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

Trapezium and the Core of the Orion Nebula Using Lucky Imaging

  • Please log in to reply
16 replies to this topic

#1 james7ca

james7ca

    Cosmos

  • *****
  • topic starter
  • Posts: 9,707
  • Joined: 21 May 2011
  • Loc: San Diego, CA

Posted 17 March 2018 - 07:35 AM

Yesterday evening I had a few hours of partly cloudy skies between the rain storms and wanted to get what will probably be my last look at the Orion Nebula until next fall (looks like rain for the next week to 10 days followed by another cycle on the gibbous and full moon). So, I decided to do a high resolution shot of the core of the nebula, including the Trapezium.

 

This was going to be a so-called lucky imaging session using exposures that ranged from 1s down to 100mm using a Tele Vue NP127is with an uncooled QHY5III-178C camera. Since I wanted to take frames as quickly as possible I decided not to use Sequence Generator Pro (too slow) and thus imaged using SharpCap. I ended up taking over 60GB of images, 1000 at 1s, 4000 at 250ms, and another 800 at 100ms (ran out of disk space on the last sequence). I didn't use any guiding and I didn't even check my polar alignment (although it was probably pretty close, since I set the tripod and mount -- a Mach1GTO -- down on fixed marks on a concrete slab).

 

Later (shortly after on the same night) I also captured 256 dark fields at each exposure setting (but no flats, which I may add at a later time).

 

To qualify for lucky imaging I ran PixInsight's SubframeSelector script looking for the best 20% of the 1s subs and I ended up with a master integration that had a median FWHM of 1.86 arc seconds with a median eccentricity just over 0.3. I'm still working on the 250ms sequence, and that may/should produce an even sharper result. In the end (as shown below) I ended up with 3m 22s of total integration time (202 x 1s) at a gain setting of 30 and an offset of 128.

 

The original capture scale was at 0.75 arc seconds per pixel (660mm e.f.l. with 2.4um pixels) and I did a drizzle integration followed by a reduction to 0.94 arc seconds per pixel to produce a final image that recorded the Trapezium stars A, B, C, D, E, and F (the latter partially merged with star C). This reduction also just barely recorded the protoplanetary disks (propylids) G, H, and I. You need to look pretty closely to see the propylids at this image scale, but in the next post I've reproduced the drizzled master that has an image scale of 0.5 arc seconds per pixel and in that image the propylids are clearly visible.

 

Image processing was done with PixInsight and Photoshop CC2017.

Attached Thumbnails

  • Core of the Orion Nebula with NP127is and QHY5III-178C.jpg

Edited by james7ca, 17 March 2018 - 08:46 AM.

  • Stelios, okiestarman56, mikewayne3 and 12 others like this

#2 james7ca

james7ca

    Cosmos

  • *****
  • topic starter
  • Posts: 9,707
  • Joined: 21 May 2011
  • Loc: San Diego, CA

Posted 17 March 2018 - 07:41 AM

Here is a lightly stretched copy of the drizzled master, same data set as above, but with a crop and a reproduction scale of 0.5 arc seconds per pixel. The propylids are clearly visible in this version.

 

As I said earlier, I'm still working on the data for the 250ms subs and I think that will give me a little better resolution. More could come later.

 

[UPDATE]

Added a quick drizzle with just 4.5 seconds of total exposure (18 x 250ms), showing stars A thru F. It's going to take a little time to process and select the best subs from the 4000 x 250ms sequence. This isn't pretty, just a look at the potential resolution from the shorter exposures.

[/UPDATE]

Attached Thumbnails

  • Core Detail at One Half Arc Second Per Pixel.jpg
  • Drizzled 18 x 250ms.jpg

Edited by james7ca, 17 March 2018 - 09:56 AM.

  • Stelios, Gregory, Joe F Gafford and 7 others like this

#3 Vassar1976

Vassar1976

    Ranger 4

  • *****
  • Posts: 376
  • Joined: 24 Dec 2016
  • Loc: Raleigh, NC

Posted 18 March 2018 - 05:45 AM

Impressive capture of the Trapezium stars!!



#4 james7ca

james7ca

    Cosmos

  • *****
  • topic starter
  • Posts: 9,707
  • Joined: 21 May 2011
  • Loc: San Diego, CA

Posted 19 March 2018 - 02:08 AM

Impressive capture of the Trapezium stars!!

