Jump to content

Banner.jpg.b89429c566825f6ab32bcafbada449c9.jpg

NGC1333 amidst a sea of dust

Barry-Wilson

By Barry-Wilson
in Imaging - Deep Sky

Recommended Posts

Barry-Wilson

Barry-Wilson

Posted
Posted

Taken over the last two nights, this LRGB image shows the beautiful and complex blue reflection nebula surrounded by the molecular cloud within Perseus.

I've had to push the data quite hard to bring out the browny molecular clouds - perhaps too far - and the dusty halos of the larger stars need a little more work.  I'm hoping for clear skies to collect some more data tonight.

Lum - 18 x 600s bin 1

RGB - 12 x 300s bin 1 for each channel

6 hours total integration

Tak FSQ85 at F3.9, Avalon Linear FR mount, QSI683-wsg, Astrodon filters

Captured with SGP and processed in PI.

I've not processed many dusty LRGB images - plenty still to learn.

Thanks for looking & clear skies!  Barry

post-28392-0-25486900-1444939237_thumb.j

Link to comment
Share on other sites
  • Replies 38
  • Created
  • Last Reply
Barry-Wilson

Barry-Wilson

Posted
  • Author
Posted

Great result!

Jeremy

Thank you Jeremy.

Well I won't bother with this region now as too many people are imaging it :grin:

Nice work Barry, perhaps stretch it a little less until you have collected more data, it will still be a very pleasing image.

Thanks Stuart.  You're right and I'm in the middle of a revised processing.  This target has been on my list for sometime and not often seen . . . be good to see yours :icon_salut:

That is really beautiful Barry.  Not heard of this one before.

Carole

Thanks Carole, much appreciated.

Nice one Barry! With a tad more data it shall become poster material :D

Thanks silos.  I'm sitting here hoping it will clear.  I'm in a wide river valley and it does attract mist, despite what Clear Outside tells me.

Link to comment
Share on other sites
Barry-Wilson

Barry-Wilson

Posted
  • Author
Posted

Excellent work Barry...a fine image. With the extra data this will I'm sure be top drawer. Well done.

Pete.

Thanks Pete.  No luck with the weather last night despite the forecast, just low cloud.

Nice one Barry - One that has been on my list for a few years now and never even attempted! Looks great!!

Thanks Sara.  Can't tempt you to do more LRGB?  Glad to read your cable issue is resolved.

very well done, its not easy to bring out the dusty regions.  i be interested in your processing workflow on how you did it, if your willing to share.

Paul J

 Will do Paul.  I'll compose a reply when I'm from work this evening and post here for the benefit of PI users.

Yes, that it very nice indeed.

Chris

Thanks Chris - the nebula is like a little beacon shining away, quite entrancing.

Link to comment
Share on other sites
Barry-Wilson

Barry-Wilson

Posted
  • Author
Posted

Fantastic Barry!

Thanks John, there's so much going on in this area, it makes it easier to get a cracking image.

Cable issue not resolved .............. :( In fact nothing is resolved, I'm more at the giving up stage if I'm honest :(

Sorry to hear that Sara.  From your last post on the SGP forum I thought it was resolved.

brill image Bary

Much appreciated Chris.

Link to comment
Share on other sites
Barry-Wilson

Barry-Wilson

Posted
  • Author
Posted

OK, here goes my PI workflow for NGC1333 – happy to take supplementary questions too.

  1. Batch Pre-processing script, using cosmetic correction along with darks, bias, flats.
  2. Image Integration to stack each channel, winzorised sigma clipping for rejection.
  3. Dynamic crop.
  4. DBE to each channel – carefully selected points, hand placed because there is so much large-scale structure.

RGB

  1. Channel Combination.
  2. Repaired HSV separation to improve star core colour.
  3. Channel combination in HSV mode – see Pixinsight Resources website for Bob Andersen’s tutorial.
  4. Background Neutralisation & Colour Calibration.
  5. Masked Stretch, followed by a couple of small iterative Histogram Transformations.
  6. Ready for LRGB combination.

