Pixinsight pixelmath problem...

[cloudbuster]

hi all....as ever a pixinsight problemo!

in pixelmath i was getting horrible dayglow greens and yellows at the autostrech....then realized that i had loaded the wrong flats into DSS!

i re-stacked my images, getting rid of any light or poor images...'.

Luminence: 30x10min

Red: 14x10min

Green: 18x10min

Blue: 18x10min

both the red and green stacks came out a little washed out, so i adjusted them on the luminence in DSS and saved them.

after putting them thru PI pixelmath, rather than the bright green or yellows i was getting, i am now gettin purple! which i think is a slight improvement!

i just wondered....would my 0.001 bias frames shot thru the Ha filter effect them? they are in with the stacks, and never gave it a thought until now.

please need your help as ever!

bob

[cgarry]

ps....sorry about the black txt at the beginning!

The black text at the beginning is all I can see since I use the IP.Board theme. Why change the colour to white?

Cheers,

Chris

[martin_h]

It doesnt matter what filter you shoot the bias frames with, they are taken in the dark!

[swag72]

If you use channel combination and get a vile colour (I often get yellow or green!!) then once it's DBE'd that sorts it out to a proper colour. Have you tried using DBE of the colour calibration tool after the image is combined into an RGB?

[cloudbuster]

sorry chris....

and sara....thanks. ive not tried the DBE tool yet. i just ran the stacks thru pixelmath first, in the hope that it might give a rough idea as to how the pic would come out. you mention 'channel combination'...should i of been using that instead of pixel math?

bob

[cloudbuster]

yes sara....i used channel combi and it is a vile yellow as u stated. i'll run it thru the DBE tool in a mo!

bob

[swag72]

I can't imagine why you'd use pixelmath for channel combination when it has its own process, but that's just me!

[Gib007]

I too use PixelMath to combine channel images into an RGB image. DynamicBackgroundExtraction will sort out your disgusting colour cast. Once that's done, ensure you use BackgroundNeutralization and ColorCalibration, followed by SCNR (to remove the overall green cast). Only then will you have a nice, colour-calibrated RGB image. This tutorial of mine goes through the entire process and much more, but it covers my procedure for colour combination:

http://lightvortexastronomy.blogspot.com/2013/02/tutorial-pixinsight-workflow-for-hdr.html

[cloudbuster]

thanx gib007....much appreciated.

will give it a whirl!

cheers

bob

[cloudbuster]

thanks all.

rough process attached, but would love to know how and at what stage in workflow to add the Ha i have....any thoughts peops?

cheers

bob

[billhinge]

Edit - oops just reread your question and noticed I missed the point , never mind the SHO AIP script is pretty good

Have you tried using the SHO AIP script ? You can preview different percentage changes quickly and dynamically (under script > multichannel synthesis)

eg

ScreenTransferFunction: Processing view: MIXSHO_AIP4

0.0300 s

PixelMath: Processing view: MIXSHO_AIP4

Writing swap files...

123.60 MiB/s

Executing PixelMath expression: channel #0: Ha*1+Ha*0.72+OIII*0+Ha*0+Ha*0: done

Executing PixelMath expression: channel #1: Ha*0+Ha*0.09+OIII*0.48+Ha*0+OIII*0: done

Executing PixelMath expression: channel #2: Ha*0+Ha*0+OIII*0.15+Ha*0+UV*0.76: done

Rescaling to [0.0000000000,1.0000000000]: done

3.060 s

BackgroundNeutralization: Processing view: MIXSHO_AIP4

Writing swap files...

123.60 MiB/s

Evaluating background: done

* Initial mean background values:

BR : 2.51153e-03

BG : 7.93203e-02

BB : 2.57728e-02

Applying background neutralization: done

1.690 s

MIXSHO_AIP4:

channel #0

c0 = 0.047998

m = 0.005735

c1 = 1.000000

channel #1

c0 = 0.047998

m = 0.005735

c1 = 1.000000

channel #2

c0 = 0.047998

m = 0.005735

c1 = 1.000000

ScreenTransferFunction: Processing view: MIXSHO_AIP4

0.0330 s

[Gib007]

thanks all.

rough process attached, but would love to know how and at what stage in workflow to add the Ha i have....any thoughts peops?

cheers

bob

Very nice result there! To add the Hydrogen-Alpha, you can simply add it to the Red channel. I would add both the Red and the Hydrogen-Alpha images together first and then with this new blended Red, I'd do the RGB combination you did. All within PixelMath of course. I say just add it to the Red because I'm pretty sure your Hydrogen-Alpha data for a galaxy is not extremely plentiful so it shouldn't saturate pixels. If it did, you could scale the addition process by adding a certain percentage of each (e.g. (0.9*Red)+(0.75*Ha), or whatever you feel fits).

[cloudbuster]

Very nice result there! To add the Hydrogen-Alpha, you can simply add it to the Red channel. I would add both the Red and the Hydrogen-Alpha images together first and then with this new blended Red, I'd do the RGB combination you did. All within PixelMath of course. I say just add it to the Red because I'm pretty sure your Hydrogen-Alpha data for a galaxy is not extremely plentiful so it shouldn't saturate pixels. If it did, you could scale the addition process by adding a certain percentage of each (e.g. (0.9*Red)+(0.75*Ha), or whatever you feel fits).

thanks gib007....i have an appointment this morning, but will give this a bash later!

cheers

bob