Jump to content


Photo

Uphill battle. opinions wanted. M63

  • Please log in to reply
64 replies to this topic

#1 Aircrftr

Aircrftr

    Apollo

  • -----
  • Posts: 1090
  • Joined: 29 Sep 2006
  • Loc: Claremore OK

Posted 24 March 2014 - 09:33 PM

Well, between the moon, weather and equipment, so far I've managed to get 6 hours of LUM and 2hrs ea. of RGB. Now, for the life of me, I can't seem to get much color in the galaxy itself. the background, I believe, is a lot of IFN and the galaxy itself shows some nice tidal streams. That's what I THINK. The red at the bottom LH corner I believe is real. It's there in all my frames and doesn't seem to be artifact or gradient. Same goes for the subtle blues in the background elsewhere. I couldn't find a wide enough field of view in other pictures to verify this, but just wanted to see what you all think. My opinion is I just need more integration time of everything L&RGB. So what do you all think? Is the background actually IFN or is this just noise? I know the darker area in the upper RH side is there also in all frames, so likely it's real which suggests the IFN below it is also real.

Posted Image

#2 alpal

alpal

    Soyuz

  • -----
  • Posts: 3660
  • Joined: 15 Jun 2009
  • Loc: Melbourne Australia.

Posted 24 March 2014 - 10:01 PM

Hi Dan,
there is colour there.
I hope you don't mind?
I checked for colour using LAB mode as per here:

https://www.youtube....h?v=fE2GS_87Ct8

I did however come out all yellow.

My result is here & I did use a mask for the background.

Attached Files



#3 Jim Thommes

Jim Thommes

    Soyuz

  • ****-
  • Posts: 3768
  • Joined: 20 Sep 2004
  • Loc: San DiegoCA USA

Posted 24 March 2014 - 10:38 PM

Dan,
You don't describe how you processed the color. It would be useful to know. One thing I have realized is that high luminance signal level is the enemy of color. That may be why your background is more colorful than your galaxy. It would be nice to know your color calibration method (G2V factors, eXcalibrator, ??)

If you haven't already done so, you might try blending your luminance layer to RGB first at 1/2 level. In Photoshop that would be setting the input level at 255 and the output at 128. (Not the same as setting opacity at 50% for the layer). Flatten then mostly reverse the level (i.e. input at say 160, output 255). Those steps should bring out more color. If you want, you can saturate deeper in LAB color space. Then you can reapply the Luminance data as a luminosity layer.

#4 Aircrftr

Aircrftr

    Apollo

  • -----
  • Posts: 1090
  • Joined: 29 Sep 2006
  • Loc: Claremore OK

Posted 24 March 2014 - 10:40 PM

Hi Alpal;
No, I don't mind at all! I was able to get the same color while messing with it, but It just doesn't seem to be the "right" color.... You look at other images and you see a lot of them with much more blue in the arms. Maybe I just need more integration time in RGB? This is where I'd like to end up...

#5 Aircrftr

Aircrftr

    Apollo

  • -----
  • Posts: 1090
  • Joined: 29 Sep 2006
  • Loc: Claremore OK

Posted 24 March 2014 - 10:48 PM

Hi Jim;

I've never really gotten deep into photoshop but I've been immersing myself in PixInsight and likeing it quite a bit. What I done was got my G2V balance from the tool in CCDAutopilot and used that setting in CCDStack to make my LRGB. Converted the color image from CCDStack to a 16bit TIF file to use in PI. In PI, I've gotten used to all the functions covered by Harry's Astroshed tutorials, but just couldn't bring out the colors I wanted. Thinking along your same lines, I done just and RGB version and ended up with the same dull colors.

#6 alpal

alpal

    Soyuz

  • -----
  • Posts: 3660
  • Joined: 15 Jun 2009
  • Loc: Melbourne Australia.

Posted 25 March 2014 - 01:01 AM

Hi Alpal;
No, I don't mind at all! I was able to get the same color while messing with it, but It just doesn't seem to be the "right" color.... You look at other images and you see a lot of them with much more blue in the arms. Maybe I just need more integration time in RGB? This is where I'd like to end up...



