Square Stars

[Jules]

Hi

I shot this image last year. It has been on the back burner as I am tying to figure out what is happening in the image. The attached image is integrated with no rejection. The integrated image comprises of 46 x 900s Lum subs.

Batch preprocessing and alignment are all in PI.

I have tried Blink, and tediously adding/rejecting subs in integration but I always end up with square stars?

I would appreciate some insight if anybody has seen this please?

Thanks

Julian 

[Cleon_Wells]

Julian
What do the individual subs look like?
       I've been playing around with some of my 155-sec subs of M99 and noticed that Star Alignment will align, with some difficulty (I assume because these subs have  poor SNR and also Hot pixels)  these subs incorrectly.
Cleon

[Jules]

Hi Cleon

Individual subs look fine to me?

Julian

[Cleon_Wells]

I still think it's an SA error, I would take  6 subs from the set, two from the top, the middle,  the bottom of your stack and manually Cal and SA and Int them and see what you get.
Cleon

[Juan Conejero]

Please upload a relevant set of raw data to reproduce the issue. Otherwise it is impossible to say anything useful about this.

[Jules]

Cleon and Juan

I have tried multiple subs, starting with 3 then adding individually up to 46. I have tried pixel interpolation algorithms, Lanczos 3 to 5 and experimenting with the clamping threshold.

Juan

Where can upload this dataset?

Thanks

Julian

[vicent_peris]

Hi,

The problem is that your dark signal was not well subtracted in the image calibration process and StarAlignment is detecting the hot pixels as stars and aligning the images to the hot pixels and not to the stars.

Try to raise the "Hot pixel removal" parameter in StarAlignment. Once your images are well aligned, ImageIntegration will reject all the residual hot and cold pixels.


V.

[Jules]

Thanks Vicent I will give it go.

Saludos

Julian

[Jules]

Hi Vicent

Just as a follow up and to say thanks. I have remedied my image issues with a defect map to reduce the hot pixels. The subs register a lot better now.

Thanks to Cleon as well

Regards

Julian

[vicent_peris]

Great. It should work also by using the dark optimization threshold parameter in BPP.

V.

[Cleon_Wells]

Julian, I’ve been dealing with similar problems with my Canon T1i, a lot of hot pixels. When I use SA and see over a hundred reference stars found, not all hot pixels I know I’ve got to start tweaking, this is one reason I don’t use BPP. I do see how BPP could give me Calibrated and Dedayered subs in the morning, Vicent I guess I should try it.

         Attached is my latest try on M99, this is a crop, I’m impressed on how well PI can turn my low quality subs into this, especially considering I don’t have a good handle on using PI
Cleon
PS Vicente I was at the Adler Planetarium workshop way back when.

[Jules]

Vicent

Interestingly enough I have the dark optimisation threshold selected in BPP as default.

Julian