APP can be fooled....

[tomato]

I'm trying to process six hours of data captured last night and I couldn't figure out why the stack had weird artefacts around the stars. I checked the ones with a low quality score and removed them from the stack  as I usually do and so I had no choice but to check every one of the 350 frames.

And guess what, the best quality frames  from each scope looked like this:

I remember banging my head against the scopes in the dark so this must be it. I guess APP has identified the juddered stars as separate and hence the star count has gone up compared to the rest of the subs and this has made it the top scorer?

Moral of the tale is don't take the analysis for granted, check your subs visually, even if you have 350 of them.

Edited February 11, 2022 by tomato

[scotty38]

Just curious but if it's only one sub wouldn't or shouldn't the "outliers" have been removed just like star trails would be?

[gilesco]

Would be good to know as to why you have the juddering?

 

[ONIKKINEN]

I had the exact same thing happen with DSS, all of my images were stacked on one trailed image and it looked a bit like yours. Was going mad with this and tried lowering the stack best % setting and did not work. After going through all the subs visually i found that the highest score frame was the one bad frame.

Since then i make NINA write some statistics of the sub to the filename itself, so that i dont have to visually inspect a thousand fits files.

Along time and date, exposure time and the usual stuff i have NINA write guiding RMS error from the duration of the sub, star HFR and number of stars in the frame. Without looking at the subs themselves i can see a few obvious outliers in the below list and i just delete them without thinking about it too much. This removes most of the obviously unsuitable images before stacking.

[tomato]

It was my own fault really, I saw the juddered frame come through on NINA, so I knew it was in there and I didn't see it in the frames I removed, so I should have gone looking for it before starting the normalisation and integration steps.  It just never occurred to me that it could be the reference frame.

1 hour ago, scotty38 said:

Just curious but if it's only one sub wouldn't or shouldn't the "outliers" have been removed just like star trails would be?

It was the reference frame, I'm not sure how well APP deals with pseudo stars, it might not see them as outliers.

1 hour ago, gilesco said:

Would be good to know as to why you have the juddering?

 

I banged my head on the scopes while adjusting the dome controls, so it was a one off event (until I do it again!)

[Tomatobro]

I wonder if the CMN Viewer software would allow a rapid scroll through of the frames to pick out the bad one would work. I guess the individual frame size would slow it down a bit but it could be worth a try.

if something like bumping into the scope happens during the night I make a note of the time so I can seek out those frames for special attention.