narrow band processing without rgb

[Rkonrad]

Unfortunately, my availability to go out and take my own images are very limited - so I have been getting practice through processing other's data. In particular this one is Scott Rosen's data from https://groups.yahoo.com/neo/groups/dsl ... ssing/info. I had been experimenting on the pacman nebula, just the Ha and Oiii data and was wondering how I achieve colour out of this. In the end I -

1. Opened up the Ha data as L and R channels, increased the R ratio an amount (can't remember what) to give it a red colour. I then processed this fully.
2. Then I opened up the Oiii data as G and B channels. I increased this ratio somewhat more. I processed this fully as well.
3. Combined the two with the blending mode ("if fg is dark, use background" - something like this) thus getting my final version.

I quite like the detail and the way the blending revealed the Oiii. I gave quite a bit of emphasis on the Oiii. But the colours are unusual and unlike anything I've seen. It seems somewhat unatural. Adding the the rgb data was just too tricky for me to work with so I left it out. Here is the image. I'd love some comments. Thanks.

General permission granted to post to other websites (with attribution to
Scott Rosen at http://www.astronomersdoitinthedark.com/
and DSLR-AIP at
https://groups.yahoo.com/neo/groups/dsl ... ssing/info)

Cheers,

Richard

This may be a better image quality than the attachment: https://www.flickr.com/gp/rkonrad/bHAk2F

[admin]

Hi,

I've been waiting a good few days for access to the Yahoo Group, but to no avail.
If you'd like to share the data privately with me, that would speed things up...

[admin]

When it comes to narrowband imaging, there is obviously no one "correct" or "natural" way of rendering the colours, so really, anything goes.
Steve Cannistra has a preferred method of combining Ha and OIII into bi-color images, which he describes here; http://www.starrywonders.com/poss2.html

Or go with any of the techniques 3 discussed here; http://www.lightvortexastronomy.com/tut ... tions.html

At the end of the day, it really doesn't matter as there really are no rules, just what you find pleasing.

Personally, I prefer just loading Ha as red and OIII as blue in the LRGB module (the green channel will be automatically synthesised), which should get you something like this;

Pacman.jpg (462.15 KiB) Viewed 9586 times

(data acquisition by Scott Rosen, provided by DSLR-AIP)

Note, you could also create a synthetic luminance frame by weighting the Ha and OIII properly, but since the exposure times weren't that different I lazily didn't bother...

[Rkonrad]

Thanks!

I'm working on the bicolour option but struggling as I'm trying to transfer the photoshop instructions to Gimp 2.9. I've never worked with layers before - except with StarTools. Will let you know if I have any questions.

Richard

[Guy]

Hi Ivo and Richard,

I had a quick go at the variation on Steve Cannistra's bi-colour technique described in the Light Vortex reference - with (Ha : Ha*OIII*1.5 : OIII) using StarTools' LRGB and Layer Modules. I used POSS2 data as Steve had. I've added the method I used to the Special Techniques section of Layer Module Use.
The method produced similar results to that on Steves web site - but please let me know if I've missed something in the 'translation' from PixInsight.
It was very straightforward - but that may be because I have oversimplified things.
The biggest problem I had was dealing with the different image sizes in the POSS2 data - which I did using the approach described in the Background Notes of LRGB Module Use.

Guy

[Rkonrad]

Hi Guy,

Thanks for your advice. I followed your tutorial quite closely and got these examples. The first one was with my processed images of Ha and Oiii - the second one with with unprocessed images. With the latter I let startools set the colour balance. The colours seem quite wild and something doesn't seem right. Perhaps there's something missing in your instructions or I didn't follow correctly? Here are the examples:

https://www.flickr.com/gp/rkonrad/P5EAns (pre-processed starting from nonlinear)

https://www.flickr.com/gp/rkonrad/VJ8295 (starting from linear)

These are my pre-processed Ha and Oiii

https://www.flickr.com/gp/rkonrad/8cyj4C
https://www.flickr.com/gp/rkonrad/VVn94G

Thanks!

Richard

[Guy]

Hi Richard,
Don't get the wrong idea - I'm learning here too.
I've adjusted the method now - reflecting the Light Vortex PixInsight approach better I think.
I tried the revised method with your pre-processed Ha and OIII jpegs with minimal processing and got the result below - which is perhaps a bit more 'natural'.

NGC281-HaOIII-Cannistra-small.jpg (81.04 KiB) Viewed 9523 times

Working with the original linear fit or tiff files may produce better results.

Edit: Here is the Ha:OIII:OIII (R:G:B) version for comparison - minimal processing.

NGC281-Ha- OIII- OIII-small.jpg (97.59 KiB) Viewed 9521 times

Here is the Ha:Interpolation:OIII (R:G:B) version - letting the LRGB Module interpolation algorithm to create the Green channel - minimal processing.

NGC281-Ha-Interpolation-OIII-small.jpg (90.57 KiB) Viewed 9517 times

Don't forget that with narrow-band techniques like this there is no 'right' result. You can make further adjustments in the Color Module if you want.
Let me know how you get on.

Guy

[Rkonrad]

Hi Guy,

Well in the end, I found I got the best result (for me) simply loading the pre-processed Ha under red, and the Oiii under green and blue. I went to the colour module and masked all but the stars, selected "sample" and worked from there with dark saturation controls. I have the rgb data for the stars but couldn't get a pleasing result so just removed them. Here is the result:

https://www.flickr.com/gp/rkonrad/N7M1K0
and with stars - https://www.flickr.com/gp/rkonrad/0n0YLc

I also tried to add a luminance file to this , using a blended Ha with Oiii (fg is dark use background) but the result seems too unnatural to me.

https://www.flickr.com/gp/rkonrad/1qVce6

Ah...its great to work with such amazing data which really spoils me for when I actually get out to do some imaging of my own!

I will try your new methods shortly.

Cheers Richard