Quote:
Originally Posted by Ross G
Hello Mike and Trish.
That is one amazing galaxy photo.
So sharp with excellent detail.
I love the colours in your final version.
Ross.
|
Quote:
Originally Posted by Retrograde
Very nice as always M & T.
Being from the 'too much saturation is barely enough' school myself I prefer the second one  (but both are great). |
Thanks Ross and Pete. Seems the red version wins hands down.
We wanted to be able to say in simple English, and in a package-independent way, where the green tinge came from in the first version, and what we did about it.
We started with a publication-ready RGB, and a publication-ready H-alpha. In version 1, we took the average of the "H-alpha as red", and the red from RGB. That felt right, and worked fine where there was strong H-alpha, but where there was no special nebulosity, averaging the natural red with nothing left a blue-green cast.
In version 2, we simply applied an arcsinh stretch to the red channel from the first version, increasing the red in the non-H-alpha parts by roughly a factor of two, so they came out looking normal again, with no cast. The result was that the H-alpha rich parts came out very strongly red, showing where the H-alpha was, without burning out.
There are a thousand legitimate ways of achieving the same result (natural-looking galaxy where there is no nebulosity, but strong representation of the nebulosity). This way has the advantage of being very easy to explain not just what we did but why it worked, in a package-independent manner.
Best,
MnT
