Jump to content


Photo

Detail in the Soul Nebula

  • Please log in to reply
41 replies to this topic

#1 Madratter

Madratter

    Fly Me to the Moon

  • *****
  • Posts: 6347
  • Joined: 14 Jan 2013

Posted 16 December 2013 - 10:31 PM

This object is normally done as a wide field, but there is plenty of detail in it as can be seen here. This is 35 HA subs of 5 minutes each binned 2x2 (total of 2 hour 55 minutes).

Posted Image

Attached Files



#2 terry59

terry59

    Aurora

  • *****
  • Posts: 4831
  • Joined: 18 Jul 2011
  • Loc: Colorado, USA

Posted 16 December 2013 - 10:50 PM

That looks nice. I was wondering which part you were shooting

#3 jbalsam

jbalsam

    Messenger

  • *****
  • Posts: 492
  • Joined: 06 Jul 2012
  • Loc: Darnestown, MD

Posted 16 December 2013 - 11:09 PM

Your image (at full resolution) shows a denoising issue that I've been curious about while playing with PixInsight. I use their a trous wavelets transform, and I see the same thing as in your image (when I'm denoising images that I know could use more data): fairly smooth results, with randomly interspersed darker pixels. I've played with settings quite a bit to try to remove them, but I can't do it without producing an objectionably "over smoothed" look. Have you noticed the same thing?

#4 terry59

terry59

    Aurora

  • *****
  • Posts: 4831
  • Joined: 18 Jul 2011
  • Loc: Colorado, USA

Posted 16 December 2013 - 11:38 PM

Your image (at full resolution) shows a denoising issue that I've been curious about while playing with PixInsight. I use their a trous wavelets transform, and I see the same thing as in your image (when I'm denoising images that I know could use more data): fairly smooth results, with randomly interspersed darker pixels. I've played with settings quite a bit to try to remove them, but I can't do it without producing an objectionably "over smoothed" look. Have you noticed the same thing?


Not trying to usurp his thread but have you tried cosmetic correction?

#5 CounterWeight

CounterWeight

    Cosmos

  • *****
  • Posts: 8160
  • Joined: 05 Oct 2008
  • Loc: Palo alto, CA.

Posted 16 December 2013 - 11:49 PM

Nice close up of what I call the mouth part of this nebula :). You have a lot of detail in there given your exposure time. The lower level signal starts becoming granular, only thing I think would help is longer sub exposures.

#6 jbalsam

jbalsam

    Messenger

  • *****
  • Posts: 492
  • Joined: 06 Jul 2012
  • Loc: Darnestown, MD

Posted 17 December 2013 - 12:02 AM

Terry - no I haven't. If you look at Madratter's image at full resolution you'll see what I'm talking about. There are literally hundreds of thousands of darker pixels fairly evenly distributed all over the image (noise remnants). Cosmetic correction as I understand it would not be feasible (i.e. going through the image in PS with the spot removal tool or something like that). The only way I've solved it is by taking more data. The only time it's really an issue is when I want to share an image after one or two nights of integration (before it's really complete). It seems like the sort of issue that PI would be able to solve easily. All of the problem pixels are single pixels (first level wavelets?), but for some reason they don't get removed. Maybe they're just too far beyond the standard deviation of other nearby pixels to be considered as noise?

#7 Madratter

Madratter

    Fly Me to the Moon

  • *****
  • Posts: 6347
  • Joined: 14 Jan 2013

Posted 17 December 2013 - 08:07 AM

Thanks Terry. This section is centered around IC 1871 instead of the Soul Nebula proper which is IC 1848.

Thanks Jim. I had planned to collect more data on it the other night. Two hours in, I discovered I had centered it wrong. Doh! I then switched over to the Jellyfish.

Josh, I have certainly noticed what you are talking about and it is a pain. I think you can reduce it by doing the initial integration with Winsorized and the lower bounds at 2 instead of the default 4. But I have stopped doing that because of the risk of clipping real signal.

