ICEINSPACE
Moon Phase
CURRENT MOON
First Quarter 53.1%
|
|
06-09-2015, 12:36 PM
|
|
Registered User
|
|
Join Date: May 2014
Location: Darwin
Posts: 737
|
|
RGB ratios
Hi Spacers,
Is there a way to determine RGB ratios for colour processing?
I've tried eXcalibrator but ended up with some weird ratios
Eg RGB 1:1.48:1.77
Saw something on the net that says to use G2V stars but no method attached.
I have CCDStack, is there a process in this?
Cheers
George
|
06-09-2015, 12:57 PM
|
|
PI cult recruiter
|
|
Join Date: Apr 2010
Location: Brisbane
Posts: 10,584
|
|
There are lots of ways to determine RGB ratios but I don't think any is a magic bullet.
Some of the ones I have tried include: - eXcalibrator
- using a G2V reference star
- using pre-calculated G2V ratios adjusted for extinction
- roughly lining up the R, G & B histogram peaks
- Unlinked stretch in PixInsight
- ColorCalibration in Pixinsight using a variety of white references (a galaxy, the star field, the whole image)
I didn't find that any of these did a good job in every situation. These days I use whatever gives me a result I like. Sometimes I blend a couple of different versions.
Cheers,
Rick.
|
06-09-2015, 03:21 PM
|
|
Registered User
|
|
Join Date: Jan 2012
Location: Melbourne
Posts: 3,630
|
|
Quote:
Originally Posted by vlazg
Hi Spacers,
Is there a way to determine RGB ratios for colour processing?
I've tried eXcalibrator but ended up with some weird ratios
Eg RGB 1:1.48:1.77
Saw something on the net that says to use G2V stars but no method attached.
I have CCDStack, is there a process in this?
Cheers
George
|
Hi George,
I've always wondered about that.
Also - what about using different exposure times for each of Red, Green & Blue
so that the final result is always 1:1 ?
( you would also have to allow for the band pass of the filter used in front of all other filters
such as the CLS-CCD filter that I use &
the quantum efficiency of the CCD chip at different wavelengths )
cheers
Allan
|
06-09-2015, 04:48 PM
|
|
Mostly harmless...
|
|
Join Date: Jul 2008
Location: Brisbane, Australia
Posts: 5,719
|
|
Quote:
Originally Posted by RickS
There are lots of ways to determine RGB ratios but I don't think any is a magic bullet.
Some of the ones I have tried include: - eXcalibrator
- using a G2V reference star
- using pre-calculated G2V ratios adjusted for extinction
- roughly lining up the R, G & B histogram peaks
- Unlinked stretch in PixInsight
- ColorCalibration in Pixinsight using a variety of white references (a galaxy, the star field, the whole image)
I didn't find that any of these did a good job in every situation. These days I use whatever gives me a result I like. Sometimes I blend a couple of different versions.
Cheers,
Rick.
|
I'm always getting in trouble for my crazy analogies, but it almost sounds like polar alignment. The search for the holy grail of perfection, that isn't ever quite possible!
Rick, I've been in love with "linear fit" within PI recently. I know its not a proper color calibration tool, but it seems to do a great job of getting R, G and B in the same ball park before serious processing begins.
|
06-09-2015, 04:53 PM
|
|
Narrowfield rules!
|
|
Join Date: Nov 2006
Location: Torquay
Posts: 5,064
|
|
I had endless trouble with this. Now I just use Astronomy action tools (Action plug in for PS) http://www.prodigitalsoftware.com/As...l_Version.html
The Construct RGB Image from Channel Files is a few clicks and its done, lifes too short to muck around.
|
06-09-2015, 05:03 PM
|
|
PI cult recruiter
|
|
Join Date: Apr 2010
Location: Brisbane
Posts: 10,584
|
|
Quote:
Originally Posted by RobF
I'm always getting in trouble for my crazy analogies, but it almost sounds like polar alignment. The search for the holy grail of perfection, that isn't ever quite possible!
|
The problem is that it isn't a single one-off ratio. Atmospheric effects mean that the sensitivity of a scope/camera at different colours is changing all the time.
Quote:
Originally Posted by RobF
Rick, I've been in love with "linear fit" within PI recently. I know its not a proper color calibration tool, but it seems to do a great job of getting R, G and B in the same ball park before serious processing begins.
|
Nothing wrong with that, Rob. Using LinearFit will just balance the channels. I often use this technique for narrowband. An unlinked stretch does much the same thing.
