Author Topic: Another combining subs of different length question.  (Read 3101 times)

Offline Buzz

  • PixInsight Enthusiast
  • **
  • Posts: 80
    • The Astrophotography Manual - support
Another combining subs of different length question.
« on: 2014 September 18 04:00:27 »
I purposefully exposed M31 at long, medium and short times to ensure the core was not blown out. Whilst I know I can combine stacked, DBE'd images together with HDRCombine - that also means I lose the benefit of the data from the shorter exposure times in the peripheral areas when it is combined with the deeper exposure time.

Does the following make sense?
Use Integration to combine the DBE'd stacks of 120, 300 and 600s subs. This will combine using noise levels to set the proportion but produce a blown out core. The fainter periphery will benefit from the longer overall exposure contribution. Then - use HDR to just substitute in the blown out core with the short exposure stack?
« Last Edit: 2014 September 18 06:00:12 by Buzz »
regards
Chris

Offline pfile

  • PTeam Member
  • PixInsight Jedi Grand Master
  • ********
  • Posts: 4729
Re: Another combining subs of different length question.
« Reply #1 on: 2014 September 18 09:14:54 »
yeah, i think that makes sense. it's worth trying anyway. you can probably just integrate all your subs of differing lengths together as an alternate technique.

rob

Offline Buzz

  • PixInsight Enthusiast
  • **
  • Posts: 80
    • The Astrophotography Manual - support
Re: Another combining subs of different length question.
« Reply #2 on: 2014 September 18 12:38:40 »
I gave it a go -  I took three DBE'd stacks and used integration on them, simple averaging and no rejection. The resulting file had very slight posterization at the core - coincidental with the clipping boundary of the medium and long exposures.   The SNRweighting level, measured by subframe selector was better however than a simple HDRcombine.

I then used HDRcombine to smooth in the core with the longest sub exposure DBE'd stack.  As the file was a simple average of three files, not surprisingly, the threshold was about 0.3.  I think I now have the best of both worlds - smooth cores without clipping and better background noise levels. 
regards
Chris