I did the noise reduction here using tgvDenoise so it is not just a atrous wavelets issue.

And hijack away guys. As I have stated in the past, I never mind technical discussions hijacking my threads.

#8 terry59

terry59

    Aurora

  • *****
  • Posts: 4831
  • Joined: 18 Jul 2011
  • Loc: Colorado, USA

Posted 17 December 2013 - 08:58 AM

Maybe I'm thinking of the wrong process. It's been a while since I had the issue some time ago.

#9 terry59

terry59

    Aurora

  • *****
  • Posts: 4831
  • Joined: 18 Jul 2011
  • Loc: Colorado, USA

Posted 17 December 2013 - 09:03 AM

Thanks Terry. This section is centered around IC 1871 instead of the Soul Nebula proper which is IC 1848.

Thanks Jim. I had planned to collect more data on it the other night. Two hours in, I discovered I had centered it wrong. Doh! I then switched over to the Jellyfish.


Did you intend to image IC 1848? I shot that last night with better stars in the lower right hand portion of the image after removing the filter from the carousel and putting it back in.

#10 Footbag

Footbag

    Fly Me to the Moon

  • *****
  • Posts: 5841
  • Joined: 13 Apr 2009
  • Loc: Scranton, PA

Posted 17 December 2013 - 09:34 AM

Looks good. I've passed this up as a long FL target for a while. You found a spot with some very nice interest.

#11 DrDispatch

DrDispatch

    Ranger 4

  • *****
  • Posts: 375
  • Joined: 26 Oct 2012
  • Loc: Spf, Mo

Posted 17 December 2013 - 09:37 AM

Very kewl image MR, as usual !
Terry- remove & replace that filter made the corners better?
wonder if the male threads are cut a little off on the filter itself?

#12 Madratter

Madratter

    Fly Me to the Moon

  • *****
  • Posts: 6347
  • Joined: 14 Jan 2013

Posted 17 December 2013 - 10:35 AM

Terry, I deliberately went after this section. I had spotted it in other wider images and knew I wanted to do it.

Thanks Adam and Dr.

#13 Madratter

Madratter

    Fly Me to the Moon

  • *****
  • Posts: 6347
  • Joined: 14 Jan 2013

Posted 17 December 2013 - 10:38 AM

Terry, I'm happy to hear you got some better results. With my carousel the filters are held flat with some rubber rings.

#14 terry59

terry59

    Aurora

  • *****
  • Posts: 4831
  • Joined: 18 Jul 2011
  • Loc: Colorado, USA

Posted 17 December 2013 - 10:45 AM

Very kewl image MR, as usual !
Terry- remove & replace that filter made the corners better?
wonder if the male threads are cut a little off on the filter itself?


Interesting hypothesis. I was assuming that the glass was tilted but maybe your explanation is correct. I have the Astrodon ordered and will see what it does :fingerscrossed:

#15 DrDispatch

DrDispatch

    Ranger 4

  • *****
  • Posts: 375
  • Joined: 26 Oct 2012
  • Loc: Spf, Mo

Posted 17 December 2013 - 11:41 AM

i used to be a machinist. was the 70's, still these long runs of parts on these modern machines have to have variances, be my luck to get one of the several cut during setup process:)

some of this stuff you can throw $$ at it and not get rewarded. the astrodons seem to be the best in this case

#16 DrDispatch

DrDispatch

    Ranger 4

  • *****
  • Posts: 375
  • Joined: 26 Oct 2012
  • Loc: Spf, Mo

Posted 17 December 2013 - 11:43 AM

terry, the glass may be tilted!! I had not thought of that:))
opinions are like.....:))))
but someone will guess it right!!!

I had the QHY FW and the 2" baaders caught about 1.5 threads, that was all there was, so thin!

