StarTools

Camera lenses are designed to be at prime focus for the green light the way I understand it, which results in red and blue being a little out of focus. Especially for modded cameras, you end up with red halos due to the out-of-focus red channel. So if you're not using a narrowband filter and end up with stars that have red halos, what is the best way to deal with it in Startools? Most of the advice I've seen involves processing just the R channel, but I couldn't figure out how to do that with my DSLR data. Any ideas? I feel like there is an easy answer and I'm too dumb to figure it out.

Thanks for the help.

Hi,

Let me say first off that there is no such thing as dumb question! (except for the one never asked ). Personally I'm pretty impressed with your break-neck progress - love your Orion wide field on Cloudy Nights; perfectly framed and bringing out Barnard's loop like that, is just the icing on the cake!

Yep, chromatic aberration can be a real pain, but putting up with it (and working around it) lets you use some, otherwise really fantastic, bang-for-your buck gear.
There is a number of things you can do in StarTools. Indeed, knowing that only certain channels exhibit the out-of-focus behavior can be used to your advantage.
The Green channel is typically the most reliable and there are a number of ways to take advantage of that knowledge.
For example, you could launch the LRGB module, click 'Green', loading just the green channel of the data. StarTools will automatically propagate the green data to the red and blue channel, so that, without any further loading, you'll end up with a grayscale image that represents the green channel's luminance data; i.e. a synthetic luminance frame.
If all is well, the halos in the synthetic luminance frame should be small or non-existent.
You can then process this image as normal. Then, when you're done, you can reintroduce color using the LRGB module, by loading the image you just processed as a luminance frame, while you specify the original image for the red, green and blue data. This way the luminance data will have come from the green channel only, but you still have a color image.

Another, quicker way of doing something (effectively) similar, is to load your image, process your image as normal (don't modify color yet) and at the end launch the Develop module (stretch as-is - don't redo global stretch!). There's a nifty feature in the Develop module that will allow you to specify the luminance contribution of the individual channels. Simply set green at 100% and read and blue to 0% (or anything in between that looks good). Again, color will be retained, but the luminance contribution will change.

Lastly there is the fringe killer functionality that will help reduce aberrant color. Using it in conjunction with the magic module (star size manipulation through, for example the Tighten and/or Shrink algorithms), should allow you to effectively tame the offending stars.

Hope this helps!

Cheers,

Great reply, thank you! I saw the tutorial about using a "synthetic" luminance frame, but I did not think about how it could be implemented with a DSLR.

I tried the method here:



I believe it worked somewhat. In any case I decided to process a more natural look. Your method was great because it allowed me to keep the luminance data more sharp and then be heavy-handed on de-noising the color. My previous effort looks a bit garish by comparison... perhaps a little overdone in some places. Live and learn I guess. The red channel was insanely noisy due to my unmodded camera. On my to do list is get the camera modded and stop down to f4.5 next time and avoid the CA!

Thanks again for the help.

Nice! Great colour balance too.

If you'd like to bring out the loop a bit more, have a play with the Filter module (Nudge or Conservative Nudge).