Another Go At Some Old Photos with PixInsight

[Enterprise]

now that is what i call a HD difference!

[Barry]

Holy moley! That Rosette it stunning!

[rcj]

marvelous attempts via PI! You'd finally believed the holy grail of gradient removals....DBE since film days was used extensively. Because of the custom bg modeling character, once or even twice applied, with proper control boxing will remove most gradients, unless it is affected adversely by asymmetric local light cones.

the backgrounds are kind of clipped though as a result of DBE after-effects, and it may have sub-dued the faint nebulous regions (end points of rosette and m42). It is always good to use STF to auto-stretch the image first (remember to click 'A' symbol) to check for how much faint region data you potentially have in the stacked image. Colour calibration / background neutralization has to be performed too, to set the base. Although with time, you will find that the handy Pixelmath tool will be best for performing this.

Some equations that I have tested ok for neutralization would be:

$target - (Med( background ) - Med( background[2] ))
$target - (Med( background ) - Med( background[2] ))
$target

for the R, G, B channels respectively. This would be best on a single preview sample, which is good enough if data is taken from dark skies, and if the region is not over nebula rich.

Otherwise, another customized set of equations that can fulfill the second case would be:

$target - (Avg( Med( background1 ), Med( background2 ) ) - Avg( Med( background1[2] ), Med( background2[2] )))

$target - (Avg( Med( background1 ), Med( background2 ) ) - Avg( Med( background1[2] ), Med( background2[2] )))

$target

For both sets of equations, the Blue channel can be left as it is ($target).

[KKK26]

Super nice! M42 is so colorful and alot of details. Thanks

[cataclysm]

marvelous attempts via PI! You'd finally believed the holy grail of gradient removals....DBE since film days was used extensively. Because of the custom bg modeling character, once or even twice applied, with proper control boxing will remove most gradients, unless it is affected adversely by asymmetric local light cones.

the backgrounds are kind of clipped though as a result of DBE after-effects, and it may have sub-dued the faint nebulous regions (end points of rosette and m42). It is always good to use STF to auto-stretch the image first (remember to click 'A' symbol) to check for how much faint region data you potentially have in the stacked image. Colour calibration / background neutralization has to be performed too, to set the base. Although with time, you will find that the handy Pixelmath tool will be best for performing this.

Some equations that I have tested ok for neutralization would be:

$target - (Med( background ) - Med( background[2] ))
$target - (Med( background ) - Med( background[2] ))
$target

for the R, G, B channels respectively. This would be best on a single preview sample, which is good enough if data is taken from dark skies, and if the region is not over nebula rich.

Otherwise, another customized set of equations that can fulfill the second case would be:

$target - (Avg( Med( background1 ), Med( background2 ) ) - Avg( Med( background1[2] ), Med( background2[2] )))

$target - (Avg( Med( background1 ), Med( background2 ) ) - Avg( Med( background1[2] ), Med( background2[2] )))

$target

For both sets of equations, the Blue channel can be left as it is ($target).

Thanks Bro! Oh boy, really got to delve deeper into this.....