#17 rigel123

rigel123

    Voyager 1

  • *****
  • Posts: 10891
  • Joined: 29 Jun 2009
  • Loc: SW Ohio

Posted 17 December 2013 - 12:40 PM

That's close to the same FOV I used for my image of this one, great job!

#18 SergeC

SergeC

    Apollo

  • *****
  • Posts: 1141
  • Joined: 23 Mar 2012
  • Loc: Gainesville, FL

Posted 17 December 2013 - 02:29 PM

Nice shot, there's some cool detail and structure there.

It does look a bit grainy in the low signal area, but I don't think it detracts from the image overall. I'm finding I need to grab at least 4-6hr of Ha to smooth out that low signal.

#19 terry59

terry59

    Aurora

  • *****
  • Posts: 4831
  • Joined: 18 Jul 2011
  • Loc: Colorado, USA

Posted 17 December 2013 - 05:26 PM

Nice shot, there's some cool detail and structure there.

It does look a bit grainy in the low signal area, but I don't think it detracts from the image overall. I'm finding I need to grab at least 4-6hr of Ha to smooth out that low signal.


Yes, I am finding that low signal areas at f/7.5 to be tough with just a few hours anyway.

#20 Madratter

Madratter

    Fly Me to the Moon

  • *****
  • Posts: 6347
  • Joined: 14 Jan 2013

Posted 17 December 2013 - 05:37 PM

My guess is that the ideal exposure where read noise gets overwhelmed is probably at least ten and more probably twenty plus minutes. Unfortunately, until I get OAG going, I need to live with five minute exposures. That is fine for LRGB where it does hit or come close to ideal exposure times.

My telescope is running at roughly f/5.8.

#21 Madratter

Madratter

    Fly Me to the Moon

  • *****
  • Posts: 6347
  • Joined: 14 Jan 2013

Posted 17 December 2013 - 05:39 PM

I should add that I definitely could have used more aggressive noise reduction but then you have the problem with things looking artificial and plastic like.

#22 jbalsam

jbalsam

    Messenger

  • *****
  • Posts: 492
  • Joined: 06 Jul 2012
  • Loc: Darnestown, MD

Posted 17 December 2013 - 05:53 PM

Agreed, Madratter - too much denoising makes things look far worse.

I'm curious about why you are limited to 5min exposures with a guide scope. What problems are you seeing at longer exposures?

#23 Madratter

Madratter

    Fly Me to the Moon

  • *****
  • Posts: 6347
  • Joined: 14 Jan 2013

Posted 17 December 2013 - 06:32 PM

I have enough flexure in the system that stars go noticeably out of round in reasonable seeing after that amount of time. I actually could go longer with my C8 despite mirror flop. My guess is that the flexure in the AT8RC is from the way the focuser is mounted to the primary. Just .2 arc-second of movement a minute leads to an error of 2 arc-seconds which is easily noticeable after 10 minutes. I hate eggs. Hence why I will probably be getting a OAG soon.

#24 jbalsam

jbalsam

    Messenger

  • *****
  • Posts: 492
  • Joined: 06 Jul 2012
  • Loc: Darnestown, MD

Posted 17 December 2013 - 06:40 PM

Weird. It does look like most people on AstroBin who use the AT8RC use an OAG. I see a few examples of people taking 10min images with guidescopes, but they are few and far between.

#25 terry59

terry59

    Aurora

  • *****
  • Posts: 4831
  • Joined: 18 Jul 2011
  • Loc: Colorado, USA

Posted 17 December 2013 - 06:47 PM

My guess is that the ideal exposure where read noise gets overwhelmed is probably at least ten and more probably twenty plus minutes. Unfortunately, until I get OAG going, I need to live with five minute exposures. That is fine for LRGB where it does hit or come close to ideal exposure times.

My telescope is running at roughly f/5.8.


Hmmm....I'm shooting 10 and 15 min subs now in NB.I should probably try 30 but I'd hate to lose any of those!






Cloudy Nights LLC
Cloudy Nights Sponsor: Astronomics