Jump to content


Photo

M101 new revision: a challenging object to process

  • Please log in to reply
16 replies to this topic

#1 Ron (Lubbock)

Ron (Lubbock)

    Messenger

  • -----
  • Posts: 439
  • Joined: 17 Aug 2012
  • Loc: West TX

Posted 15 June 2013 - 12:46 PM

Despite being one of the apparently larger galaxies in the sky, M101 has a reputation for being one of the most difficult major DSOs to image and process. This past spring, I found out the hard way that its reputation is well-deserved.

I have lost track of how many revisions of this image I have made. Quite simply, it was difficult to process because the color data did not present much contrast. In order to bring out the pink H-alpha regions and get some color in the spiral arms, one has to push the saturation to painful levels. As a result, any noise in the color data or gradients in the background will become obnoxiously apparent. Scores of nearly colorless M101 images appearing on AstroBin are a testament to this situation. I have not found the same to be true for M33, or other galaxies, which seemingly need less saturation boost to show their colors.

FWIW, this image had 10.75 hrs. of integration time, and most of it was spent on RGB data so I could have deep color data and increase the saturation. Only two hours were pure luminance, but I made a master synthetic luminance channel by stacking all of the frames. With 10.75 hrs. integration, many tiny background galaxies are visible in the full version I have uploaded to AstroBin (worth clicking and enlarging).

Link to full sized image on Astrobin

Posted Image

#2 Wouter D'hoye

Wouter D'hoye

    Surveyor 1

  • *****
  • Posts: 1519
  • Joined: 27 Jun 2003
  • Loc: Belgium

Posted 15 June 2013 - 12:56 PM

Hi,

Nice image but there is a very strong magenta cast all over; especially visible in the stars and in the core of the galaxy. That cast shouldn't be there.

Other than that. Nice image with plenty of detail and good depth.

Thanks for sharing.

Wouter.

#3 Ron (Lubbock)

Ron (Lubbock)

    Messenger

  • -----
  • Posts: 439
  • Joined: 17 Aug 2012
  • Loc: West TX

Posted 15 June 2013 - 08:48 PM

Wouter,

I honestly cannot see any trace of magenta on my monitor. I am left puzzled by your comment, except to say that our monitors must have different opinions of color.

Can anyone else see a pink/magenta cast to the image? Most of the stars are blue on my monitor, with a few orange ones in the corners.

#4 zerro1

zerro1

    Fly Me to the Moon

  • *****
  • Posts: 5862
  • Joined: 02 Aug 2009
  • Loc: Smokey Point , 48.12°N 122.25°W Elevation:512 ft

Posted 15 June 2013 - 10:13 PM

Can anyone else see a pink/magenta cast to the image? Most of the stars are blue on my monitor, with a few orange ones in the corners.


the core of 101 and stars are pinkish/magenta

#5 CounterWeight

CounterWeight

    Cosmos

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

Posted 16 June 2013 - 12:22 AM

Yes, I am seeing it everywhere. Also note some dimpling (dark spot at center) in some of the larger/brighter stars.

#6 Wouter D'hoye

Wouter D'hoye

    Surveyor 1

  • *****
  • Posts: 1519
  • Joined: 27 Jun 2003
  • Loc: Belgium

Posted 16 June 2013 - 03:19 AM

Hi,

monitor calibration can (and most likely will) be the issue here. To prevent that I usually measure the colors in photoshop. This will most likely lower the risks of strong casts.

Wouter.

#7 Ron (Lubbock)

Ron (Lubbock)

    Messenger

  • -----
  • Posts: 439
  • Joined: 17 Aug 2012
  • Loc: West TX

Posted 16 June 2013 - 12:17 PM

I guess I must have an unusual monitor. Never would have known about the rosy glow if you hadn't pointed it out! I also do not have any dimpling on my screen, and I didn't turn up the highlight setting much at all in Deep Sky Stacker, so that has me puzzled too.

I sometimes use Excalibrator to adjust colors before processing the final image, but it's too late for that. Also, I'd have to start the processing over to get rid of any dimpling. In any event, I have tried two new tweaks to the color balance. Without any elaboration on what they were, here are the new versions. Let me know if the color looks better on your monitors, and which one you prefer!

New Version 1:

Posted Image

New Version 2:

Posted Image

#8 Wouter D'hoye

Wouter D'hoye

    Surveyor 1

  • *****
  • Posts: 1519
  • Joined: 27 Jun 2003
  • Loc: Belgium

Posted 16 June 2013 - 12:57 PM

Hi,

of the two you just posted the second one looks best. But still the magenta cast is there. I hope you don't mind but I took a go at it too. see the attached image.

