PixInsight Forum

PixInsight => General => Topic started by: chboing on 2018 July 22 14:08:07

Title: Bad filter I think ?
Post by: chboing on 2018 July 22 14:08:07
Ola todos

TLDR : you should be able to understand by reading bold text only ^-^

First, let me say I bought Pixinsight 2 days ago, after the trial period. And i'm quite happy with this ! :)
I started this hobby last december, I'm on a low budget, but I have to say, this is a damn good software ! great work, could not do without it now. Thanks for this.
Now i've had some problems with some of my stacked images (after all calibration) the results had a greenish vignetting, that was sometimes here, always almost the same, and sometimes not.
So i've been puzzled for quite some time before realising yesterday that it was when I used a cheap (actually not THAT cheap) CLS filter (i live in town...) that i had those green halo.

shot with : Canon 1000D, Newton Astrograph 150/600 -> f/4
OPTOLONG Telescopes CLS Clip filter
often 240 seconds subs
See here :

What do you think, could this be it ? a bad filter not filtering uniformely ?
I'm definitely not using this filter again alright BUT, i was thinking since i have the best software for this in my hand, as wel as darks & flats & bias. That maybe i could sort this.

So first (during the 45 periods trial) as a noob, i tried to solve this with the great background extraction tools... but well, on big nebulae like rosetta neb, well it doesn't work well cause the nebula is everywhere, and i could not do without alterate it. So i had some results, but I'm sure it's really below what i could do.

Correct me if i'm wrong, but the flat are used only for "lightness" if i understood, that's why the "color" of flats does not matter, we finally divide the Light with a normalized flat. (and all the correct calibration before all right but i think we can ignore this here.)
So my flats are correcting perfectly the vignetting effect, good...

But I was thinking, if this green thing comes from the filter, and i shot my flats with the filter, then surely, the "gradient in the filter" that would cause this is somewhere in the flat right ?
So how could I extract this, and then ... well subtract/correct this from my light ?
I was thinking right now (but did not tried yet) :
-correct the "rgb-flat" with the normalized flat ? as if it was a light frame ?
-I would obtain a "flat" containing no more vignetting data, but only color data ? the green halo ?
-substract it from stacked image maybe
-since the color will probably won't be good anymore, use photometricMagicCalibration :)

That's where i need some help, i don't know if that would work, and i'm thinking pixel math will be of some help here, but even simply applying a flat to a L seems not as simple as i thought.
L / F -> not good.
So maybe some basic sample pixel math expressions would help me better understand, but i could not find some...
Or just some rightly ordered process to apply different files would be of great help !! i'll find the time to fine tune this but i need help to have a better picture of what to do.

here are some files in case someone has some free time ;) (https://we.tl/dj3lWegm9Y)
contains master bias,dark,flat, light, and a light frame :

Thank you !

Title: Re: Bad filter I think ?
Post by: chboing on 2018 July 22 14:42:50
I think my halo is there ... I have no idea how to do now :p

I definitely think i'm on the right track. But i don't know how to correct the Light with my "result file"
I calibrated the flat with a lightness extraction of the same flat as a master flat ... and obtained this :

RGB linked STF
RGB NOT linked STF

Title: Re: Bad filter I think ?
Post by: chboing on 2018 July 22 15:50:45
Well, in case anyone has this problem, i have a solution that suits me, even if it's not the right one.
Since last post, I calibrated my flat with "the lightness channel of my flat" as a master flat (image calibration). Let's call this "the thing"
Then i created a NEW homogenous image with pixel math that has the value of the green in the center + 0.0004 (R 0.9942 + 0.0004, G 0.9921 + 0.0004, B 0.9932 + 0.0004)  -> "green"
then pixel math : "green" - "the thing" -> "Image22"

and finally :
myStackedImageWithGreenHalos +0.48 * "Image22"  (0.48 after some tries ... )
BAM ! no more halo !
a light automatic brackground extractor after this, a rapid curve and :


I'm so happy right now, knowing i can do this for my other pictures
+ this lagoon nebula is really something ! i was so disappointed i never really took the time to enjoy the view :D