Bright star control?

[ollypenrice]

How does everyone else deal with this? I have taken short exposures in the past but also tried this with some success;

Process the LRGB image till nearly finished, leaving the bright star (in this case the dreaded Alnitak) as a white supernova the size of a fried egg on the screen.

Alnitak was already saturated on the linear L data but not on the RGB so I took that and did a special curve - see below - intended to brighten the environs of Alnitak far more than its core. I just lifted the bottom of the curve steeply as usual but rolled it off very early with a strqight line to the white point. It took two or three of these curves to get to this stage. By now the area round Alnitak is not too much fainter than in the LRGB image but the star is resolved into a double. I have some hope of blending them, therefore, and put the Core Curve RGB underneath. The rest is ludicrously simple. I took a large soft erasor bigger than the over exposed area and set only to about 10%. Dab it over the area centred on Alnitak. Reduce the brush diameter and dab again. And so on and on. I wouldn't have expected it to work, really, but it generally does. In fact it worked better for me than a proper layer mask.

Olly

[Catanonia]

Nice work mate and well thought out. Might give this a go myself next time.

Any chance of a large image to see what you were doing.

I tend to use layer masks, or PI wavelets on the selected area and merge over the top of the bad image.