RGB star color preservation with NB

[johnnyK]

Processing RGB and NB subs together using the script utility NBRGB the stars seem to lose there original colors obtained from the RGB subs. Is there a best practice or tutorial on how to create NB and RGB images while bringing out the star colors.
Converting from PS to PI so I am still new with all the tools available. I was able to generate decent star colors in PS by whiting out ans shrinking the stars in NB subs and overlaying the RGB over the NB with a Lighten Blend mode.
Thanks

[akulapanam]

I was noticing that myself today.  It seems like the HaRVB-AIP script does a better job here

[hvb356]

You may check out my script "EmissionLineIntegration.js", which can be found on Herbert's page:  https://www.skypixels.at/pixinsight_scripts.html
EmissionLineIntegration tries to remove unwanted frequencies from the NB and adds the emission line intensities to the channels. Play with contingents and leave Mue correction unchanged.
This Rosette https://www.flickr.com/photos/hvb_astro/39547057252/sizes/l was made from RGB + Ha + OIII

[johnnyK]

Thanks I will check it out.
JK

[akulapanam]

I played with that script a bit and looked at the PowerPoint here.  http://www.skypixels.at/downloads/PixInsight/Tagung_VIII/EmissionLineIntegration.pdf .  It seems to work fairly well but produces a noiser image and the color tends to be a bit hot pink compared to the RGB frames but maybe better than Salmon?

[monkeybird747]

There is this tutorial that uses a good star mask and MorphologicalTransformation to erode/remove the unwanted stars from the NB image.

http://www.lightvortexastronomy.com/tutorial-reducing-star-sizes.html#Section3

You could try that before combining using one of the above mentioned methods. I'm going to try this on my next processing attempt with some OSC + Ha data. My first attempt shows some of what you are talking about, i.e. some star color loss that is hard to get back. Here is a screen shot of, from left to right, RGB fully processed, Ha integration, RGB+Ha fully processed.

[johnnyK]

I have seen that tutorial and tried it with some results. There seems to be a separation between true color and personal interpretation in the various tutorials.
Since the Emission Line Integration process is done in the linear steps, one may use a star elimination in the NB images prior to applying ELI. Then perform the Photometric Color Calibration(G5V) prior to going non-linear. This may have the benefit of enhanced NB integrated into the RGB image with ELI and true star color with the Color calibration. Then perform other processes to enhance the total image. I would have to try this on old data since we are in the monsoon season.