|
Hi, Glen,
This is a good image, and I'm surprised it hasn't received lots of positive support.
The very best way to handle the overwhelming excess of H-alpha over the other channels is to take a much longer set of subs in the weaker channels. With an object like this, that can border on the insanely impractical.
Another is to use 2x2 on-camera (not after the event) binning, preferalby with dithering. However, if you are undersampled to start with, that can lead to loss of resolution in the OIII and SII channels.
But even if none of that is possible, you can still enhance the OIII and SII channels in processing. Most folk do that routinely.
Your black points for OIII and SII look pretty spot on. So all you have to do is to increase the brightness for OIII and SII. For concreteness, a linear increase in OIII of 150% would mean that 0, 1000, and 2000 become 0, 1500, and 3000 counts.
Another trick that can help a bit is to carefully increase the contrast in the H-alpha channel, so long as you don't overdo it and posterize, this will reduce the excessive H-alpha in the weaker parts of the image.
Here I've had a go at doing that: increasing the H-alpha contrast a bit, and then increasing the OIII and SII linearly until the mean counts for OIII and SII exactly match the H-alpha mean count for the whole image.
The resulting image no longer pretends to show which of OIII, SII, and H-alpha is the strongest. But that often isn't very interesting anyway. Think of a map showing the distribution of dirt, coal, and industrial diamonds in Australia. There is overwhelmingly more dirt, but we want to know where the diamonds are. The approach I'm advocating shows where the SII is concentrated, and how it looks, not how much there is.
It's never a good idea to do this to someone's finished image. It would be far better if you did it yourself at an early stage in processing. That will avoid the OIII burning out in the very brightest parts.
I've also zapped the OIII and SII stars, so there are only H-alpha stars, mapped here to white. I've only done that very very roughly, just to give the general idea. Two reasons for doing this. The first was that in this particular image, your channels weren't quite aligned properly in the corners, suggesting your registration software isn't quite right. The proper way to fix that is with better registration and stacking software that can handle rotation, change of scale, and other distortions between frames. The other reason is to reduce the inevitable magenta haloes that the technique I've described lead to.
Hope I'm being useful here, rather than a pain in the potato.
Best,
Mike
|