A different Corona...

[lazjen (Chris)]

Corona Australis - at least the parts that contain NGC6729, NGC6726, IC4812 and the nearby globular cluster NGC6823.

This was a lot more painful to process than I thought it would be. And I think I will be going back for at least one more reprocess in the future, as I'm not happy with the star colours.

Astrobin with details, much larger pic: https://astrob.in/2oaru1/0/

[strongmanmike (Michael)]

Gee the dusty parts really look like dust Chris, nice work there. As you say the star colours are perhaps just a tad mute I guess but for me it looks like something in your processing has rendered much of the image kind'a two toned ie cyan and brown, the glob is very cyen/blue too some noise/colour noise reduction can do this, not sure..?

Still, the dusty look is prominent .

Mike

[lazjen (Chris)]

Yes, the glob is the red flag that indicates me having messed up the stars. I just didn't control the stars properly when pushing the colour/saturation. There's very limited noise reduction done in this image - BackgroundNeutralization, maybe whatever was in the script DarkStructureEnhance. No other NR that I can remember.

So anyway, consider this version more the nebula parts.

I also did a small crop on this to get rid of the bright star (Epsilon Corona Australis) as the ASI1600MM microlens reflections were terrible. I've now got a tutorial that I'm going to try to see if I can remove those reflections and keep the star (and surrounds in the pic). that plus better masks, etc should get me further. It'll have to wait until the next lot of cloudy/rainy weather though for the next attempt.

[gregbradley]

Lovely image Chris.


Star colours is the area I find these CMOS cameras weak in general and an area that really needs to be watched compared to CCD.

I am not sure why but many of these CMOS based camera images share this characteristic and the more colourful star images on Astrobin are invariably CCD cameras.

One approach that seems to work and is common with DSLR imaging is to use shorter exposures for the stars.

Its like imaging with too high an ISO with a DSLR. The stars end up all white with no colour.

It probably has to do with the fact that you can adjust gain on a CMOS which is most likely a lot higher than that on a CCD which then blows out the stars quicker than normal. The pixel sizes are also usually smaller to meaning smaller wells that fill more quickly (also due to the high QE of most of these sensors - not so much the 1600 though).

Greg.

[codemonkey (Lee)]

That's really nice, Chris, well done! The star colours do look a bit off, perhaps PCC in PixInsight might help with that?

[lazjen (Chris)]

Originally Posted by gregbradley

Lovely image Chris.


Star colours is the area I find these CMOS cameras weak in general and an area that really needs to be watched compared to CCD.

I am not sure why but many of these CMOS based camera images share this characteristic and the more colourful star images on Astrobin are invariably CCD cameras.

One approach that seems to work and is common with DSLR imaging is to use shorter exposures for the stars.

Its like imaging with too high an ISO with a DSLR. The stars end up all white with no colour.

It probably has to do with the fact that you can adjust gain on a CMOS which is most likely a lot higher than that on a CCD which then blows out the stars quicker than normal. The pixel sizes are also usually smaller to meaning smaller wells that fill more quickly (also due to the high QE of most of these sensors - not so much the 1600 though).

Yes, all the above is true, but I have used shorter exposures. I'm almost certain the data is (mostly) ok, and that it was my processing at fault. I've still got a long way to go to learn (and practise) some techniques.

Originally Posted by codemonkey

That's really nice, Chris, well done! The star colours do look a bit off, perhaps PCC in PixInsight might help with that?

PCC failed to work for this image too, but I think I might have another idea to retry it next time around.

One of the hassles with all this processing is how long it takes. Even with my decent machine here, it takes hours to get to the point for the 4 channels of drizzle integration (this started with over 200 images per channel). Then I do all the processing on these larger drizzle images, and only resize when publishing, so each operation takes just that little bit longer. That's one of the reasons why I'm taking a break before having another go at this image for a reprocess. At least when I do, I can start at the drizzle images and go from there.

[codemonkey (Lee)]

Originally Posted by lazjen

PCC failed to work for this image too, but I think I might have another idea to retry it next time around.

One of the hassles with all this processing is how long it takes. Even with my decent machine here, it takes hours to get to the point for the 4 channels of drizzle integration (this started with over 200 images per channel). Then I do all the processing on these larger drizzle images, and only resize when publishing, so each operation takes just that little bit longer. That's one of the reasons why I'm taking a break before having another go at this image for a reprocess. At least when I do, I can start at the drizzle images and go from there.

Ah, right, I forgot you were having issues with PCC. Yeah, that makes things hard.

It's been a long time since I've used drizzle, but I recall with PI you can use drizzle without increasing the image size. I would hope that would help with the negative effects of being undersampled (e.g. square stars) while keeping the integrated data and thus speeding up post processing

[lazjen (Chris)]

Ok, might have made some progress with PCC - I was using automatic magnitude limits (which has worked fine for me in the past). When I turned that off and increased the mag limit value, things started working better. I also had to adjust the noise reduction setting as well (which makes sense).

It gets past the plate solving, but is failing with the photometry. Doing more fiddling with the params there to see if I can get that to work.

I'll look into the drizzle stuff more - maybe I can adjust the settings better for faster processing.

[ChrisV (Chris)]

Nice image Chris. The dust looks, well dusty. I can't even get PCC through the plate solve ...

[lazjen (Chris)]

Thanks Chris, but I'm still having problems with it too.

I managed to get PCC to work with the above image, but the results weren't great. I suspect because I had ruined it a bit too much. So I went back to the start and worked out how to get the star colours much better, but PCC isn't liking the new image.

And I'm having trouble getting a good star mask that includes the glob, but not the nebula, which means when I try to improve the colour of the nebula I wreck the glob.

And the microlens star that I cropped out for this image is not helping either.

I wasn't going to do the processing today, but I got these ideas/tutorials/etc and wanted to try them out.

I think the next attempt will be to crop out the microlens star from all 4 channels first and leave attempts to fix it to some distant future. I have to think about this starmask some more too - or a workaround.