Synthetic Green Channel

[Redshift]

Hi,

Is it possible to make a Synthetic Green Channel from Ha and OIII date in PI, so I can make a BiColour Image?

I've been playing around with PixelMaths for a few hours but I'm not getting anywhere.

Thanks

Dave

[Andres.Pozo]

I don't know where do you want to get  but with Pixel Math is easy to get a green channel mixing other channels. Please watch the attached screen capture: using PixelMath I got an RGB image with a green channel composed of 66% Ha and 33% O3. After this, I applied a BackgroundNeutralization, a ColorCalibration and a STF.

[Nocturnal]

I assume you want to create a TRIcolor image then, right? I don't see the appeal of sticking R and B values into G but I know it's commonly done. Pixelmath is the way to do it. Which formulas have you tried? More importantly, what are you trying to do? You can't -create- Green afterall. The best you can do is mathematically combine R and B which means the signal in G is essentially the same as in R and B. Anyway, what formula do you want to use? Which PM formulas have you tried?

[Redshift]

Andres and Sander,
Thank you for your replies. I'm trying to emulate Steve Cannistra's technique here http://www.starrywonders.com/bicolortechniquenew.html he uses Photoshop I was hoping to do it in PixInsight. Here is one of his images using the process http://www.starrywonders.com/rosettebicolor.html.

I've only tried the PixMaths process that Andres suggests using different weights for the Ha and OIII in the Synthetic Green channel.

Regards

Dave

[Nocturnal]

Hmm, that technique sounds like alchemy and I don't subscribe to that Hopefully someone has translated all those PS operations into what's actually going on underwater so it can be written as PM or other processes. It'll take a careful read and understanding of what the PS layers really do.

[Redshift]

Sander,

I like it  and he's turned out a good looking photograph.
Yes, what do the Screen and Multiply Layers etc do.

Dave

[Nigel Ball]

Dave

The various PS blending modes can be implemented fairly easily in PixelMath once you know the formulae

Rogelio has a guide to PS Blending Modes on his website which I found very useful and was able to implement those I needed in PixelMath

Have a look at http://blog.deepskycolors.com/archivo/2010/04/21/formulas-for-Photoshop-blending-modes.html

HTH

Nigel

PS If you have problems with PixelMath just call .......  

[Redshift]

Nigel,

Thank you, I've just seen your message. I don't know how I missed it.

Regards

Dave

[dlp]

I'm trying to use Pixel Math as per the formulae on the deep sky colors website listed above and am having a problem parsing a formula for 'soft light' :

if (blend > 1/2) R = 1 - (1-target) * (1-(blend-1/2))
if (blend <= 1/2) R = target * (blend+1/2)

Gives 'unknown function identifier'. I have my images named as 'blend' and 'target', and symbol 'R' just entered into the Symbols box.

Can anyone enlighten me please?

David

[oldwexi]

Hi David!
When i want to use the so called "Soft Light" function from PS in PI
i paste into PixelMath this formula:

iif(Blend > 0.5, 1-(1-Target)*(1-(Blend-0.5)), Target *(Blend+0.5))

The names of the images (layers) therefore are:
Blend and Target ....

Hope this helps
Gerald

[dlp]

Thanks Gerald - worked a treat!