Thanks for the notice. I've now started working on the data set for the 250ms exposures and they have allowed me to fully resolve the "F" component. The median FWHM was 1.77 arc seconds on a total integration time of 2m 33s (613 x 250ms), that was about 15% of the 4000 subs. Not as "deep" as the 3m 22s that I got on the 1 seconds subs, but the stars look better on the selected 250ms subs (the 1s master had a FWHM of 1.86 arc seconds). I may try to combine the 1s and 250ms masters to produce a kind of HDR that could allow me to keep the better resolution on the Trapezium while giving even less noise in the faint "wings" on the nebula.

 

I should also get a slightly better resolution after I perform a deconvolution, but I'll probably do that after a drizzle integration which generally helps with the roundness of the stars while just slightly increasing the FWHM measurements, so the net gain after the drizzle will probably go toward improving the eccentricity (which isn't too bad right now -- in the 0.35 range).



#5 james7ca

james7ca

    Cosmos

  • *****
  • topic starter
  • Posts: 9,707
  • Joined: 21 May 2011
  • Loc: San Diego, CA

Posted 19 March 2018 - 09:23 AM

Here is my "final" shot of the Trapezium in Orion taken with a bunch of 0.25s exposures. In this reproduction you can definitely make out the stars A, B, C, D, E, and F and the protoplanetary disks (propylids) G, H, and I.

 

The total integration time was 2m 28s (592 x 0.25s). I used PixInsight's SubframeSelector script to select only the best subs, cutting pretty heavily to accept only 15% of the 4000 total subs. The drizzle integration produced a final master with a median FWHM of 1.92 arc seconds and an eccentricity of 0.32 (a deconvolution reduced the FWHM to 1.83 arc seconds). The image shown below was cropped and reduced in size to produce an image scale of 0.71 arc seconds. The original capture scale was 0.75 arc seconds -- 660mm focal length with a pixel size of 2.4um -- as produced at the prime focus of a Tele Vue NP127is and an uncooled QHY5III-178C camera.

 

Given that this was taken with a one-shot-color (OSC) camera it's a bit of a challenge to show this at near to a 1:1 reproduction from the original capture scale. Normally I don't like to go beyond about a 70% reduction when working with a OSC camera.

 

Image processing was done in PixInsight and Photoshop CC2017. I changed the color balance from my initial effort with the 1s subs, since I think the color on that was rather "cartoonish."

Attached Thumbnails

  • The Trapezium with Tele Vue NP127is and QHY5III-178C.jpg

Edited by james7ca, 19 March 2018 - 09:53 AM.

  • Starman27, jsmoraes, gvk and 4 others like this

#6 gvk

gvk

    Vostok 1

  • -----
  • Posts: 167
  • Joined: 05 Feb 2015
  • Loc: Mystic, CT, USA

Posted 19 March 2018 - 08:07 PM

Nicely done!

 

Gerry



#7 canukLX90

canukLX90

    Viking 1

  • -----
  • Posts: 724
  • Joined: 25 Dec 2008
  • Loc: B.C. Canada

Posted 20 March 2018 - 12:59 AM

Very very impressive effort!!!



#8 james7ca

james7ca

    Cosmos

  • *****
  • topic starter
  • Posts: 9,707
  • Joined: 21 May 2011
  • Loc: San Diego, CA

Posted 20 March 2018 - 09:05 PM

Gerry and canukLX90, thanks for the notice.

 

I just realized that I gave the wrong image scale for the reproduction shown in post #5, it's actually at 0.56 arc seconds per pixel and that's just over a 1.3X enlargement from the original capture. That's probably too much for a one-shot-color camera, almost twice what I would normally attempt with that kind of sensor (because of the Bayer pattern).

 

I had tried several different image scales and the 0.71 arc seconds that I mentioned earlier was from a different reproduction (an 800 x 800 sample, rather than the 1024 x 1024 that I uploaded).

 

Having said that, I think I'm going to revisit the processing and scaling on this image. I may try an HDR which should allow a slightly lower histogram stretch on the core while still capturing some of the fainter portions of the nebula. I guess this is another way of saying that I'm not really happy with the current image quality. I guess I knew that it would be a bit of a challenge to do this with a 5" scope with only 660mm of focal length. Since this will likely be my last opportunity this year to image the Orion Nebula I'm going to try to make the most of the data I have.

