Hi Everyone,
I hope I'm not taking this discussion too far afield, but I thought I'd ask some questions about how these HaLRGB `blending' techniques fit into other parts of the overall `PI workflow'... Feel free to let me know if I should make this a new thread in a different part of the forum.
I recently acquired data on M16 (the Eagle nebula) from a very dark site. I got 2 hours (=24x5min) each of L, R, G, and B, all unbinned. I also got 5 hours (=20x15min) of Ha, also unbinned. All of these data have been calibrated, registered, and integrated in PI. I now have a master frame for each of: Ha, L, R, G, B.
Clearly, at this point, I can make an RGB image from my R, G, and B. I could also make an LRGB image. And I could dig into various posts listed in this thread, to see how to make some sort of HaLRGB image.
What I'm wondering is... where should things like sharpening and smoothing fit into all of this? It seems to me that the overall philosophy of PI is, broadly speaking, rather like the `Astro Zone System': 1) Use some sort of sharpening on the high-SNR areas (e.g. Deconv or an ATWT-based method), and 2) Use some sort of smoothing on the low-SNR areas (either ACDNR or an ATWT-based method). Juan has given us some nice instructions on how to do these things, in various posts. There are also processing examples on other parts of the PI site. Some methods should be applied to linear data, some to non-linear, but that's all been discussed elsewhere.
Mostly, I'm wondering when I should do these sharpening and smoothing operations. Should I do some sort of blending of the Ha, L, and RGB data while still linear, and THEN do the sharpening and smoothing on the resulting image? Or should I do various things to the Ha, the L, and the RGB, (probably not the same things to each image) and then combine?
It looks like there are various ideas for how to combine Ha, L, and RGB, but I'm just confused as to how these steps relate to the OTHER parts of the workflow. If anyone has any thoughts on this, I'd be very curious to hear.
Thanks,
- Marek