Jump to content

  •  

CNers have asked about a donation box for Cloudy Nights over the years, so here you go. Donation is not required by any means, so please enjoy your stay.

Photo

AutoStakkert Issue with Cloud Affected Solar Image

  • Please log in to reply
7 replies to this topic

#1 MichaelTeoh

MichaelTeoh

    Lift Off

  • -----
  • topic starter
  • Posts: 23
  • Joined: 16 Jul 2019
  • Loc: Penang, Malaysia

Posted 05 December 2020 - 03:03 AM

When doing solar surface imaging in h-alpha, sometimes thin cloud would cross during the session briefly.

 

When I try to stack the sequence with AS!3, after the "Analyse" step, the processing is limited to a small area (indicated by the red box) of what appears to be the least affected area of the FOV. 

 

is there a way to force AS!3 to process the whole frame instead of limiting it to the small area?

Thanks.

 

Attached Thumbnails

  • SunCloud.jpg


#2 MalVeauX

MalVeauX

    Cosmos

  • *****
  • Posts: 8,348
  • Joined: 25 Feb 2016
  • Loc: Florida

Posted 05 December 2020 - 07:49 AM

Heya,

 

Clouds pose big problems because the contrast changes through the sequence will appear to be different "lines" to the laplace algorithm that is looking for contrast line transitions to consider it the same data to then align it. So it will pitch most of the frames.

 

Increase the anchor window as big as you can (+ key on keyboard).

Try increasing noise robust to be larger than 6 and see if it handles the loss in contrast changes better.

 

For the reference frame, click the double-stack option and uncheck automatic reference and choose most of your frames (the # of frames) for the reference frame option.

 

Very best,


  • dragracingdan and walkman like this

#3 MichaelTeoh

MichaelTeoh

    Lift Off

  • -----
  • topic starter
  • Posts: 23
  • Joined: 16 Jul 2019
  • Loc: Penang, Malaysia

Posted 05 December 2020 - 08:08 AM

Thanks Marty, I apply all your suggestions and it works. 

Cheers.


  • MalVeauX likes this

#4 MalVeauX

MalVeauX

    Cosmos

  • *****
  • Posts: 8,348
  • Joined: 25 Feb 2016
  • Loc: Florida

Posted 05 December 2020 - 08:19 AM

Thanks Marty, I apply all your suggestions and it works. 

Cheers.

waytogo.gif

 

The double-stacking of the reference frame does the initial stack to generate a reference frame, then takes the stacked image and generates a new reference from to then generate a stack of reference for. This is good for things that move a lot, like solar. And when you uncheck auto on the reference stack size, it generally takes 50% of the frames on auto. If you're losing frames to low contrast or movement, increase the reference stack size to lots more frames, it will then stack the "problems" into the final reference stack image and then the other frames are more likely to align to the final data as it included them in the reference. Then when doing your align points, do small align points and use multi-scale, this will help fitting the aligned data back together where it should be and have less error/artifact. The higher noise robust will allow for lower contrast when it comes to laplace calculations, so it will attempt to align things even if they don't perfectly match to get a chance to use the data more often.

 

Very best,


  • walkman and MichaelTeoh like this

#5 rigel123

rigel123

    Hubble

  • *****
  • Posts: 18,743
  • Joined: 29 Jun 2009
  • Loc: SW Ohio

Posted 05 December 2020 - 08:33 AM

If all else fails with Marty's suggestions you might want to give Registax a chance at the data.  Sometimes when AS!3 fails at images because of things like clouds Registax can go ahead and stack based upon the frame you pick as the reference.  The other option is to clip the video at the point the clouds are obvious and run the shortened video through AS!3.


  • BYoesle and walkman like this

#6 MichaelTeoh

MichaelTeoh

    Lift Off

  • -----
  • topic starter
  • Posts: 23
  • Joined: 16 Jul 2019
  • Loc: Penang, Malaysia

Posted 05 December 2020 - 10:30 AM

Hi Marty, out of curiosity, I try to stack a second time with the same settings as you suggested, but this time the processing area is slightly smaller than the first try. No matter how many time I retry, I won't get the almost full processing area like in the first try. Weird.

Hi Warren, I have tried both Registax and AviStack before posting, both failed. Registax freezes after clicking align. AviStack wouldn't load the video at all. 



#7 MalVeauX

MalVeauX

    Cosmos

  • *****
  • Posts: 8,348
  • Joined: 25 Feb 2016
  • Loc: Florida

Posted 05 December 2020 - 10:40 AM

Hi Marty, out of curiosity, I try to stack a second time with the same settings as you suggested, but this time the processing area is slightly smaller than the first try. No matter how many time I retry, I won't get the almost full processing area like in the first try. Weird.

Hi Warren, I have tried both Registax and AviStack before posting, both failed. Registax freezes after clicking align. AviStack wouldn't load the video at all. 

Ok, then probably just need to remove the worst clouded frames. You could use PIPP to quality weight and reorder the frames and then only keep the best frames before the clouds come in. Then run that output through AS!3.

 

Very best,


  • rigel123 and walkman like this

#8 MichaelTeoh

MichaelTeoh

    Lift Off

  • -----
  • topic starter
  • Posts: 23
  • Joined: 16 Jul 2019
  • Loc: Penang, Malaysia

Posted 05 December 2020 - 10:49 AM

Never mind, I still keep the resulting stacked image from the first try, so I'll just use that.

 

I was just curious to know which of the four settings (larger anchor box, higher noise robust, double stack reference, manual size) or any combo of those would have given the desired result. But seem like the magic has disappear, and it only happens once. 


  • MalVeauX likes this


CNers have asked about a donation box for Cloudy Nights over the years, so here you go. Donation is not required by any means, so please enjoy your stay.


Recent Topics






Cloudy Nights LLC
Cloudy Nights Sponsor: Astronomics