Quote:
Originally Posted by Atmos
Although it does depend on the data and object a bit as to whether I do it during the linear stage or not but usually I'll stretch both the RGB and Ha data separately, get their histograms similar and then run in PixelMath R*R+Ha*(1-R) on the red channel.
What this does is keep the brighter regions mostly R (especially for stars) but the fainter regions become more and more Ha dominant. This particular expression doesn't work while in linear, sometimes I'll do a LinearFit on Ha using R as a reference and then use the Max(R,Ha) expression which just takes the brighter nebulosity of Ha and adds it into the R. It isn't always as neat though as it sometimes doesn't make the stars look that nice; they like they don't match the nebulosity.
|
Thanks Colin - That seems really useful especially for bringing out the faint dust. Maybe a combination of these two (Marco and this) methods could be interesting... Props for a pixinsight only combination