#1  
Old 12-11-2012, 12:11 AM
PRejto's Avatar
PRejto (Peter)
Registered User

PRejto is offline
 
Join Date: Jan 2011
Location: Rylstone, NSW, Australia
Posts: 1,397
Stars Lack Colour

I'm wondering if some of you experienced imagers might be able to tell me why I am able to get plenty of colour in photos of the Helix nebula, but not in the stars?

Using my TEC140/Moravian G2-8300 camera my image scale is 1.14 arcsec. I've not collected a huge amount of data.....5 min sub frames of 1 hour each in RGB + L using the Hutech IDAS. The nebula has good colour, but even completely over saturating to the point of destroying colour in the nebula hardly generates any star color. I've binned 2x2 for RGB, then processed in CCDStack, and used the mean combined stacks to generate a colour image.

I suppose a problem that might be related is that the DDP stretched images look pretty good (except minus star colour), but if I try to open an unstretched image in Photoshop I really cannot generate a useable colour image using levels; there appears to be very little usable signal...so little in fact that the entire image seems compressed into a single vertical line in the histogram. If, however, I save the image stretched by DDP, etc, that image looks fine in Photoshop, though somewhat noisy.

Are all of these issues just due to not enough sub exposre time, either lengths of sub frames, and or the total # of subs?

Thanks for any help!
Peter

Last edited by PRejto; 12-11-2012 at 05:33 AM.
Reply With Quote
  #2  
Old 12-11-2012, 06:39 AM
Rigel003's Avatar
Rigel003 (Graeme)
Registered User

Rigel003 is offline
 
Join Date: May 2006
Location: Adelaide, South Australia
Posts: 1,079
Doing a search on Helix images seems to reveal the same problem pretty much everywhere. You should select the stars only before trying to increase saturation. There are lots of descriptions of this technique on the web.
Reply With Quote
  #3  
Old 12-11-2012, 07:43 AM
Geoff45's Avatar
Geoff45 (Geoff)
PI rules

Geoff45 is offline
 
Join Date: Jul 2006
Location: Sydney
Posts: 2,631
See http://www.astropix.com/HTML/J_DIGIT/STARCOLR.HTM#
Reply With Quote
  #4  
Old 12-11-2012, 07:04 PM
PRejto's Avatar
PRejto (Peter)
Registered User

PRejto is offline
 
Join Date: Jan 2011
Location: Rylstone, NSW, Australia
Posts: 1,397
Thanks very much. I appreciate the input! Pretty obvious cause after reading about it, but I couldn't get the solution in the second post to work. I think I need better colour data for the stars. I might try some shorter exposures unbinned and see where this gets me.

Peter
Reply With Quote
  #5  
Old 22-11-2012, 05:24 PM
astrospotter (Mark)
Registered User

astrospotter is offline
 
Join Date: Jan 2010
Location: San Jose, CA, USA
Posts: 146
I tend to use one layer of fairly short exposures like 12x30sec for each of RGB and this is my 'star color' layer. When all of R/G/B are very close to 1.0 (full CCD range) there is no color data any more. To get star colors one must avoid over-saturating any of the R/G/B layers. Even if one layer is saturated to max CCD ADC or even over 80-90 percent you distort the color data.

You can then separate this color layer out after much boost to it's saturation to peel off the color data without the luma (separate out to Yuv sort of layers). This color data can be combined back into your star luma data for more colorful stars.

Be careful in your stretching of r/g/b and use a curve that brings up lower levels but does not as greatly amplify over 40% signal. This curve must always be increasing slope (gain) but just not as much gain to the over 40 percent already hot stars.

