Log in

View Full Version here: : M20 .. no frost


Jeffkop
07-08-2008, 12:52 PM
Hi everyone.

OK, back in action after the frosting episode last week .. look back to M20 31/7
So I decided to do M20 again, but this time armed with the processing tips from Jase and Greg and others from my post "Does only exposure determine sub times", in particular the info that I learned from this link http://www.rc-astro.com/resources/ta...StarColors.htm (http://www.rc-astro.com/resources/talks/PreservingStarColors.htm) about preserving star colours. (PSC from now on)
Well I dont know how well I did that.

Info about the image:

SUBS
Lum .. 4 x 20min 1x1
Red .. 4 x 15min 2x2
Green 4 x 15min 2x2
Blue 3 x 15min 2x2 (the object disappeared behind the huge jacaranda in the backyard before the 4th one finished)

Following the guidelines in the PSC, I did a Digital Development (DD) on the luminance and then the same on the LRGB image, all the while making sure that the histogram showed no clipping. This method is in stark contrast to my previous ways, however I was warned about black clipping in previous posts and this was about my only consideration in the histgram regard.

Apart from that I have done little else to this image, and to be honest, little else I have tried made any significant difference to its quality.

I am still unrewarded as far as deep colour goes compared to many posts of this object with less imaging time than this.

Ive a sneaky suspicion that I may have been better served to do 15min lum and 10 min RGB, perhaps the luminance signal has washed out the colours.

With so many variables to consider in this processing stage, its difficult to pick a strategy and refine it. So many things affect so many other things. How on earth did these guys who are good at it sus it all out ???

So I dont want to turn this forum into a help Jeff forum, but anyone who can comment is most welcome.


Thanks

beren
07-08-2008, 03:59 PM
:) Very nice Jeff , nice star colours too {thanks for the link}. Doh :doh: I had a frosting experiance last night that cut short my night :lol:

gregbradley
07-08-2008, 06:26 PM
Oohh nice image. A stunner.

If you want to enhance the stars more (your call they don't really need it but it can be an interesting experiment to see if you like it).

Then try this:

1. Select a more bloated star with the colour range tool and adjust the slider until it selects most of the stars.

2. Increase the size of the circles of the selection by select/modify/expand (say 4-6 pixels), also you can use select/feather (4=8 pixels) here as well to spread the effect a bit and blend it into the surrounds of the stars

3. Now you can increase saturation (perhaps do this last).

4. Now you can use the minimum filter filter/other/minimum (1 pixel) watch the stars shrink.

5. Deselect.

Another little trick is once something is selected hit control H to hide the selection circles which can be distracting and you can see the result of your processing more easily.

You can overdo this easily.

Or you can use Noel Carboni photoshop actions and one of these is make stars smaller and it works reasonably well (you can run it several times).

Greg.

Bassnut
07-08-2008, 06:57 PM
Thats an excellent image Jeff, suprisingly deep, your being hard on your self, and the colour looks pretty good.

One thing you missed. ALL the detail of an image comes from the lum channel. The RGB can be relatively poor (as long as its not too noisy), the biggest improvement you can make to this image now is megadata on the Lum channel. The 1 hr or so exposure each on the RGB youve done is fine, just a tweak in PS on RGB saturation is all thats required (or if yr keen, blending lum into RGB to boost colour ). I would recommend you try a few more hrs just on Lum to allow more sharpening without noise appearing. Its not unusual to take 10 times longer exposure on Ha/Lum (for Lum channel) over RGB, not the equal exposures youve done. I have often taken 10 hrs or more Lum/Ha to go with 1 hr RGB exposures.