Any way of removing diffraction star pollution from a photo near a bright star

Dingo Taz

Hello. I am doing a mosaic of witch head and want to extend out to include Rigel (12 panel mosaic with a 533mc pro camera). Rigel is obviously very bright. The panel with Rigel has it to the far right of the panel. The diffraction etc on this panel i just love. 

The surrounding panels though get diffraction spike pollution from Rigel coming through. I asked on the PI forum, it was explained this is probably because my scope (Askar FRA600) has a wider FOV and can probably see Rigel towards its edge of FOV, even though Rigel itself doesn't show up on the photo from the 533mc. But nobody knew of any way to remove the spikes (other than photoshop which I neither want or have). I have shifted the panel to the left of Rigel further to the left to increase distance, but it's got to the point now where there is no overlap for the mosaic to happen and there are still faint marks. 

I did try removing the stars and using the clone tool in PI but I couldn't get it anywhere close to un-noticeable.

I'm considering upgrading to a 2600mc camera which has a much wider FOV (which will result in less panels for all my mosaics). But I'm hoping maybe there's some magic in PI to solve this and remove the diffraction spike pollution.

masterLight_BIN-1_3008x3008_EXPOSURE-120.00s_FILTER-UV-IR-Cut_RGB_PANEL-9_drizzle_1x.xisf is the stacked problem child. Rigel is in Panel 10.

https://1drv.ms/u/s!AqZNofwpcMYnjJkMMKGLA_mdK9AkSg?e=PEMhoS

BlockHead

HI Dingo,

You are correct concerning the diffraction spike can extend for many degrees depending on how deep your data is. This is normal. It is also a fact that this diffraction is every bit as "real" as the rest of the photons that formed the images of the nebulae and everything else. So there isn't a terribly clever way to remove the spike.

When you use a star removal tool... it will usually put the diffraction into the extracted stars frame. In this frame it is easier to kill the excess spike perhaps.  (I can't look at your data today... so this is in principle). This means you would wait to complete the mosaic and then apply star removal. Anyway... something to think about.

-the Blockhead

Dingo Taz

thanks mate.

I can send the other panels if needed, but I think the four there would be enough, rigel (top right edge) and the three surrounding ones. I was able to shift the panel under rigel (i think #6) to not get the spikes in that panel and still mosaic, but panel to the left of rigel I couldn't. These are only test panels to see if it would work, so only about 10 or 15 shots stacked.

I do like the spikes on the Rigel panel itself and would like to keep those, but they don't line up with the spikes in the surrounding shots.

Have mosaiced the 4 panels ("4 panel mosaic rigel diffraction.xisf"):

• spcc on each panel
• Register by coordinates
• Trim (3 on outside edges, 10 on inside edges)
• Photometric Mosaic 

   1. ref panel 10, target panel 6  -> creates "mosaic"

      Defaults (i think) on everything (eg. average, 2% outlier, join 90%) 

  Haven't touched any of the other settings (photometry, gradient, etc).

   2. ref panel 9, target panel 5  -> creates "mosaic 1"

   3. Ref panel "mosaic", target panel "mosaic 1"

Might need better settings on the photometric mosaic, the brightness of rigel is mucking up the panel under it as well.

Whilst the diffraction spikes to the left of rigel are faint, the issue is when I make the witch come out these come out as well. Maybe a mask on the witch is needed to solve that one.

Would really appreciate any comments if you do get a chance to look.

PS I am fairly new to this, working my way through and really enjoying fundamentals thank you.