Jump to content


Photo

IC 434, Barnard 33 - Horsehead Nebula

  • Please log in to reply
40 replies to this topic

#26 fco_star

fco_star

    Surveyor 1

  • *****
  • Posts: 1771
  • Joined: 13 Dec 2010
  • Loc: Texas, Midland

Posted 27 November 2012 - 07:25 AM

like always your pictures are really good...

#27 Ivo Jager

Ivo Jager

    Vendor ( Star Tools )

  • -----
  • Vendors
  • Posts: 155
  • Joined: 19 Mar 2011
  • Loc: Melbourne, Australia

Posted 28 November 2012 - 02:46 AM

That's a cracker of an image!

I find one thing puzzling/intriguing though in your (really good and thorough) processing document. You seem to perform deconvolution on data that is already stretched globally, modified locally, as well as smoothed.
Mathematically applying deconvolution at that point does not make a whole lot of sense, as the point spread function that needs to be recovered will now be varying all over the place. Wouldn't you get much better (and mathematically correct) results by applying deconvolution prior to stretching as soon as you load the data?

#28 mmalik

mmalik

    Fly Me to the Moon

  • *****
  • Posts: 5390
  • Joined: 13 Jan 2012
  • Loc: USA

Posted 28 November 2012 - 03:07 AM

Thanks Carlo/Francisco/Ivo.

Ivo, I'll defer the complete answer for Mike Unsold (writer of ImagesPlus) to respond; in the meantime here... is one example where Mike U. is using similar chronology. Not sure if you have read this... thread through its entirety, I came up with this routine as an end result of that dialogue. Let’s wait for Mike U. to chime in. Your feedback is always welcome and greatly appreciated; keep it coming. Regards

#29 Mike Unsold

Mike Unsold

    Vendor, MLUnsold - ImagesPlus

  • *****
  • Vendors
  • Posts: 245
  • Joined: 21 May 2009

Posted 28 November 2012 - 07:39 AM

That's a cracker of an image!

I find one thing puzzling/intriguing though in your (really good and thorough) processing document. You seem to perform deconvolution on data that is already stretched globally, modified locally, as well as smoothed.
Mathematically applying deconvolution at that point does not make a whole lot of sense, as the point spread function that needs to be recovered will now be varying all over the place. Wouldn't you get much better (and mathematically correct) results by applying deconvolution prior to stretching as soon as you load the data?


You can applying deconvolution to a well calibrated stacked image before the initial stretch. If you are working with a color camera I would suggest creating a luminance frame from the stacked color image and applying deconvolution to it rather than the color image. Use the deconvolved and stretched luminance image as the the luminance layer for the stacked color image.

Also use a feathered luminance range perhaps with a mask to target just the background, brighter areas with detail, stars, or hue range when using smooth, sharpen, or color adjustment tools. Not much reason to smooth an area only to sharpen it later.

Mike M has posted a very nice Horse Head image that has received many very positive comments. People should experiment with different images and decide what works best for them.

#30 mmalik

mmalik

    Fly Me to the Moon

  • *****
  • Posts: 5390
  • Joined: 13 Jan 2012
  • Loc: USA

Posted 30 November 2012 - 01:10 AM

Thanks Mike U/Ivo for the feedback.

#31 mrturtle

mrturtle

    Mariner 2

  • -----
  • Posts: 248
  • Joined: 06 Aug 2010

Posted 30 November 2012 - 09:09 AM

wow......

#32 bigdeal

bigdeal

    Messenger

  • -----
  • Posts: 446
  • Joined: 26 Feb 2012
  • Loc: Waxhaw, NC

Posted 01 December 2012 - 10:14 AM

The contrast and details around "The Head" are remarkable. Excellent job on the processing and thanks for posting. That's a wall hanger for sure...


+1

Really amazing...

#33 mmalik

mmalik

    Fly Me to the Moon

  • *****
  • Posts: 5390
  • Joined: 13 Jan 2012
  • Loc: USA

Posted 01 December 2012 - 04:19 PM

Thanks mrturtle/Curt.

Direct link... to high res. 602KB JPG download.

Feel free to create/share your own renditions with unprocessed FIT provided in the first post.

#34 Mike7Mak

Mike7Mak

    Apollo

  • *****
  • Posts: 1277
  • Joined: 07 Dec 2011
  • Loc: New York

Posted 02 December 2012 - 02:54 AM

I played with the jpeg a bit in StarTools. Cropped off the left edge a bit, tightened the stars, removed the streak with 'Heal' and got a little more definition in the foreground dust with the 'Filter>nudge' tool.

Nice image, thanks for the processing practice. I plan to try for the HH soon myself.

Attached Files



#35 mmalik

mmalik

    Fly Me to the Moon

  • *****
  • Posts: 5390
  • Joined: 13 Jan 2012
  • Loc: USA

Posted 03 December 2012 - 04:58 AM

Welcome Mike; I think you would get much better results if you processed the FIT... instead of the JPG. Regards

#36 fco_star

fco_star

    Surveyor 1

  • *****
  • Posts: 1771
  • Joined: 13 Dec 2010
  • Loc: Texas, Midland

Posted 03 December 2012 - 01:06 PM

mmalik, Here is my rendition on your excellent data, thank you for letting me practice with it. Do this data was taken in a dark site?
I use a software I found called NASA FITS LIBERATOR 3 to change from .FIT to .TIFF I almost pass 3 hours trying to learn how to use the action and script with photoshop because it works with 3 different layers in gray scale RGB, first time I do something like this but I really enjoy it at the end. I try to make it look smooth like yours but I couldn't, any tip I can get from you ...thank you
Hope you like it

