Pompey Monkey Posted August 30 Share Posted August 30 Sometimes you just can't win. lol I wonder if I could do something clever with one of the statistical rejection methods in PI? Any suggestions? 1 Link to comment Share on other sites More sharing options...
Vroobel Posted August 30 Share Posted August 30 Wow, this might be funny if it wasn't a problem. 🤔 Did you try stars removal and any form of clone on the final image? 1 Link to comment Share on other sites More sharing options...
Elp Posted August 30 Share Posted August 30 Make sure the bug(ger) setting is set to off... 1 Link to comment Share on other sites More sharing options...
Elp Posted August 30 Share Posted August 30 Realistically if doing it manually it's a simple case of using one "clean image", copying the region you want to copy, then layering it over and blending it with the stacked image. 1 Link to comment Share on other sites More sharing options...
ONIKKINEN Posted August 30 Share Posted August 30 The GESDT (Generalized extreme something something) clipping algorithm might work. Depends on how many subs there are and how large the movement of this thing in the image is between subs. If it stays in the stack, should be a simple fix in Photoshop with content aware fill. Not sure what the right play for that would be in PI though. 2 Link to comment Share on other sites More sharing options...
Pompey Monkey Posted August 30 Author Share Posted August 30 (edited) My friend suggested that it might be a reflection from a bright star moving around due to the dithering process. I've (just about) fixed it using a mask created with the GAME script, some delicate histogram transformations in the blue channel, and a double dose of SCNR to erode the resulting green sky. Not perfect, but a lot better Note that the following are just capture of my unstretched RGB. Before and after: Edited August 30 by Pompey Monkey 2 Link to comment Share on other sites More sharing options...
Pompey Monkey Posted August 30 Author Share Posted August 30 50 minutes ago, ONIKKINEN said: The GESDT (Generalized extreme something something) clipping algorithm might work. Depends on how many subs there are and how large the movement of this thing in the image is between subs. If it stays in the stack, should be a simple fix in Photoshop with content aware fill. Not sure what the right play for that would be in PI though. I used GESDT something in the stacking. However, there are only 16 subs in the blue channel. So maybe one of the more aggressive rejection algorithms might do a better job. You've got me thinking, thanks Link to comment Share on other sites More sharing options...
Mr Spock Posted August 30 Share Posted August 30 I thought I had a couple of nasty looking bugs in the kitchen a while back. When I put my glasses on they turned out to be a couple of sultanas... 3 Link to comment Share on other sites More sharing options...
ONIKKINEN Posted August 30 Share Posted August 30 23 minutes ago, Pompey Monkey said: I used GESDT something in the stacking. However, there are only 16 subs in the blue channel. So maybe one of the more aggressive rejection algorithms might do a better job. You've got me thinking, thanks GESDT prefers stacks of a hundred or more subs to do its magic if i recall correctly. How about normal sigma clipping with a very aggressive 2 or even 1.5 hot sigma? You could then clone stamp tool just the affected area to the GESDT normal stack and the trickery might not be too noticeable, if all this needs to be done in PI. 1 Link to comment Share on other sites More sharing options...
Pompey Monkey Posted August 30 Author Share Posted August 30 58 minutes ago, ONIKKINEN said: GESDT prefers stacks of a hundred or more subs to do its magic if i recall correctly. How about normal sigma clipping with a very aggressive 2 or even 1.5 hot sigma? You could then clone stamp tool just the affected area to the GESDT normal stack and the trickery might not be too noticeable, if all this needs to be done in PI. I have since tried the sima clipping very aggressively, down to about 0.8 sigma, It worked a bit, but not before destroying the overall image. I see from other images that there's no obvious blue signal in that area. I could blend the green or lum channels in the affected area, perhaps. And yes, I never got on with PS, so it's PI for dinner every time! Link to comment Share on other sites More sharing options...
ollypenrice Posted August 31 Share Posted August 31 On 30/08/2024 at 18:43, Mr Spock said: I thought I had a couple of nasty looking bugs in the kitchen a while back. When I put my glasses on they turned out to be a couple of sultanas... Good to see you raisin this matter! On 30/08/2024 at 20:05, Pompey Monkey said: I have since tried the sima clipping very aggressively, down to about 0.8 sigma, It worked a bit, but not before destroying the overall image. I see from other images that there's no obvious blue signal in that area. I could blend the green or lum channels in the affected area, perhaps. And yes, I never got on with PS, so it's PI for dinner every time! With Ps you don't have to worry about the whole image. All you have to do is fix the bit in question and blend it in as a layer with the standard one. Once de-starred, content aware fill would be perfect, too, as Oskari said. I'm always mystified by not getting on with Ps: I feel that working in PI is like trying to go rock climbing wearing boxing gloves and roller skates. (As a unicyclist you would doubtless find this a doddle, though!) lly 2 Link to comment Share on other sites More sharing options...
wimvb Posted September 1 Share Posted September 1 (edited) Pixinsight has large scale pixel rejection. I've never needed it, but it might be worthwhile to check it out. Since there is no real detail in that section of the image, the easiest solution is the one you used. Create a mask and use scnr on blue, or simply desaturate the blue to get rid of the artefact. Edited September 1 by wimvb 1 Link to comment Share on other sites More sharing options...
Recommended Posts
Create an account or sign in to comment
You need to be a member in order to leave a comment
Create an account
Sign up for a new account in our community. It's easy!
Register a new accountSign in
Already have an account? Sign in here.
Sign In Now