 

In any case, below is the smaller reproduction (800 x 800 pixels at 0.71 arc seconds per pixel), looks better in terms of sharpness.

 

 

Attached Thumbnails

  • The Trapezium with Tele Vue NP127is and QHY5III-178C 800 x 800.jpg

  • Quaternion, bmhjr and cyber like this

#9 entilza

entilza

    Soyuz

  • *****
  • Posts: 3,799
  • Joined: 06 Oct 2014
  • Loc: Oakville, ON, Canada

Posted 21 March 2018 - 07:51 AM

Very awesome project James always enjoy your efforts.



#10 james7ca

james7ca

    Cosmos

  • *****
  • topic starter
  • Posts: 9,707
  • Joined: 21 May 2011
  • Loc: San Diego, CA

Posted 24 March 2018 - 12:39 AM

Very awesome project James always enjoy your efforts.

Thanks.

 

I took some flats which I think will help with the color balance since I used a Baader SEMI-APO filter to help with the resolution with this shot (the NP127is isn't perfectly corrected for one-shot-color, you really want to refocus between the red, green, and blue channels by using a mono camera and filters). However, the SEMI-APO filter makes the one-shot-color results even more green than you'd get with a typical Bayer pattern and that made the color balance in the above images a little hard to correct. I also added 16GB of additional DRAM to my new Windows PC so PixInsight should now handle these large sets of subs with fewer issues.


Edited by james7ca, 24 March 2018 - 12:42 AM.


#11 james7ca

james7ca

    Cosmos

  • *****
  • topic starter
  • Posts: 9,707
  • Joined: 21 May 2011
  • Loc: San Diego, CA

Posted 25 March 2018 - 06:08 AM

Okay, can't give up yet. Here is another version with just 114 of the best subs from the 4000 sub sequence (that's just 2.9%). This only has 28.5 seconds of total integration time (114 x 0.25s) but the median FWHM is now 1.79 arc seconds (down from the earlier 1.92, both without deconvolution). This also had flats and I think the color is better (although not as bright or saturated). The "true" color of the core of the nebula is supposed to be green/blue (teal) and my earlier efforts were too green.

 

This certainly resolves A thru F and the propylids are still faintly visible. Once again, I did a drizzle integration which was followed by a crop and size reduction to produce an image scale just under 0.5 arc seconds per pixel. The image was captured at 0.75 arc seconds per pixel, so this is still an enlargement which is kind of hard for a one-shot-color camera.

 

Amazingly, I still have another version in the works, but with about twice the integration time (nearly one minute total).

 

Next year I may try to resolve H1/H2 as I've only seen one earth-based image that contained even a suggestion of doing that (those objects are apparently separate by about 1.6 arc seconds and in my shot they appear as a single object). I doubt that it is possible with a 5" scope, I may have to try my SCT.

Attached Thumbnails

  • Trapezium with Tele Vue NP127is and QHY5III-178C.jpg

  • Stelios, bill w, Joe F Gafford and 3 others like this

#12 canukLX90

canukLX90

    Viking 1

  • -----
  • Posts: 724
  • Joined: 25 Dec 2008
  • Loc: B.C. Canada

Posted 25 March 2018 - 10:00 PM

Very interesting project.  Love the colors...so many images wash out the "teal" of the inner core.  The following URL provides some info on the Trapezium stars:

 

https://maas.museum/...ebula-in-orion/

 

where it states that H1 / H2 are mag. 14.5 / 15.5 proplyds and if I understand the web page info correctly would be better imaged in the infrared bands which is an interest of mine.

 

I used the  "lucky" image technique on the Parachute awhile back to increase the resolution of my setup in combination with using a 2.5X PowerMate.  My QSI632 is not really suitable for such short exposures so I may have to invest in another CCD more suitable for the technique as your results are inspiring and impressive.

 

Looking forward to reading/seeing your future posts.

 

PJ

 

 



#13 james7ca

james7ca

    Cosmos

  • *****
  • topic starter
  • Posts: 9,707
  • Joined: 21 May 2011
  • Loc: San Diego, CA

Posted 25 March 2018 - 10:17 PM

PJ, thanks.

