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

Crosshatch pattern on SII stack

  • Please log in to reply
15 replies to this topic

#1 Wyvernp

Wyvernp

    Sputnik

  • -----
  • topic starter
  • Posts: 33
  • Joined: 22 Feb 2020
  • Loc: Haywards Heath, UK

Posted 13 July 2020 - 03:11 PM

Hi All.

I'm having a strange issue with my SII stacks. I used the same batchpreprocessing script in PI to do Ha, OIII and SII all at 300s so also used the same dark frames. I cant seem to see this pattern at all within the individual subs although there was a lot of high level cloud when they were taken. The flats were taken this morning using the SGP flats wizard again using the same process as usual with a white lit ipad screen with a white tshirt over the scope. Target ADU was 22000. subs were 4.5 secs-ish. I also tried stretching the master flat and could not see this pattern.

 

SIIintegrationsmall.jpg

 

Any help would be appreciated.

 

Cheers,

 

Wyvern



#2 Stelios

Stelios

    Cosmos

  • *****
  • Moderators
  • Posts: 9,477
  • Joined: 04 Oct 2003
  • Loc: West Hills, CA

Posted 13 July 2020 - 03:19 PM

Are you dithering? Dew possible?



#3 Wyvernp

Wyvernp

    Sputnik

  • -----
  • topic starter
  • Posts: 33
  • Joined: 22 Feb 2020
  • Loc: Haywards Heath, UK

Posted 13 July 2020 - 03:31 PM

Yes I was, medium dither on each frame, although I did the same with the OIII and Ha too.

Dew heaters were on so I don’t think so but I didn’t actually check so I can’t say that with 100% certainty.

#4 SilverLitz

SilverLitz

    Viking 1

  • -----
  • Posts: 816
  • Joined: 17 Feb 2018
  • Loc: Louisville, KY

Posted 13 July 2020 - 03:54 PM

Way too regular to be dew.  I have seen something similar on some Ha frames taken on the Fireworks Galaxy.  I assume it might be because the signal was so low and therefore the stretching so great to magnify small problems in calibration frames.



#5 Der_Pit

Der_Pit

    Surveyor 1

  • *****
  • Posts: 1,513
  • Joined: 07 Jul 2018
  • Loc: La Palma

Posted 13 July 2020 - 05:14 PM

To me it looks like an interpolation moire pattern that you can easily get if you have to rotate the image for alignment.  It should however go away if you have enough dither positions with sufficient offset.  

That could at least explain why you can't see it in single subs.


  • sharkmelley likes this

#6 smccully

smccully

    Vostok 1

  • *****
  • Posts: 171
  • Joined: 25 Dec 2018
  • Loc: Washington

Posted 13 July 2020 - 06:37 PM

My guess would be flats, I have had this kinda of issue with some Light Panels but usually with LRGB filters. I would take a second a look at your Flats and inspect for anything that could give off a similar pattern in the MasterFlat, as well as each Flat frame. 

 

Second thing to do, is to do the Calibration one at a time. First just bias, then just darks, and then just Flats, and then Dark and Bias, etc.



#7 ChrisWhite

ChrisWhite

    Fly Me to the Moon

  • *****
  • Posts: 5,931
  • Joined: 28 Feb 2015
  • Loc: Colchester, VT

Posted 13 July 2020 - 09:08 PM

Der pit is correct I believe.

Try a different interpolation method for registration. With batch processing im not sure how to do this, but if you can manually select one try cubic-b-spline instead of the default which I believe is lanczos.

CBS should fix the pattern issue. Its not quite as sharp as lanczos but much better than this.

#8 pfile

pfile

    Fly Me to the Moon

  • -----
  • Posts: 5,564
  • Joined: 14 Jun 2009

Posted 13 July 2020 - 09:48 PM

it's is probably an interpolation artifact, but it could be caused by bad calibration. you need to add a pedestal when calibrating the SII frames. in dark skies with narrowband filters, what can happen is that there's so little background signal that the background is essentially a dark frame. 

 

then you come along and subtract a master dark from that "dark" frame. since there's shot noise in the dark current that's not in the master dark, some pixels are less than the master dark and some greater... and the pixels that are less work out to a negative value, which then gets clamped to 0. so then you have a bunch of 0 value pixels in your calibrated frames, and when you register and stack them you get these weird patterns.

 

so if the change in interpolation method doesnt help, try an output pedestal while calibrating. the problem is that i don't think you can do this in WBPP, so you'd have to do it manual.

 

rob


  • SilverLitz likes this

#9 Wyvernp

Wyvernp

    Sputnik

  • -----
  • topic starter
  • Posts: 33
  • Joined: 22 Feb 2020
  • Loc: Haywards Heath, UK

Posted 13 July 2020 - 11:36 PM

Thank you all for all your suggestions. I think the interpolation artifacts looks like the most likely at this point due to low signal, I have just changed from 8 to 3.5nm filters which I think were fine in the OIII and Ha. So here is the plan ...

