Jump to content

  •  

CNers have asked about a donation box for Cloudy Nights over the years, so here you go. Donation is not required by any means, so please enjoy your stay.

Photo

PixInsight Arcsinh Stretch

  • Please log in to reply
57 replies to this topic

#51 sharkmelley

sharkmelley

    Apollo

  • -----
  • topic starter
  • Posts: 1121
  • Joined: 19 Feb 2013

Posted 22 March 2017 - 01:56 AM

 

Here it looks like you are calculating luminance by giving equal weight to each color channel. By default PixInsight does not weight the channels equally when extracting luminance, as discussed in documentation of RGB Working Spaces (such as  https://pixinsight.c.../why_rgbws.html). Would you get better results by using the coefficients defined in the RGB Working Space?

 

 Hi John,

 

Thanks for this observation. I've never really played with RGBWorkingSpace, so I've taken the time to research what it does and to incorporate it into a test version of the code.  Using the RGBWS luminance coefficients when calculating the luminance does have some subtle but useful effects on the outcome of the stretch.  For instance it can be used make H-alpha pop out more from the background during the stretch.  It doesn't change the colour balance of those individual pixels but makes pixels dominated by that colour appear brighter in the stretched image.

 

I'll certainly put this in the next version so the more sophisticated user can take full advantage of it but I won't get a chance to do this for a little while.

 

Mark


Edited by sharkmelley, 22 March 2017 - 02:26 AM.

  • johnpane and bmhjr like this

#52 JukkaP

JukkaP

    Mariner 2

  • -----
  • Posts: 262
  • Joined: 14 Aug 2016

Posted 22 March 2017 - 06:04 AM


Thanks for looking my data. I had to give this a thought and read couple times. The result you got whit multiple iterations was good. Dont really understand why it works differently. And when it would not be good to do it in iterations to all images if it preserves the fainter contrast in images. Why not do it in many iterations allways as we do masked strech?

It really depends on the distribution of data in the image. The effect of applying the Arcsinh stretch in a number of iterations is to compress the highlights more but if you keep the total stretch the same (in the above example 144) then the gradient at the bottom end of the response curve will be identical.

Here's a graph that illustrates the response curve of those 3 stretches:
Arcsinh1.JPG

Changing the scale of the graph's x-axis shows that the gradient at the bottom end in each case is the same (i.e. 144):
Arcsinh2.JPG

Some images will respond very well to having their highlights extremely compressed - especially if the highlights only contain stars. Your image was an example of that.

Mark
The 3.5 strech looks like normal histogram transformation. While preserving star colour. Looks to me that this multiple iteration aproach should be ingluded to the aplication. You might get the code from masked strech, if you are interested in that.

I did some testing and when you adjust the background slider litle bit from the suggested one, you will have some room in your histograms left side. I did see some clipping using the automated background value.

Enyway great aplication and thanks for sharing!
  • sharkmelley likes this

#53 federico bonino

federico bonino

    Vostok 1

  • -----
  • Posts: 123
  • Joined: 21 Jun 2012
  • Loc: Argentina

Posted 24 March 2017 - 07:06 PM

Mark Good tool !

Thanks very much to share.

I Must to check it asap !!

Federico


Edited by federico bonino, 25 March 2017 - 12:27 PM.


#54 jammeymc

jammeymc

    Vostok 1

  • -----
  • Posts: 143
  • Joined: 19 Mar 2016

Posted 25 March 2017 - 10:18 AM

Very nice - my new favorite stretching tool. Thank you for the all you did to provide this.


  • sharkmelley likes this

#55 Diane Miller

Diane Miller

    Mariner 2

  • -----
  • Posts: 230
  • Joined: 28 Sep 2016

Posted 25 March 2017 - 11:30 AM

I've been following this discussion sporadically and wondering if there is a summary tutorial for those of us who are new at this sort of thing.  I would also find it helpful to have information about the sort of image for which it is more advantageous than the usual PI process.


  • Islandwatcher likes this

#56 Islandwatcher

Islandwatcher

    Viking 1

  • -----
  • Posts: 528
  • Joined: 12 Nov 2009
  • Loc: Wales,U.k.

Posted Yesterday, 08:02 AM

I am very interested in this topic but am fairly new to Pixinsight , have a Macbook that I use for processing and was after a few tips how to actually get this process working / installed into my version of Pixinsight...

 

Thanks

 

Steve .



#57 sharkmelley

sharkmelley

    Apollo

  • -----
  • topic starter
  • Posts: 1121
  • Joined: 19 Feb 2013

Posted Yesterday, 05:04 PM

Below are the instructions for installing it under Windows.  I'm guessing guess that for the Mac you simply substitute Arcsinh-pxm.dylib for Arcsinh-pxm.dll

 

Installation instructions for Windows:

  • Put Arcsinh-pxm.dll in a chosen folder.  For Windows this might be C:/Program Files/PixInsight/bin but you can choose anywhere convenient to you.
  • In PixInsight navigate the Menus to: Process->Modules->Install Modules
  • Use the "Browse..." button on the top right of the "Install Menus" dialog to navigate to your chosen folder and then hit the "Select Folder" button.  The name of your folder will appear as the "Search Directory".  Now hit the "Search" button.  If more than one module appears you may want to to double-click the irrelevant ones, leaving a green tick only against Arcsinh-pxm.dll
  • Hit the "Install" button.
  • The Arcsinh process will now be available from the Process Explorer window or the Process menu in the usual way.

Mark



#58 Islandwatcher

Islandwatcher

    Viking 1

  • -----
  • Posts: 528
  • Joined: 12 Nov 2009
  • Loc: Wales,U.k.

Posted Today, 03:11 AM

Hi Mark ,

 

OK thanks...that seems to have done the trick...thanks for the extra clarification . I've had a quick play and it really does preserve the colours very well ! Many thanks to you...and John for this tool !

 

Steve .




CNers have asked about a donation box for Cloudy Nights over the years, so here you go. Donation is not required by any means, so please enjoy your stay.







Cloudy Nights LLC
Cloudy Nights Sponsor: Astronomics