Hi Dan,
yes I know what you mean - it's all yellow.
It could be the way you've processed it.
What did the blue signal look like?
Did you bin the colours 2x2?

cheers
Allan

#7 Aircrftr

Aircrftr

    Apollo

  • -----
  • Posts: 1090
  • Joined: 29 Sep 2006
  • Loc: Claremore OK

Posted 25 March 2014 - 05:12 PM

I'll post up my R,G, and B calibrated/registered frames in a while when I get back out to the observatory. Funny thing is, I don't usually have trouble with color. Oh, and I bin all frames 1x1.

#8 alpal

alpal

    Soyuz

  • -----
  • Posts: 3660
  • Joined: 15 Jun 2009
  • Loc: Melbourne Australia.

Posted 25 March 2014 - 05:45 PM

I'll post up my R,G, and B calibrated/registered frames in a while when I get back out to the observatory. Funny thing is, I don't usually have trouble with color. Oh, and I bin all frames 1x1.



OK - I now bin the colour 2x2 so I can get more signal.

This might be your problem.

#9 vpcirc

vpcirc

    Soyuz

  • -----
  • Posts: 3962
  • Joined: 09 Dec 2009
  • Loc: Merced CA

Posted 25 March 2014 - 05:48 PM

There should be more there. Hopefully I'll finish mine tonight. I think your issue may be gradient related, but I'm no expert. Here's mine from last year shooting with a refractor.

M 63

#10 Aircrftr

Aircrftr

    Apollo

  • -----
  • Posts: 1090
  • Joined: 29 Sep 2006
  • Loc: Claremore OK

Posted 25 March 2014 - 07:32 PM

OK.. Here is my calibrated/registered RED which is 8X900 Bin 1

Posted Image

#11 Aircrftr

Aircrftr

    Apollo

  • -----
  • Posts: 1090
  • Joined: 29 Sep 2006
  • Loc: Claremore OK

Posted 25 March 2014 - 07:34 PM

Here is the Green channel. All channels are same number and duration.

Posted Image

#12 Aircrftr

Aircrftr

    Apollo

  • -----
  • Posts: 1090
  • Joined: 29 Sep 2006
  • Loc: Claremore OK

Posted 25 March 2014 - 07:36 PM

Here is the blue channel.

Posted Image

#13 elbee

elbee

    Viking 1

  • -----
  • Posts: 689
  • Joined: 02 May 2009
  • Loc: Arizona

Posted 25 March 2014 - 07:40 PM

can you zip those raw RGB Master FITs files and put them in a cloud drive to share. the data looks nice!

#14 Aircrftr

Aircrftr

    Apollo

  • -----
  • Posts: 1090
  • Joined: 29 Sep 2006
  • Loc: Claremore OK

Posted 25 March 2014 - 07:43 PM

Here are all three channels put together in CCDStack (No luminance) and untouched from there.(accept conversion to the JPG format..)
Posted Image

#15 Aircrftr

Aircrftr

    Apollo

  • -----
  • Posts: 1090
  • Joined: 29 Sep 2006
  • Loc: Claremore OK

Posted 25 March 2014 - 07:48 PM

You know, I've just thought of something... The G2V color balance was obtained before I put my AO-X in the imaging train. I wonder if the coatings on that glass plate affected my color balance? I'll have to experiment with that.

#16 Aircrftr

Aircrftr

    Apollo

  • -----
  • Posts: 1090
  • Joined: 29 Sep 2006
  • Loc: Claremore OK

Posted 25 March 2014 - 09:50 PM

Here are the LRG&B Fits files from CCDStack. They are all averaged when combined.

#17 alpal

alpal

    Soyuz

  • -----
  • Posts: 3660
  • Joined: 15 Jun 2009
  • Loc: Melbourne Australia.

Posted 25 March 2014 - 10:14 PM

Hi Dan,
I'll download & process those files & put
the result on Astrobin.
It looks like you have plenty of colour signal.

cheers
Allan

#18 Aircrftr

Aircrftr

    Apollo

  • -----
  • Posts: 1090
  • Joined: 29 Sep 2006
  • Loc: Claremore OK

Posted 25 March 2014 - 10:39 PM

Thanks Alan. I'm looking forward to seeing what the problem is... In aviation, we call it "Loose nut behind the yoke" or a "short between the headset". We'll see. :)

#19 alpal

alpal

    Soyuz

  • -----
  • Posts: 3660
  • Joined: 15 Jun 2009
  • Loc: Melbourne Australia.