 

Yes, H1/H2 is pretty faint in my image and the pair certainly isn't resolved. I did find one image taken with a 25" refractor that just barely resolved H1/H2 (see link below to Flickr, a very nice image of the core).

 

  https://flic.kr/p/e6FcJX

 

The proplyds are definitely easier to image with H-alpha and I have a shot done with my 5" Tele Vue that clearly resolves the proplyd that is just beside star F (very rarely shown in any image and perhaps just barely suggested in the above image taken with the 25" scope).

 

My RGB image shows 4 proplyds (faintly) near to the core while the shot done in H-alpha has at least 6 (but H1/H2 is still unresolved). Here is a link to the H-alpha image:

 

  https://flic.kr/p/DWFvtH


Edited by james7ca, 25 March 2018 - 10:20 PM.


#14 james7ca

james7ca

    Cosmos

  • *****
  • topic starter
  • Posts: 9,707
  • Joined: 21 May 2011
  • Loc: San Diego, CA

Posted 07 April 2018 - 04:29 AM

It looks like I might have just barely detected both H1 and H2 in a drizzled stack of the very best 100 subs (median FWHM on the stack was 1.8 arc seconds). It's more like a suggested detection, since one component of the pair is very faint. The position angle looks correct and I measured the separation at something just over 1 arc second (online sources suggest the separation is 1.6 arc seconds, but there is the potential for a lot of error in my measurement since I was using a 200% resize on an already 2X drizzled image). I guess I should try to finish up on this project, do an HDR with the center core based only upon the best of the best subs.

 

At the risk of some ridicule (about detecting H1/H2), here is an annotated closeup of the Trapezium that was produced from the best 100 subs (total integration time 25 seconds, 100 x 250ms).

Attached Thumbnails

  • Trapezium Annotated.jpg

Edited by james7ca, 07 April 2018 - 05:18 AM.

  • Quaternion, Jon Rista and tkottary like this

#15 Jon Rista

Jon Rista

    ISS

  • *****
  • Posts: 24,854
  • Joined: 10 Jan 2014
  • Loc: Colorado

Posted 07 April 2018 - 04:26 PM

James, have you considered trying this with a longer focal length, so your image scale is truly high? I think a 0.75"/px scale is a good balance of high res and SNR, but you could probably eek out some more separation at 0.5"/px or around there.



#16 james7ca

james7ca

    Cosmos

  • *****
  • topic starter
  • Posts: 9,707
  • Joined: 21 May 2011
  • Loc: San Diego, CA

Posted 07 April 2018 - 07:28 PM

James, have you considered trying this with a longer focal length, so your image scale is truly high? I think a 0.75"/px scale is a good balance of high res and SNR, but you could probably eek out some more separation at 0.5"/px or around there.

Yes, I will probably try my EdgeHD next year, but it's always fun to try to see what can be done with smaller scopes like my SV80 or NP127is. I also have Tele Vue's 1.5X extender for the NP127is, but it doesn't do very well when photographing stars, although I have gotten some pretty good images of the moon using the extender. Tele Vue stopped making that extender several years ago and it was pretty expensive as new (I got mine used and fairly cheap). It converts the NP127 to a 990mm focal length f/7.8 system that is supposed to cover up to a full frame sensor, but in my hands the image quality has always been lacking (stars don't come out very round or small, but I don't think I've ever tried it for narrow band or even with separate RGB channels (where it might do better for DSO work).

 

In any case, if I simply switch to my mono ASI178 then I'll get a notable improvement in resolution, enough that I think I'll be able to resolve H1/H2 given some good seeing. In fact, I may revisit some of the sessions that I did in H-alpha to see whether I can clearly resolve that pair of propylids (the components G, H, and I are not stars but proto-planetary disks).

 

Oh, and here is a thread where I used the Tele Vue with a Powermate to cleanly resolve a sub-arc-second double star (near to the absolute limit for a 5" scope):

 

 https://www.cloudyni.../#entry8041810 


Edited by james7ca, 07 April 2018 - 09:12 PM.

  • Jon Rista likes this

#17 Dan M

Dan M

    Stellar Skies

  • *****
  • Vendors
  • Posts: 212
  • Joined: 11 Oct 2008
  • Loc: Central Texas

Posted 07 April 2018 - 10:37 PM

Very impressive James!  Thank you for sharing it and all your efforts to find more detail.




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