Reducing the size/intensity of one star in an image

Comments

• BlockHead

  January 2021

  Lots of ways to do it... 

  Few of them are "passable." 

  None of them could be explained by descriptions easily... they need to be demonstrated.

  Finally, without seeing an image- even more challenging to suggest.

  Alnitak typically has a bright saturated part with a halo of some sort and maybe other scattered light effects (diffractions, other reflections..etc). Reducing the halo is different than reducing the saturated part of the star. 

  I do have a lesson in the queue that is a dramatic way to lesson the brightness/halo around stars like Alnitak. 

  I hope to publish it soon. I am trying to finish up FastTrack for beginners at the moment.

  -the Blockhead
• Bruce Donzanti

  January 2021

  That would be great, Adam.  

  
  

  Here is the image I took and my attempt at reducing the overwhelming brightness Alnitak has on the image.  I used multiple masks of different sizes to reduce it because of the halo vs. the core saturation effect. 

  

  

  IC434_LRGB_v3.jpg

  2021 x 1326 - 988K
• Brent Jaffa

  February 2022

  Adam can the GAME script be modified to help reduce the large reflection halo around Alnitak. For example the latest version can write zeros on a file list for your selective rejection method for a subset of artifacts on images. On the large halo around Alnitak there are several concentric halo intensities. If the GAME script would allow writing a small intensity level rather than a zero on a monochrome image several iterations of the GAME script for varying concentric circles, could be used to easily reduce the halo.

  I hope this makes sense - Brent 

  

  

  BLU_B_clone_clone1.jpg

  4371 x 3315 - 1M
• BlockHead

  February 2022

  Brent,

  
  

  Whoah..that is a tough one.

  Yes, I asked Hartmut to put that functionality into the script.

  
  

  The zeros are really specifically for rejection. This single value is used to signify these pixels need to be rejected.

  
  

  I think what you want to do is use the GAME script in the normal way. You can create a mask and modify the values (from all black/white to any intermediate value). Then you would use Curves transformation or HistogramTransformation to lessen the brightness of the halo. There are some more sophisticated ways of doing this. You can operate on many images at once with an image container (and a process container if you want more than one thing done).

  
  

  By the way... in Bruce's case, the Alnitak probably looks better with less of a halo in the red data? I wonder what happens if you put the red version (in the gray/luminance) of Alnitak in place of the blue. Wouldn't this lessen the bluish scattered light a bit?

  
  

  -the Blockhead
• Brent Jaffa

  February 2022

  Adam,

  Thanks for your suggestions. Its funny that the Green is the worst followed by the blue. The red is almost

  nonexistent. And there is none in the HA. I read where it is a reflection off the back of the filters in the filter wheel. I believe there is a calculation based on the reflection diameter to determine the distance where the reflection is occurring, and it matches the distance to the filter.

  Anyway, I love your instruction both in fundamentals & horizons.

  
  

  Thanks,  Brent

  
  
• Roger Smith

  February 2022

  Bruce, 

  I tackled this same star with similar huge blue halo when I imaged Horsehead. The second time I was successful. 

  I made a Game mask (as big as the blue), carefully centered it, and feathered (convolution) the outer edge. I had to be careful the feathering matches the blue taper profile. The mask protects everything except the star area.  

  I then concentrated on reducing the blue channel until the intensity looked good (not overpowering) with the other colors. 

  Roger
• Bruce Donzanti

  February 2022

  Thanks, Roger.  Greatly appreciate the suggestion.