Posted 26 March 2014 - 12:02 AM

Hi Dan,
You're right - this one was tricky.
I'll post some smaller versions here instead of Astrobin.
I had to play around with the colours before combining the channels & do
quite a bit of LAB mode boosting.
Here is a full frame.

Attached Files



#20 alpal

alpal

    Soyuz

  • -----
  • Posts: 3660
  • Joined: 15 Jun 2009
  • Loc: Melbourne Australia.

Posted 26 March 2014 - 12:03 AM

Here is a rotated closeup.
I recommend that you always take colour as 2x2 binning from now on.

Attached Files



#21 Aircrftr

Aircrftr

    Apollo

  • -----
  • Posts: 1090
  • Joined: 29 Sep 2006
  • Loc: Claremore OK

Posted 26 March 2014 - 09:06 AM

Thanks Allan, It definitely looks better. I was taking advise from Gendlers book "Lessons from the Masters" that said (and I forget which one of the guys said it) that he likes to get 1x1 color data as well as LUM. It was in a section about the debate of doing 2x2 data Vs 1x1 for color. I guess I'll give 2x2 a try.
That doesn't really explain what's going on here tough. You would think that whether 1x1 or 2x2 that the colors should be closer to true color than this (leaving luminance out of the picture) So I wonder why all the yellow instead of the blues and reds that I get? Without all the manipulation you done, did it still turn out wanting to be looking all yellow or mainly like a straight luminance to you?

Dan

#22 alpal

alpal

    Soyuz

  • -----
  • Posts: 3660
  • Joined: 15 Jun 2009
  • Loc: Melbourne Australia.

Posted 26 March 2014 - 09:38 AM

Hi Dan,
the 2x2 binning increases the signal strength.
What I had to do was try & balance the colours better.
I tried combining as is but ended up with the same all yellow result.
I therefore stretched & adjusted the levels to give a boost to blue & some Green before combining the channels.

It was a bit of trial and error.
The background was adjusted with a mask & I had to use selected
colour to try & get rid of the magenta that resulted in the galaxy.
I actually played with all the selective colours & the neutral colours in selective colour.
In order to see what I was doing I boosted first with LAB colour.

The problem is that the colour pass band of the RGB filters is also affected by the efficiency
of the CCD at their respective pass bands.

You'd need to know all that technical information & do it
mathematically to end up with the right colours.
I just did it by sight.

It would help to have more colour signal to start off with
which is why I recommend 2x2 binning.


Another issue:
Also - the 4 FITS LRGB frames were not aligned & they wouldn't align in DSS until I changed them to TIFFS
& checked the Lum frame as a reference.

You need to make a master reference frame when stacking.

I have never had your above colour problem before.

I did sharpen at 2 pixels but used no noise reduction
so you can see a bit of noise at the edge of the galaxy.


#23 elbee

elbee

    Viking 1

  • -----
  • Posts: 689
  • Joined: 02 May 2009
  • Loc: Arizona

Posted 26 March 2014 - 01:01 PM

i think it is pretty nice data, dan. i use excalibrator for color ratios to apply in ccdstack. i used RGB weights of 1,0.85,0.96. i brought a Lum TIFF and combined RGB TIFF into photoshop. stretched with curves, applied some saturation and mild sharpening. there was a slight gradient i fixed with gradient xterminator. i did not apply any "color correction". i would be pleased to call this my own. i didn't see any instrument data in the FITS. did you collect the data with your AGO 17in?

Large version Aircrftr's Sunflower processed by elbee

Attached Files



#24 DeanS

DeanS

    Gemini

  • *****
  • Posts: 3236
  • Joined: 12 Jul 2005
  • Loc: Central Kentucky

Posted 26 March 2014 - 01:10 PM

Nice processing job Lee!

#25 dp297

dp297

    Mariner 2

  • -----
  • Posts: 285
  • Joined: 14 Dec 2010

Posted 26 March 2014 - 01:43 PM

I think u have beautiful data here....very nice. U just need to find the right way to control the color and the detail.
Here's my bit too on Pixinsight

Attached Files








Cloudy Nights LLC
Cloudy Nights Sponsor: Astronomics