Author Topic: NGC 6992  (Read 8789 times)

Offline Nocturnal

  • PixInsight Jedi Council Member
  • *******
  • Posts: 2727
    • http://www.carpephoton.com
Re: NGC 6992
« Reply #15 on: 2011 September 27 13:26:43 »
Julian,

looks like you're trying to understand masking and deconv at the same time. I recommend you browse Harry's tutorials and see if they get you started.
Best,

    Sander
---
Edge HD 1100
QHY-8 for imaging, IMG0H mono for guiding, video cameras for occulations
ASI224, QHY5L-IIc
HyperStar3
WO-M110ED+FR-III/TRF-2008
Takahashi EM-400
PIxInsight, DeepSkyStacker, PHD, Nebulosity

Offline Jules

  • PixInsight Guru
  • ****
  • Posts: 513
Re: NGC 6992
« Reply #16 on: 2011 September 30 09:53:13 »
Hi

I am relatively happy with star mask. My only query was if a star mask could be applied to a linear image.

My biggest problem is with deconvolution, sharpening the detail you want to sharpen but not impacting on the stars. So far it is not working very well for me. I want to keep using deconvolution as it is such a powerful tool.

Regards

Julian

Offline Juan Conejero

  • PTeam Member
  • PixInsight Jedi Grand Master
  • ********
  • Posts: 7111
    • http://pixinsight.com/
Re: NGC 6992
« Reply #17 on: 2011 September 30 10:42:03 »
StarMask can be applied to a nonlinear image to create a mask that can be used with linear and nonlinear images. In fact, the StarMask process already includes several parameters to delinearize the image: the shadows, midtones and highlights parameters work just like a histogram transformation applied as an initial step prior to mask generation.

With some practice, StarMask can be used to create a mask including virtually *all* of the stars in an image. In difficult cases you can apply HDRWaveletTransform to a stretched version of the image before StarMask. Rogelio Bernal has recently published a tutorial where he demonstrates this technique. The HDRWT+StarMask combo is a killer combination. Using it we have been able to create a mask of the M42 region including all the stars within the core of M42 and M43.

Another common source of confusion is deconvolution and its deringing support image. A deringing support is not a mask. It works by driving a special routine that limits growth of ringing artifacts at each deconvolution iteration, but it doesn't work as a mask because it does not perform a blend of two images. Fortunately, a deringing support for deconvolution is usually very easy to build: you just need to include relatively bright stars. Jordi Gallego has a nice presentation on his website in PDF format:

http://astrosurf.com/jordigallego/articles/Pixinsight_avanzado_JGallego.pdf

This document is in Spanish but the examples relevant to this discussion are very easy to understand. Jump to slide #  15 and you'll see the typical ringing artifacts that one may expect when applying deconvolution to a linear image. In slide 16 you can see the effect of using our global deringing algorithm. In slides 18, 19 and 20 Jordi puts a very nice example of global + local deringing using a deringing support image (a star mask including only bright stars). In slides 21, 22 and 23 you can see a comparison of the result before deconvolution, with deconvolution + global deringing, and deconvolution + global + local deringing.

Note that deringing with a deringing support can be used along with a regular mask, which is often necessary for deconvolution to protect low SNR areas. However, a star mask (working as a mask) is almost never necessary for deconvolution.
Juan Conejero
PixInsight Development Team
http://pixinsight.com/

Offline Jules

  • PixInsight Guru
  • ****
  • Posts: 513
Re: NGC 6992
« Reply #18 on: 2011 October 04 06:06:48 »
Juan

Thank for taking the time to answer this post. I have worked my way through Jordi's Deconvolution article and found the application of star masks on a linear image through deconvolution fairly straight forward.

I deconvolved each linear integrated image, Ha and OIII prior to blending to make the third one.

Have a look at the previous image posted and compare it with this one.

Please consider it is only a small 236kb image

Regards

Julian


Offline Jules

  • PixInsight Guru
  • ****
  • Posts: 513
Re: NGC 6992
« Reply #19 on: 2011 October 09 14:36:53 »
Hi

I have translated pages 19 to 24 of Jordi Gallego's presentation on Deconvolution that Juan pointed me to. I have found this very useful.

I love 'mascara de estrellas', what a great language!

It needs to be read in conjuction with the presentation, see above in Juan's post.

Regards

Julian