I read the link you provided plus the links within those links and it seems like the LRGB subject has been kicked around and around without much resolution (pun) The way I have always understood it, you shoot more luminance at the best times (good transparency and seeing and up towards zenith) This gives you great detail. Then shooting the RGB will always be lower resolution because its not all shot right at zenith. For instance, I always shoot the red channel first in the series because the red penetrates the thicker air near the horizon better than the other colors. The reason for shooting the RGB in 2X2 is that it would give better SN ratio and the fact that the chrominance only serves to colorize the detailed luminance channel. ( I actually do shoot RGB in 1X1 when the conditions will allow, to get better resolution when I can) So in essence, you don't need highly detailed RGB because the way it is combined (in the LRGB combine process) the chrominance doesn't add much (perceived by the human eye) detail, but only colorizes the detailed luminance channel. Now to my original question, It would be great to be able to match the sizes of the stars in each channel so that you don't get that multicolored star which is also bloated due to the RGB channels..(compared to the luminance) Doing this via deconvolution for each channel separately would be hit and miss and pretty time consuming in my opinion. It would be great to have a process that compares the PSF of the stars in each channel and run a deconvolution process that would match all three channels to the better stars in the luminance.
