Page 1 of 1

M45 can I minimize diffraction spikes

Posted: Wed May 06, 2020 6:57 pm
by Troy Galebach
I have some clean M45 data from what I thought was an FSQ-106, but when I processed it, I have what I believe are diffraction spikes. Perhaps it isn't from an FSQ, or perhaps I should not have used AutoDev, should I try and use Dev instead? I'm happy with the image color, but I'd appreciate any ideas on how I might minimize the spikes in this image. I'm new to LRGB processing, and wonder if I'm somehow responsible for these spikes. Any help will be greatly appreciated. If you want to see the image please follow the link, and scroll down to the last image. Thanks!

https://tgalebachmo.net/astronomy/

Regards,
Troy

Re: M45 can I minimize diffraction spikes

Posted: Thu May 07, 2020 9:30 am
by admin
Hi Troy,

Diffraction spikes are usually considred a fact of life, however you can try to heal them out.
Would you be able to post a screenshot/example of the issue you are seeing?

Clear skies!

Re: M45 can I minimize diffraction spikes

Posted: Thu May 07, 2020 11:14 am
by Troy Galebach
admin wrote:Hi Troy,

Diffraction spikes are usually considred a fact of life, however you can try to heal them out.
Would you be able to post a screenshot/example of the issue you are seeing?

Clear skies!
Hi Ivo, I hope all is well with you.
I'd love to post a shot, but the link I left in my message is the only way I know how to post a pic. Did I miss something which would allow me to post on this site?
In case you missed it, here is the link to the image. You have to scroll down to the very last image, it is the one I'm asking about (M45).

https://tgalebachmo.net/astronomy/

Regards,
Troy

Re: M45 can I minimize diffraction spikes

Posted: Fri May 08, 2020 11:38 am
by admin
Thank you Troy. It's a lovely image. Diffraction spikes on these bright stars are usually par for the course in M45 renditions. You really kept the stellar profiles under control well.
What does throw me off a little though, is that the spikes are all oriented differently... :think:

As for dealing with them in post-processing, they only thing I can think of would be to put them in a mask and using the Heal module to heal them out.

Re: M45 can I minimize diffraction spikes

Posted: Fri May 08, 2020 12:25 pm
by Troy Galebach
Thanks for your reply Ivo,
admin wrote:Thank you Troy. It's a lovely image. Diffraction spikes on these bright stars are usually par for the course in M45 renditions. You really kept the stellar profiles under control well.
What does throw me off a little though, is that the spikes are all oriented differently... :think:

As for dealing with them in post-processing, they only thing I can think of would be to put them in a mask and using the Heal module to heal them out.
the spikes were driving me a bit nutty for the same reason. I have never processed LRGB data from Lights, Darks, Flats, and Bias frames. To do so I got a trial license for PixInsight. The data were collected over multiple nights, and I wonder if the left side of the pier/right side are causing the issue. I'm leaning in that direction, but I don't actually know. Combine my lack of knowledge on multi-date imaging, along with never registering/stacking in PI, and I didn't know where to look. I have all the data, and perhaps I can find the time to blink though it, compare dates & times, and see if the different angled spikes are being caused by different pier sides, and dates. I never did that when I processed the data, just wanted to learn how to use PI to register & stack so I could use StarTools to process the data out of PI. God knows after looking at PI that I don't have the desire, IQ, or time left to learn all there is to know about that Program! If I ran Windows, I'd go with DSS, if Siril were better documented, I'd use it.

Stay safe Ivo, and thanks for StarTools.

Regards,
Troy

Re: M45 can I minimize diffraction spikes

Posted: Sat May 09, 2020 6:12 am
by admin
Aha! The multi-date acquisition may have something to do with it. If the orientation of the camera vs object changes, then the diffraction pattern will change as well.
This can be a good thing too, as some stacking rejection algorithms will be able to reject the spikes if they keep changing per frame.

If PI is not your cup of tea, you could try Astro Pixel Processor as well.

Clear skies!

Re: M45 can I minimize diffraction spikes

Posted: Sat May 09, 2020 11:14 am
by Troy Galebach
Thanks Ivo, as you may recall, I don't think as well as I used to. I'll bet that's the issue, but if it were taken by an FSQ 106, there is a camera rotator which should have cancelled that out if it were used correctly. Perhaps that is the root cause of the uneven and multi angled spikes. I can live with that, it still is 10s of times better data than I can collect with my FSQ 106 with a good mount on wheelie bars, even if I were willing to upgrade to a fancy camera! :lol: I bought a new iOptron CEM40 which arrived here in November, it still hasn't seen any use because the weather has been so awful. Hopefully, by months end, I'll have it out at the observatory.

Best Regards,
Troy

Re: M45 can I minimize diffraction spikes

Posted: Sat May 09, 2020 1:34 pm
by Troy Galebach
admin wrote:Aha! The multi-date acquisition may have something to do with it. If the orientation of the camera vs object changes, then the diffraction pattern will change as well.
This can be a good thing too, as some stacking rejection algorithms will be able to reject the spikes if they keep changing per frame.

If PI is not your cup of tea, you could try Astro Pixel Processor as well.

Clear skies!
Ivo, I don't know if I should reply here, or in the feature request section. I tried to repair using lasso but my mousing skills evidently need help. For quick removal of spikes, satellite trails, meteor trails, plane lights, etc from a frame, a feature which allowed me to select a start point on the offending line, with a visible rubber band line to the mouse cursor which I would use to select an second point on the line by a second click. Repeat as necessary for each diffraction line, then repair/heal all at once. I think that would be much easier and quicker than the lasso. I'll bet you already considered that, or it is already there and I haven't found it, but on the slight chance I thought of something useful, which might be useful for others, and doable by you, I thought I'd share the idea. Ordinary people probably can use lasso, but I think the 2 point rubber band method would be quicker and more accurate.

Regards,
Troy

Re: M45 can I minimize diffraction spikes

Posted: Mon May 11, 2020 3:16 am
by admin
Troy Galebach wrote: Ivo, I don't know if I should reply here, or in the feature request section. I tried to repair using lasso but my mousing skills evidently need help. For quick removal of spikes, satellite trails, meteor trails, plane lights, etc from a frame, a feature which allowed me to select a start point on the offending line, with a visible rubber band line to the mouse cursor which I would use to select an second point on the line by a second click. Repeat as necessary for each diffraction line, then repair/heal all at once. I think that would be much easier and quicker than the lasso. I'll bet you already considered that, or it is already there and I haven't found it, but on the slight chance I thought of something useful, which might be useful for others, and doable by you, I thought I'd share the idea. Ordinary people probably can use lasso, but I think the 2 point rubber band method would be quicker and more accurate.

Regards,
Troy
Thanks Troy - something roughly similar is indeed in place. Select the "Line Toggle" Brush Mode. Then, move your mouse to the start of the line, press the mouse button, keep it pressed while dragging to the end of the line, then let go. This will draw a line between two points.

Do this for every diffraction spike/line.

Afterwards, you can use the Grow button to make all the lines thicker at once, or use the the Grow button once and then grow individual lines to the thickness required by using the "Grow Blob" Brush Mode.

Does that do what you need?

Cheers!