Hi,
I'm working on an imaging of IC434 and I'm stuck on how to separate Alnitak from the main image and combine it with the rest of the stars. (Duncan I've progressed a little further since Friday but not much)
This is the first time I've tried this target so on reflection 10 mins subs were perhaps too long however it does present a challenging processing exercise for me. I'd planned to process the work flow as follows, I might be wrong and I'm happy to be correct.
Separate the stars and correct colour balance, reduce Alnitak
Stretch nebula and add Ha as luminosity.
Initial processing involved stacking and alignment in AA5 and stretching using arcsinh() in FITS liberator.
(http://farm9.staticflickr.com/8064/8204359392_dbd5f168d2_b.jpg)
IC434-Stars
(http://farm9.staticflickr.com/8070/8204346396_62a89372e8_b.jpg)
Without stars
(http://farm9.staticflickr.com/8200/8204346442_1114615d93_b.jpg)
Is this the best approach?
How else should I do it?
Is there something I'm missing?
PSD CS4 file can be found here https://www.dropbox.com/s/ewli8jxmrsp8ut2/ic434_original.psd (https://www.dropbox.com/s/ewli8jxmrsp8ut2/ic434_original.psd) (110mb)
Thanks
Ivor,
I find it difficult to believe Altinak is so oversaturated with only 10 min exposures - you're not using a Hyperstar! Check your subs to see. If I'm right then by changing your stretching you will make a huge difference to this image.
Mark
OK I'll try reducing the stretch and post the results up later, day job interferring again :(
I'm still struggling to capture the star halo as well as the star core, I've uploaded the unstretched FITS to see if anyone else can do it?
https://www.dropbox.com/s/ukb14077xgsckqf/ic434_RGB_600_pre_m_clipped.rar (https://www.dropbox.com/s/ukb14077xgsckqf/ic434_RGB_600_pre_m_clipped.rar)
Ivor,
Your file was monochrome - was that intentional? [Later Edit: Actually it's not monochrome but Iris opens it as monochrome and PixInsight fails to open it at all. I guess the format is a slightly strange variation.]
In any case, the Iris asinh function works a treat, as always ;)
(http://www.markshelley.co.uk/webdisk/ivor_horse.jpg)
You've got some good data there!
well there you go, what a difference
I'm glad someone can at least get something out my data, one consolation at least I know my capture workflow works.
Not sure why the file can't be read AA5 and FITS open it. I've separated the channels here
https://www.dropbox.com/s/826ej8jc9m4uaoq/mkiii.rar (https://www.dropbox.com/s/826ej8jc9m4uaoq/mkiii.rar)
I must be doing something wrong in FITS Liberator, Duncan gave me a brief demo of IRIS on Friday, I think this will need further investigation as 5 mins probably wasn't quite enough!!
Aha, the colour data was in the original file. I opened it in FITS Liberator; saved each plane separately; combined into a colour TIF in PixInsight; saved it and pumped it through the Iris asinh function:
(http://www.markshelley.co.uk/webdisk/ivor_horse_colour.jpg)
Just a very quick proof of concept - a bit more tweaking would improve it no end.
Clearly I need to sit down with someone and examine my current processing workflow.
Trying to familarise myself with IRIS asinh function, I've started with asinh 0.005 30 and I get a Andy Warhol effect. My histogram is in the middle but narrow.
Also when I type another asinh command is it compounding it onto the last and if so how do I reset back to the beginning?
Hints please!!!
To use Iris asinh function:
1) Set black level - this is critical. Do so by selecting a rectangle on the blackground and executing the command "black" in the command dialog.
2) The way I personally do the next bit is to adjust the sliders on the threshold dialog to bring out the faint nebulosity - this will also completely oversaturate the stars.
3) Now bring up the colour stretching (asinh) dialog and adjust the top slider until the stars are desaturated sufficiently - I found 0.0070 worked well for your image. I never touch the bottom slider - it is unncessary if you perform step 2 first
4) Asinh cannot be undone, but the dialog slider gives plenty of interactive opportunity to play with it before committing.
5) You can choose to do RGB colour balancing before or after asinh because asinh does not alter the colour balance of the pixels. This is the unique feature of the Iris asinh function.
Hope that helps.
Mark
That's looking a lot better now, processing is a maze!
Quote from: The Thing on Nov 22, 2012, 09:56:31
That's looking a lot better now, processing is a maze!
Agreed on both fronts !
I think Alnitak and also Gamma Casseiopea are the two worst offenders in the sky and if you can master those then you've gained a lot of valuable experience.
Alnitak core is still well blown so you may have to use some sort of reduction just on the core.
Robert
Quote from: RobertM
Alnitak core is still well blown so you may have to use some sort of reduction just on the core.
I've been thinking about another approach to this all-too-common problem using three pieces of easily available information:
1) In the area just outside the saturated core, we can extract the correct star colour.
2) From the image as a whole (the non-saturated stars) we can estimate the FHWM of the star field.
3) From the unsaturated part of the star we have an intensity curve.
Now mathematically we can do a 2D curve fit to the unsaturated part of the star and create the right shape synthetic star (i.e. a star with the correct peak value and correct FHWM) of the right colour to replace the clipped values.
Voila!
Still struggling with the first asinh process, I'm failing at the first hurdle, I can't get the mouse pointer to draw a rectangle to do the black command, is there a button I need to press first?
found it, you have to be x1 scale.
Quote from: Ivor
found it, you have to be x1 scale.
Oh yes, I've hit that before - anything that involves selecting or adding points on image will only work at 1:1 scale.
Starting to get the the stretching correct in Iris however saving it is presenting a problem, the save function in the file menu and button don't appear to save the visualised image?
Tutorial on the Iris don't mention anything so I must be missing obvious step?
Another hint please?
If you save as a BMP then you get exactly as it appears on the screen.
If you save as a TIF then first set the threshold sliders to 0 and 32767 and then multiply image by a constant "mult n" and maybe add an offset "offset n" so it looks right on the screen before saving.
Mark
After a number of emails with Mark (thanks again) and much plotting in Excel to work out how asinh works I got it straight in my head and I think I can do it correctly in FITS liberator now.
After various processing in PS this is the end result of
RGB 8 - x 10mins Ha - 11 x 10mins
Stacked in AA5
FLT110 ST8300M
(http://www.jovets.co.uk/sites/default/files/styles/large/public/ic434_RGB_600-CG--HA-IV.jpg?itok=peBvBeWr)
That's looking really good Ivor. The structure in the nebulosity behind the horsehead has come out really well.
I'm not too sure about the colour however - maybe it's a lack of blue - I'm not quite sure.
You have succeeded in getting Altinak under control.
Overall it's a very good result. Well done
Mark