So I'm reposting without the reference to the competing product which it was not my intention to promote.
You have a couple of choices to integrate HA into OSC or LRGB.
- Using Ha as the red channel - This works but doesn't account for the fact that Ha areas are also emitting blue light as a result of H-beta (and probably H-Gamma?) emission from the same atoms at a lower rate (34% although I think it varies from region to region). It also doesn't impact the intensity of the color.
-Using Ha as the luminance channel - This produces a high contrast image without strong star size artifacts but throws out the intensity of other objects that appear in OSC RGB or broadband Luminance. However it will correctly capture the Ha/Hb ratio in the Ha regions. It also adds a bunch of noise.
-Mixing Ha into the Red/Blue channels - This appears to be the basis for the SHO-AIP & HaRVB-AIP and NBRGBCombination scripts in PixInsight. We can't know for sure because PixInsight has zero documentation on the subject. If someone would post a workflow it would be appreciated or documentation it would be appreciated. The challenge here is that it merely adjusts the color but NOT the intensity.
-Synthetic Luminance - This works by combining the ha and luminance to capture the intensity of the Ha region without loosing the intensity of other areas and while keeping the correct color. Ha probably is slightly over allocated here but it seems better than just using ha as the luminance channel. It also is less noisy than merely using Ha for luminance. I think you need PixelMath to create this in PixInsight but maybe there is a script that does this?
One of the bigger challenges is deciding whether to combine the images after they have been stretched or before they are stretched. I haven't quite figured this out yet. Clearly for synthetic luminance and Ha as luminance pre-stretch combination works well at least for those frames. However, it does appear that there is a benefit to combining your HA and RGB for the other methods AFTER you have stretched and background calibrated each image.
For OSC users it should be noted you have two additional choices that impact resolution.
-You can split your CFA (color filter array) to get individual RGB channels on the RGB capture and then combine that at the same resolution with the split CFA R channel from your HA filter images. You will loose considerable resolution here but for faint objects won't hurt and the methods above are easy to use. Stacking also becomes a longer process. Just make sure you calibrate your images BEFORE extracting each channel.
-You can drizzle stack or resample align the split CFA R channel to be the same size as the interpolated RGB image. For some of the methods above you may need to split your interpolated RGB image into separate channels.
_____________________________________________> Below are notes from my first attempt.
This is one of the only objects I can think of where you can get a planetary nebula, emission nebula, and globular cluster in your field of view. The bad news is that Sh-2-72 is sooo dim that 2.5 hours of exposure without a HA filter revealed absolutely nothing. Not to be differed I went and purchased a Future Optics 12nm Ha filter and went to work. I currently use a color Atik 490 so figuring out how to process this image was a bit of a challenge. One thing I decided right off the bat is that I would not try and keep full resolution and instead opt to split the channels. I may have tried full resolution but the tool I used doesn't support drizzle for stacking. I would like to use PixInsight here but the SplitCFA tool gives me a odd pixel count error.
Method 1) Ha as Luminance. I stacked each channel separately, then processed, saved, and recombined using my Ha master as luminance.
http://astrob.in/193516/A/Method 2) Synth Luminance. Luminance added to the RGB via LRGB combo.
http://astrob.in/193516/B/Method 3) Synth Luminance with different color addition approach
http://astrob.in/193516/C/I also tried the NBRGBCombination and HaRVG-AIP scripts in PixInsight and frankly had no luck. If anyone has used either method I would love to see your work flow. The HaRVG-AIP method in particular did something that produced a FIT file that was only open able in PixInsight. Also neither one of these methods correctly accounts for the Hb inside of the Ha channel whereas the methods I used do.
