Stars getting chewed in PixInsight... what´s wrong?

[MartinFransson]

I get a problem when stacking images from multiple sessions in PixInsight. At least if the images are rotated a bit between the sessions. What happens is that the edges of brighter stars get eroded or how to put it. See attached image. To the left is an image stacked from multiple sessions, to the right is the same area from an image stacked from only one night.

What settings could cause this? I don´t get anything like that in DSS. I stack using darks, flats and bias and use Winsorized Sigma clipping with the default settings (3.00 and 4.00 or something like that).

[daz]

When you stack the multiple sessions, are all of the frames showing good star shapes?

Are you aligning each session separately, and then doing a final alignment - or are you aligning all frames as one set?

If the latter, the image you choose as the reference could make a difference...

When you do the stack, does the process window show any error messages, or numbers that are obviously different to the single session stack?

[MartinFransson]

Haven't seen any error messages.

I run all of them at once through the Batch Preprocessing. Some of the older subs might have stars that have slight trailing or are out of focus. I haven't run the subframe selector on them.

Sent from my phone using Tapatalk

[ChrisLX200]

I think you need to check all the subs going into that stack because you're losing detail and that is not something that should happen. Were some taken under a bright Moon and thus have a much higher background level?

ChrisH

[daz]

So I would pick the best frame as a reference frame, and then align each of the sessions to that same frame - and then do the stack.

And of course, make sure the best quality frames are used

[MartinFransson]

I think you need to check all the subs going into that stack because you're losing detail and that is not something that should happen. Were some taken under a bright Moon and thus have a much higher background level?

ChrisH

Yeah, some (but only a handful) were taken during a bright moon. Quite a few were also taken with bad tracking and slightly out of focus. The noise levels go down but so does the detail.

[chrisvdberge]

Try to analyse your subs with the SubframeSelector script. Then you can see if you have some bad frames in there.

Furthermore; you mention the image is rotated between sessions, do you mean the camera position in regards to the scope? So these star deformations we are seeing could be rotated?

[MartinFransson]

Try to analyse your subs with the SubframeSelector script. Then you can see if you have some bad frames in there.

Furthermore; you mention the image is rotated between sessions, do you mean the camera position in regards to the scope? So these star deformations we are seeing could be rotated?

Well, I used a tele lens. Don't use a scope yet. I rotated the lens to create better framing of the shot. It's got a ring-type adapter so I can rotate the lens and camera any way I want.

Got to try the subframe selector if I can figure out how it works.

Sent from my phone using Tapatalk

[chrisvdberge]

Well, I used a tele lens. Don't use a scope yet. I rotated the lens to create better framing of the shot. It's got a ring-type adapter so I can rotate the lens and camera any way I want.

Got to try the subframe selector if I can figure out how it works.

Sent from my phone using Tapatalk

Ok, so we're probably seeing the result of the difference in rotation between the 2 imaging sessions in terms of star deformations. But still weird that they are being 'chunked' like this. 

Do you have the ability to put your frames somewhere we can download it? i could have a look at stacking it myself. 

For the SubframeSelector script; I just wrote a post about that and explaining it here:

http://dslr-astrophotography.com/subframeselector-script/

[Phillyo]

Hi Martin,

I'm at work right now so I can't open my PI to check settings, however there's a setting I use to remove these artefacts as it has happened to me in the past, especially when I drizzle my subs. I will try to remember to check when I've finished work tonight and reply with an updated message explaining which settings I play around with to remove it. I'll have a good look through google today to see if I can find it.

Phil

[Phillyo]

OK so a very quick google search has made me think it's during the Star Alignment phase. At the very bottom of the star alignment you can select the interpolation method. Try changing that to (I think) Bicubic Spline with a clamping threshold of around 0.3 and see how your stars look? If that doesn't work then try playing with those settings a little bit to see how you get on.

Good luck.

Phil

[MartinFransson]

OK so a very quick google search has made me think it's during the Star Alignment phase. At the very bottom of the star alignment you can select the interpolation method. Try changing that to (I think) Bicubic Spline with a clamping threshold of around 0.3 and see how your stars look? If that doesn't work then try playing with those settings a little bit to see how you get on.

Good luck.

Phil

Thanks Phil! I will try that!

Sent from my phone using Tapatalk

[MartinFransson]

OK so a very quick google search has made me think it's during the Star Alignment phase. At the very bottom of the star alignment you can select the interpolation method. Try changing that to (I think) Bicubic Spline with a clamping threshold of around 0.3 and see how your stars look? If that doesn't work then try playing with those settings a little bit to see how you get on.

Good luck.

Phil

Hey! I think it worked! Going to try a bit more but it looks promising!

[Phillyo]

Fantastic. Let me know how you get on!!

Phil

[MartinFransson]

Fantastic. Let me know how you get on!!

Phil

Sorry, didn´t work. Will keep investigating..