Attached Files



#9 Wouter D'hoye

Wouter D'hoye

    Surveyor 1

  • *****
  • Posts: 1519
  • Joined: 27 Jun 2003
  • Loc: Belgium

Posted 16 June 2013 - 01:03 PM

Now, what did i do, first of all I made a number of color samples, one on the bright star, one in the core of the galaxy and one in the background. There are a few things you know, the background for instance should be more or less neutral in this part of the sky. The bright star will be near saturation and should be near white. With at most a slight blue or orange cast. the core of the galaxy is usually more yellow. These are things you find out by looking at other peopl's images. Sometimes it can be a help to check other imager's results to at least know where you want to go. With these known quatities i started to adjust the curves of your image, bot the RGB as the individual channels. This is what I came up with:

Attached Files



#10 Ron (Lubbock)

Ron (Lubbock)

    Messenger

  • -----
  • Posts: 439
  • Joined: 17 Aug 2012
  • Loc: West TX

Posted 16 June 2013 - 02:03 PM

Thanks for analyzing the color for me. Your version looks very nice on my monitor, minus the loss of resolution from JPG compression, of course. Interestingly, the pink regions in your image actually look more vivid on my screen. I unfortunately don't own Photoshop or know how to use it, but I will see if I can work out something similar in IRIS or ImageJ, which I know pretty well.

#11 SKYGZR

SKYGZR

    Viking 1

  • *****
  • Posts: 843
  • Joined: 13 Aug 2009
  • Loc: Planet Earth, Milky Way Galaxy, Speeding towards the Virgo Supercluster

Posted 22 June 2013 - 02:39 PM

I feel your pain...here's attempts also..tricky bugger..

Attached Files



#12 SKYGZR

SKYGZR

    Viking 1

  • *****
  • Posts: 843
  • Joined: 13 Aug 2009
  • Loc: Planet Earth, Milky Way Galaxy, Speeding towards the Virgo Supercluster

Posted 22 June 2013 - 02:40 PM

Another

Attached Files



#13 SKYGZR

SKYGZR

    Viking 1

  • *****
  • Posts: 843
  • Joined: 13 Aug 2009
  • Loc: Planet Earth, Milky Way Galaxy, Speeding towards the Virgo Supercluster

Posted 22 June 2013 - 02:42 PM

Different Stacks & processing results in different ends

Attached Files



#14 Ron (Lubbock)

Ron (Lubbock)

    Messenger

  • -----
  • Posts: 439
  • Joined: 17 Aug 2012
  • Loc: West TX

Posted 22 June 2013 - 04:06 PM

I like your second one the best. The third one has a pink core on my colorblind monitor, so I'm guessing it's flaming red on Wouter's!

Your images show one of the problems I was talking about. There are some H-alpha regions in your image that are clearly defined, but they don't look distinctly pink in image #2. Turning up the red or saturation on your image will probably cause the whole thing to look pink. If you look at images taken with large scopes by people with excellent processing skills, there is a lot of contrast between the pink highlights and predominantly blue spiral arms. I'm just not getting that contrast without some unwanted color balance issues.

#15 SKYGZR

SKYGZR

    Viking 1

  • *****
  • Posts: 843
  • Joined: 13 Aug 2009
  • Loc: Planet Earth, Milky Way Galaxy, Speeding towards the Virgo Supercluster

Posted 22 June 2013 - 06:06 PM

The first was 75% of the best frames, the 2nd was 80%, the third, 100%. Aprx total time is around 4-6 hrs, used in camera darks, no flats, bias, etc.

Yup..a DSLR limits details to a degree..a "true" mono CCD w/filters would be a better solution for targets like these.

#16 Wouter D'hoye

Wouter D'hoye

    Surveyor 1

  • *****
  • Posts: 1519
  • Joined: 27 Jun 2003
  • Loc: Belgium

Posted 23 June 2013 - 03:30 AM

Hi,

I agree, the second image of skygzr looks best. If you used a DSLR it's quite normal that you have to pull really hard to bring out H-alpha. Usually at the cost of overdoing it in other areas. Often other imagers actually take separate NB H-alpha images to blend in to even further boost the H-alhpa regions in the image.

Wouter.

#17 SKYGZR

SKYGZR

    Viking 1

  • *****
  • Posts: 843
  • Joined: 13 Aug 2009
  • Loc: Planet Earth, Milky Way Galaxy, Speeding towards the Virgo Supercluster

Posted 23 June 2013 - 10:52 PM

That's on the list..want to acquire some HA time on this target for blending purposes...






Cloudy Nights LLC
Cloudy Nights Sponsor: Astronomics