Cheers,
Rick.
|
06-09-2015, 05:22 PM
|
|
Drifting from the pole
|
|
Join Date: Feb 2013
Location: Brisbane
Posts: 5,443
|
|
Apologies to the OP for a slight off-topic...
Surely this would be dependent on the subject? Something like Eta Carinae is off the charts in red with my modded DSLR, but signal is signal I'm not sure I'd want to balance that out to look nice or is that the point?
|
06-09-2015, 06:18 PM
|
|
Mostly harmless...
|
|
Join Date: Jul 2008
Location: Brisbane, Australia
Posts: 5,719
|
|
Quote:
Originally Posted by Bassnut
|
Is that what Carboni's tools has grown into Fred, or completely unrelated?
|
06-09-2015, 06:19 PM
|
|
Narrowfield rules!
|
|
Join Date: Nov 2006
Location: Torquay
Posts: 5,064
|
|
Quote:
Originally Posted by RobF
Is that what Carboni's tools has grown into Fred, or completely unrelated?
|
No, thats the one, Carboni's tools. OK, maybe Carbonis name seems to have disappeared.
|
06-09-2015, 06:36 PM
|
|
Big Scopes are Cool
|
|
Join Date: Jun 2007
Location: SE Tasmania
Posts: 4,541
|
|
I think that Carboni tool justs combines the channel images into an RGB 1:1:1. I don't remember it dealing with colour balance.
|
06-09-2015, 06:38 PM
|
|
Narrowfield rules!
|
|
Join Date: Nov 2006
Location: Torquay
Posts: 5,064
|
|
Quote:
Originally Posted by peter_4059
I think that Carboni tool justs combines the channel images into an RGB 1:1:1. I don't remember it dealing with colour balance.
|
Thats interesting, and possibly disappointing. I dont know. Seems to do a better job than I do manually.
|
06-09-2015, 06:43 PM
|
|
Big Scopes are Cool
|
|
Join Date: Jun 2007
Location: SE Tasmania
Posts: 4,541
|
|
The description of the action says it prompts you to do a colour balance.
|
06-09-2015, 08:01 PM
|
|
PI cult recruiter
|
|
Join Date: Apr 2010
Location: Brisbane
Posts: 10,584
|
|
Quote:
Originally Posted by Camelopardalis
Apologies to the OP for a slight off-topic...
Surely this would be dependent on the subject? Something like Eta Carinae is off the charts in red with my modded DSLR, but signal is signal I'm not sure I'd want to balance that out to look nice or is that the point?
|
G'day Dunk,
I think the discussion covers both cases. Figuring out RGB ratios for an "accurate" image is not a trivial or easily repeatable process. Coming up with something that looks nice isn't that easy either. But you are correct... they are different goals.
Cheers,
Rick.
|
06-09-2015, 09:33 PM
|
|
Ultimate Noob
|
|
Join Date: Aug 2011
Location: Melbourne
Posts: 6,999
|
|
As has already been stated, there is no "magic scale" that will work in every situation. The Astrodon LRGB Gen II filters are supposed to be 1:1:1 for either Kodak or Interline sensors. Even though the filters may be 1:1:1, it doesn't mean that what you are actually capturing is correct!
That may sound strange but depending on where it is in the sky at the time changes the colour slightly (near horizon or zenith), imaging on different nights and seeing conditions, water vapour or dust in the atmosphere. Of course there is also dust between us and what we're imaging!
PI has some great tools for colour calibration, mostly I guess you just need to be selective about what you are using as your calibration. One good way to go, Aladin Sky Atlas!
http://aladin.u-strasbg.fr
That way you can manually align flux values in RGB (well actually Johnson-Cousins Photometric UBVRI)
|
06-09-2015, 10:53 PM
|
|
Drifting from the pole
|
|
Join Date: Feb 2013
Location: Brisbane
Posts: 5,443
|
|
Quote:
Originally Posted by RickS
G'day Dunk,
I think the discussion covers both cases. Figuring out RGB ratios for an "accurate" image is not a trivial or easily repeatable process. Coming up with something that looks nice isn't that easy either. But you are correct... they are different goals.
Cheers,
Rick.
|
Thanks Rick, good to know it's a common struggle!
|
Posting Rules
|
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts
HTML code is Off
|
|
|
All times are GMT +10. The time is now 11:16 PM.
|
|