 

I will try to use a different method, rather than interpolation. Failing that I'm going to increase my sub exposures from 5 to maybe 7-8 mins to try and gather a bit more signal in each sub. I will also redo my flats just to be sure.

 

Thank you all for your help, it is very much appreciated.

 

Wyvern



#10 ChrisWhite

ChrisWhite

    Fly Me to the Moon

  • *****
  • Posts: 5,931
  • Joined: 28 Feb 2015
  • Loc: Colchester, VT

Posted 14 July 2020 - 06:59 AM

Thank you all for all your suggestions. I think the interpolation artifacts looks like the most likely at this point due to low signal, I have just changed from 8 to 3.5nm filters which I think were fine in the OIII and Ha. So here is the plan ...

 

I will try to use a different method, rather than interpolation. Failing that I'm going to increase my sub exposures from 5 to maybe 7-8 mins to try and gather a bit more signal in each sub. I will also redo my flats just to be sure.

 

Thank you all for your help, it is very much appreciated.

 

Wyvern

 

You porbbaly wont be able to get enough signal to bury this pattern.  Thats the curse of narrowband imaging.  wink.gif

 

In my fairly dark skies I get this pattern with all three NB channels when the target is moderately dim, especially when using cameras such as the ASI 1600. 

 

All registration processes will use an interpolation algorithm.  You just need to experiment to find which works best for your data. 
 



#11 Der_Pit

Der_Pit

    Surveyor 1

  • *****
  • Posts: 1,513
  • Joined: 07 Jul 2018
  • Loc: La Palma

Posted 16 July 2020 - 12:35 PM

Hmm, I wonder if Drizzling could help there, as that doesn't interpolate the source.  In principle one should be able to use a scale of 1 and no 'droplet downscale', then it's just doing a 'normal' image, but with the drizzle summing algorithm...



#12 ChrisWhite

ChrisWhite

    Fly Me to the Moon

  • *****
  • Posts: 5,931
  • Joined: 28 Feb 2015
  • Loc: Colchester, VT

Posted 16 July 2020 - 12:49 PM

Hmm, I wonder if Drizzling could help there, as that doesn't interpolate the source.  In principle one should be able to use a scale of 1 and no 'droplet downscale', then it's just doing a 'normal' image, but with the drizzle summing algorithm...

Interesting.  I doubt it would fix it, but admittedly have never tried.  My approach to this kind of thing is to prevent it rather than correct for it. 

 

Even with drizzle the interpolation algorithm is implemented during registration, so I don't see how drizzling would have any mitigating effect after the pattern has been generated. 



#13 pfile

pfile

    Fly Me to the Moon

  • -----
  • Posts: 5,564
  • Joined: 14 Jun 2009

Posted 16 July 2020 - 12:59 PM

Interesting.  I doubt it would fix it, but admittedly have never tried.  My approach to this kind of thing is to prevent it rather than correct for it. 

 

Even with drizzle the interpolation algorithm is implemented during registration, so I don't see how drizzling would have any mitigating effect after the pattern has been generated. 

i agree that the problem should be fixed rather than worked around if possible. i guess adding a pedestal is also a workaround but it is the most reasonable one (if calibration is indeed the problem.)

 

the xdrz files need a pass thru StarAlignment so that they can be updated with the X/Y displacements for each image, but that does not mean that drizzle uses interpolation... it doesn't.

 

rob



#14 Der_Pit

Der_Pit

    Surveyor 1

  • *****
  • Posts: 1,513
  • Joined: 07 Jul 2018
  • Loc: La Palma

Posted 16 July 2020 - 01:01 PM

Even with drizzle the interpolation algorithm is implemented during registration, so I don't see how drizzling would have any mitigating effect after the pattern has been generated. 

Ah, that might be my lack of knowledge of the mainstream software....  I had assumed the drizzle implementation uses the full projection from the raw images, not (just) the scaling from the pre-aligned images.  In that case it would indeed not help.



#15 pfile

pfile

    Fly Me to the Moon

  • -----
  • Posts: 5,564
  • Joined: 14 Jun 2009

Posted 16 July 2020 - 02:07 PM

Ah, that might be my lack of knowledge of the mainstream software....  I had assumed the drizzle implementation uses the full projection from the raw images, not (just) the scaling from the pre-aligned images.  In that case it would indeed not help.

it doesn't - in PI the drizzle files refer back to the original unregistered images.

 

rob



#16 ChrisWhite

ChrisWhite

    Fly Me to the Moon

  • *****
  • Posts: 5,931
  • Joined: 28 Feb 2015
  • Loc: Colchester, VT

Posted 16 July 2020 - 02:50 PM

it doesn't - in PI the drizzle files refer back to the original unregistered images.

rob


I didnt realize that. Thanks for claeifying. Well worth a try then.


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