stars is just right and the smoothness of the darker regions is a good indication. Nice job...

@multiweb - yes a little more saturation would make it.

Last edited by rcheshire; 25-09-2018 at 02:23 PM.
Reply With Quote
  #11  
Old 25-09-2018, 09:20 AM
multiweb's Avatar
multiweb (Marc)
ze frogginator

multiweb is offline
 
Join Date: Oct 2007
Location: Sydney
Posts: 22,060
#2 with the color from #3. There is an improvement in your pics both in depth and colour balance.
Reply With Quote
Reply

Bookmarks

Thread Tools
Rate This Thread
Rate This Thread:

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off

Forum Jump


All times are GMT +10. The time is now 09:26 PM.

Powered by vBulletin Version 3.8.7 | Copyright ©2000 - 2024, Jelsoft Enterprises Ltd.
Advertisement
Testar
Advertisement
Bintel
Advertisement