Author Topic: Color Balance  (Read 2809 times)

Offline sreilly

  • PixInsight Padawan
  • ****
  • Posts: 791
    • Imaging at Dogwood Ridge Observatory
Color Balance
« on: 2014 May 12 07:49:36 »
Now that I've gotten a professional monitor, a slightly used NEC MultiSync LCD2490WUXi, which stores internally its color calibration, I see a big difference in my images. Some in color balance and mostly in background levels. But the big question remains in color balance. I do my color balance as suggested in Harry's video http://www.harrysastroshed.com/pixinsight/pixinsight%20video%20files/2013%20pix%20vids/ccfinal/ccfinal.mp4 which is very good. So now the question needs to be asked if the color calibration should be checked by any other means. I usually use all the data I have at my disposal after calibrating and culling the data by means of FWHM. I had been doing this by using a separate program "CCD Inspector" but of late have been using a script in Pix Insight. Another question should be asked about what terms to use but that's for later. After the data is culled the usually process of aligning the images and creating the master red, green, and blue images respectively is done. Then the color combine for the RGB master image. At this point it's likely I'll need to crop the image to get rid of the uneven edges before I can use DBE. At that point I'll typically do a background neutralization and then color calibration. If I have a galaxy I'll use the galaxy as my white point and if not make a aggregated preview and use as many stars as possible. So if all goes well we should have a fairly well color balanced image. What this has meant to me in the past is that if I open the Histogram process and set to track the view I should see all three peaks lined up mostly equally, correct? What is suggested as the recommended way to check that? Forget checking the web as there will be so many biased images it will make your head spin and then of course you could just adjust to what you like I suppose. Going back to the culled data, the final data set may have 14 red images, 17 green, and 12 blue so the idea of using any preset color ratios is not a good idea and then the sky conditions can affect that as well. There's eXcalibrator that uses G2V stars but I haven't really tried using that with PI lately.

I realize color balance is suggestive but getting as close to "real" is a goal I'd like to get to. I see many galaxies that seem way too blue and there are some very well know astro images that like to saturate their images for show but I'd like to get something close to realistic if possible. Any suggestions would be very welcomed.

Thanks
Steve
www.astral-imaging.com
AP1200
OGS 12.5" RC
Tak FSQ-106ED
ST10XME/CFW8/AO8
STL-11000M/FW8/AO-L
Pyxis 3" Rotator
Baader LRGBHa Filters
PixInsight/MaxIm/ACP/Registar/Mira AP/PS CS5

Offline Phil Leigh

  • PixInsight Addict
  • ***
  • Posts: 220
Re: Color Balance
« Reply #1 on: 2014 May 12 08:51:10 »
But - There is no "realistic" - none of these objects are illuminated by or emit the same wavelengths as our sky-filtered sunlight... so it's all highly subjective.

You can't say that a saturation-boosted image is not realistic... it might be - how could we possibly know one way or another? It's not like the frequencies aren't there... they are.

And what about Ha+RGB images or Ha/Oii bi-colour? - never mind Hubble pallette narrowband!

For RGB I use DBE+BN+CC to get a technically balanced baseline image on my calibrated monitor as per the Harry Page/Pi method - and then get artistic from there.

Take the Pelican or Soul Nebulae - they need to be very "red" - none of that wishy-washy salmon colour!. Andromeda needs the red and blue from the star-forming regions and young stars to look exciting.
Spiral galaxies look much more aesthetically pleasing (to me) with plenty of blue in the mix.

YMMV which is the whole point :-)