Attached Files



#37 mmalik

mmalik

    Fly Me to the Moon

  • *****
  • Posts: 5390
  • Joined: 13 Jan 2012
  • Loc: USA

Posted 03 December 2012 - 05:06 PM

Francisco, image was taken from suburban red zone, but on a clear/cold night with high ISO; IDAS LPS probably is the biggest secret. Smoothing steps for this particular image in terms of IP/PS are Multiresolution Smooth-Sharpen [IP], Smoothing (Standard Smoothing and Noise Reduction) [IP], and Noise Reduction [PS]. Details in this... doc. Other alternative is lesser use of sharpening; sharpening introduces some granularity to the image. I'll let experts chime in as well. Thanks for trying. Regards

#38 mmalik

mmalik

    Fly Me to the Moon

  • *****
  • Posts: 5390
  • Joined: 13 Jan 2012
  • Loc: USA

Posted 26 May 2013 - 06:59 AM

Yet another rendering of Horsehead Nebula; more of a PixInsight learning exercise. But I did go beyond PixInsight to include some aspects of ImagesPlus and Photoshop processing which I think helped. Comments/critique welcome.

Following is the complete workflow; don't be over-whelmed by it since most of it is well documented here.... Regards

PixInsight:
1. RAW File Processing & Calibration [Linear data]
2. Alignment (Registration), Integration and Crop [Linear data]
3. DBE (Gradient Removal) [Linear data]
4. SKIPPED---Color Calibration [Optional] [Linear data]
5. ATrous Wavelet Transform (Noise Reduction) [Linear data]

---Linear/NON-Linear demark---

6. Histogram Transformation
7. SCNR (Noise Reduction)
8. HDR Multiscale Transform [IMPORTANT]
9. SKIPPED---Local Histogram Equalization (Contrast) [Optional]
10. ACDNR (Noise Reduction)
11. SKIPPED---Curves Transformation (Saturation)
Saved as 16-bit unsigned TIFF for ImagesPlus/Photoshop processing

ImagesPlus/Photoshop:
12. Geometric Transform/Scale (ImagesPlus)
13. Smoothing & Noise Reduction (ImagesPlus)
14. ProDigital ‘Deep Space NR’ and HLVG (Photoshop)
15. Saturation (Block) and Noise Reduction (Photoshop)
16. Star Size & Halo Reduction, and Saturation (ImagesPlus)
17. Levels, Curves, Contrast, Vibrance and Color Balance (Photoshop)
Saved as JPG for posting

Larger JPG here...

Note: PixInsight-CR_Integration.fit for practice available here.... Start with DBE if you use provided FIT.

Attached Files



#39 mmalik

mmalik

    Fly Me to the Moon

  • *****
  • Posts: 5390
  • Joined: 13 Jan 2012
  • Loc: USA

Posted 26 May 2013 - 10:51 PM

PixInsight-CR_Integration.FIT for practice available here.... Start with DBE if you use provided FIT. Thx

#40 Phil Hosey

Phil Hosey

    Surveyor 1

  • *****
  • Posts: 1634
  • Joined: 29 Oct 2008
  • Loc: LaGrange, GA

Posted 27 May 2013 - 08:16 AM

Mike,
Nice image, especially like the detail around the edges of the horsehead itself. I've been following your PI thread as well, good stuff!

#41 mmalik

mmalik

    Fly Me to the Moon

  • *****
  • Posts: 5390
  • Joined: 13 Jan 2012
  • Loc: USA

Posted 01 June 2013 - 04:28 AM

Thanks Phil.

PixInsight learning series... I keep using the same image mainly for direct comparison of various processing attempts and processing variations. Note: Last time I had skipped PI color calibration; hence the color differences. This iteration is pretty much a full dress rehearsal of combined PixInsight/ImagesPlus/Photoshop processing whose details are as follows:

PixInsight:
0. DSLR_RAW Format Preferences (ONE/FIRST time setting) [Linear data]
1. SKIPPED---Calibration [Linear data]
2. Batch Debayer [Linear data]
3. Alignment (Registration), Integration and Crop [Linear data]
4. DBE (Gradient Removal) [Linear data]
5. Background Neutralization [Linear data]
6. Color Calibration [Linear data]
7. SCNR (Noise Reduction) [Linear data]
8. ATrous Wavelet Transform (Noise Reduction) [Linear data]

---Linear/NON-Linear demark---

9. Histogram Transformation
10.HDR Multiscale Transform [IMPORTANT]
11.SKIPPED---Local Histogram Equalization (Contrast) [Optional]
12.ACDNR (Noise Reduction)
13.SKIPPED---Curves Transformation (Saturation)
Saved as 16-bit unsigned TIFF for ImagesPlus/Photoshop processing

ImagesPlus/Photoshop:
14. Geometric Transform/Scale [0.5] (ImagesPlus)
15. Smoothing & Noise Reduction (ImagesPlus)
16. ProDigital ‘Deep Space NR’ and HLVG (Photoshop)
17. Saturation (Block: 1 pass) and Noise Reduction (Photoshop)
18. Star Size & Halo Reduction, and Saturation (ImagesPlus)
19. Levels, Curves, Contrast, brightness, and Vibrance (Photoshop)
Saved as JPG for posting

Larger version here...

Updated processing instructions here...

PixInsight-CR_Integration.fit for practice available here.... Start with DBE if you use the provided FIT. Thx

Attached Files








Cloudy Nights LLC
Cloudy Nights Sponsor: Astronomics