There are likely many tutorials but I had to mess with this problem for many years especially if you want those stars to have nebula all about them. Every images is a bit different so there is no one stock formula except in general terms mentioned above.

good luck,
Mark

www.astrospotter.zenfolio.com is where my images live
Reply With Quote
  #6  
Old 22-11-2012, 10:05 PM
PRejto's Avatar
PRejto (Peter)
Registered User

PRejto is offline
 
Join Date: Jan 2011
Location: Rylstone, NSW, Australia
Posts: 1,397
Hi Mark,

Many thanks for your thoughtful reply. I follow what you are saying about over-saturation, but I'm confused as to how one works around the problem even if short RGB exposures have given one a good colour star layer. I assume that most images wil have over saturated stars if they are bright and the object of interest is fairly dim just because of the sub frame length needed to catpure the dim oject. I also understand that if the luminance layer has gray, and not white, colour can be added, so, is it a question of removing over saturated stars and replacing them, or is there a method of blending an over saturated luminance layer with RGB data and actually colouring the over saturated stars? If you could point me towards a tutorial that roughly follows your method I would be very grateful!

Thanks,

Peter
Reply With Quote
  #7  
Old 23-11-2012, 06:08 PM
astrospotter (Mark)
Registered User

astrospotter is offline
 
Join Date: Jan 2010
Location: San Jose, CA, USA
Posts: 146
I'm afraid I don't speak photoshop, or PS, lingo and I don't know what you are using. I have been using PixInsight, or PI, and it fits my mindset but it has a tough learning curve. What I say below is not a good answer and sorry about that but it gives you some hints. A lot of well done tutorials that are about this PixInsight software that I highly recommend can be found on 'Harry's Astro Shed' so if nothing else from this conversation you may consider use of PixInsight but I do caution you it is kind of confusing so Harry really helps in that area with his tutorials.

You will not find this below on those great tutorials but many of the things I do have a basis in the raw tools Harry explains so well.

Some short thoughts on the tool that combines things goes like this:
I usually have 1 or two layers of very deep narrowband with Halpha for sure and then sometimes OIII. Through assorted tricks that this PI tool has that have odd names I do my best to create narrowband layers that do not have stars as white hot dots. One tool is called 'Morphalogical Transform' and that certainly has one of the most odd name of their tools but it can be used with a starmask to greatly subdue the stars in the Ha (narrowband) layer.

The Ha layer can be combined with your over-saturated nice star image that is processed in RGB by basically adding components of R,G,B together using a tool called PixelMath (some other astrophoto packages also have the PixelMath tool. Suppose you have a nice Halpha layer and you want to have it be the 'Ha Red'. Well more or less the white image you have of Ha can be 'mixed' with some other image where 'Ha' below is the image of white and we will use multiplication to 'add' it into an R, G, B layer that is from the stars. This technique can be done on several layers at once.

Red = R + (0.8 * Ha)
Green = G (We don't add in any Ha layer as green)
Blue = B + (0.2 * Ha)

So above we use a mixing formula where we use Red mostly and then 1/4 of the 'weight' of the red we also add in for Blue.

What I have 'glossed over' is how to prevent the mixing of the Ha with the RGB from leading to little rings around the stars. That is a complex issue and sorry I don't have any easy way of explaining that.

Harry has a tutorial on mixing in Ha to galaxies but that is a bit different than mixing in nebula. He adds tutorials all the time so I don't know what he has these days.

Good luck,
Mark
Reply With Quote
  #8  
Old 24-11-2012, 10:07 AM
PRejto's Avatar
PRejto (Peter)
Registered User

PRejto is offline
 
Join Date: Jan 2011
Location: Rylstone, NSW, Australia
Posts: 1,397
Hi Mark,

Many thanks for making this attempt to explain what is obviously complex. I've heard very good things about Pixinsight so maybe I will give it a go, but I'm heavily invested in Photoshop and would like to explore more ways with that tool to see if I can get to the same place.

I mentioned in my first post that I was having trouble stretching unstretched images coming out of CCDStack. I've resolved the error I was making so I am making some progress. Your explanations have given me some hints that I will further explore. Thanks again.

Peter

PS I looked at the photos on your site. Very Nice!!!
Reply With Quote
Reply

Bookmarks

Thread Tools
Rate This Thread
Rate This Thread:

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off

Forum Jump


All times are GMT +10. The time is now 05:40 AM.

Powered by vBulletin Version 3.8.7 | Copyright ©2000 - 2024, Jelsoft Enterprises Ltd.
Advertisement
Testar
Advertisement
Bintel
Advertisement