Lum

  1. Create copy Lum.
  2. Apply STF strech, clip black and set as Lum Mask.
  3. Create Star Mask, set smoothness to 8, tick aggregate and binarise.
  4. Dynamic PSF, then Deconvolution – only for 5 iterations.
  5. Create Copy.
  6. Masked Stretch to one copy.
  7. Histogram Transformation to the other copy with iterative stretches, perhaps 5 or 6.
  8. I then inspect which stretch I prefer.  Sometimes I will combine the two stretched non-Linear frames either using PixelMath or the Blend Script (downloadable from the PI forum).  The Blend script is great, created by Rick Stevenson and mimics to some degree the PS blend modes, eg lighten, darken, colour burn, vivid light etc.
  9. Ready for LRGB combination.

Non-Linear

  1. LRGB combine.  I do not boost saturation nor apply chrominance noise reduction.
  2. Apply SCNR with Green and Average Neutral set to 100%.
  3. Using Histogram Transform adjust histogram in individual channels to align the left handside of the curves.
  4. I then applied the Lum mask to protect the background, and opened Local Histogram Equilisation.  (I didn’t use High Dynamic Range Multiscale Transform on this target).
  5. I set the Kernel Radius to 312 and the amount to 0.3, apply.  Repeat with settings of Kernel Radius 194 amount 0.25, Kernal Radius 28 amount 0.17.
  6. Leaving the Lum Mask in place, use Multiscale Linear Tranform to sharpen.  Set bias for the first level to 0.05, for the second level to 0.03, the third to 0.01.
  7. Create another star mask, this time I reduced Scale to 3, Large Scale to 1, Compensation to 1, Slide Noise Threshold to 2.5, smoothness to 6.  This mask is for the smaller stars.
  8. Using PixelMath combine the star mask using Max(StarMask1, StarMask2).
  9. Subtract the Combined Star Mask from the Lum Mask in PixelMath to give a Nebula Mask.  Apply this mask to the image.
  10. Using Colour Saturation boost the colour to your liking.
  11. Apply the Combined Star Mask to protect the background but reveal the stars.
  12. Use Morphological Transformation to reduice the stars – in two stages.  Using Erosion Minimum, amount to 0.25 with 1 iteration, size 5, set to circle.
  13. Second stage, set the Operator to Morphological Selection and set amount to 0.9, selction 0.9, iterations 3.
  14. I then used Large Scale Structure Enhance script to separate the wavelet layers.  I stretched the large scale (LS) image to enhance the background dust.  Using PixelMath recombine the small scale (SS) with the LS; LS+SS, with Rescale ticked.
  15. Blend this dust enhanced version with the LRGB using Blend mode Lighten.
  16. Using the Blend script blend the Background Enhanced image with the LRGB with Enhanced as the Top layer, Blend mode Lighten and opacity set at 80%.
  17. Create a clone of the new LRGB.  Using the Nebula Mask – with some adjustment from Clone Stamp to protect the large star halos – apply to the LRGB image and using Inter Channel Curves with reference channel set to Hue, accentuate the colour of the dust to suit.
  18. Use Blend Script o once again blend the hue enhanced dust with the LRGB.
  19. I then used Curves with the Lum mask protecting the background to boost contrast a little.
  20. I used TGVDenoise for NR – Lum Mask as Local Support, reducing smoothness to 1.5, Edge Protection to 6.00 and -4, iterations at 350 (first test on a preview on 100 iterations).
  21. Darks Structure Enhance script set at 30%.
  22. Final curves to adjust.

Phew!

Link to comment
Share on other sites
gnomus

gnomus

Posted
Posted

Barry - one question. I have looked for this blend script a couple of times now (you mentioned it to me in an Astrobin comment). I cannot find it anywhere. I've found a 'Blend Script' by 'Mike Read' on the PI forum, but nothing by RickS. Do you happen to have the link for it?

Steve

Link to comment
Share on other sites
Barry-Wilson

Barry-Wilson

Posted
  • Author
Posted

Steve

Public link to my dropbox file with the script, https://dl.dropboxusercontent.com/u/32050952/Blend%20Eng.js

I remember I had seen comments from Rick Stevenson on the script and thought he wrote it, apologies Steve if I've sent you on a wild 'google' chase!  Just checked and it is Mike Read and the link above is the same as the PI one here, http://pixinsight.com/forum/index.php?topic=3902.0.

Link to comment
Share on other sites

Archived

This topic is now archived and is closed to further replies.

Go to topic listing

  • Recently Browsing   0 members

    • No registered users viewing this page.
×
×
  • Create New...

Important Information

We have placed cookies on your device to help make this website better. You can adjust your cookie settings, otherwise we'll assume you're okay to continue. By using this site, you agree to our Terms of Use.