G'day everyone, finally got some clear sky, and it's been a long wait. Anyway, this is a two panel mosaic of NGC6188 and I managed to catch most of NGC6164 in the corner. I might take a third panel oriented vertically up the right hand side to try and centre it all up a bit and catch the rest of 6164. We'll have to see how the weather goes. I'm not totally happy with this, but it's the best I can do with current knowledge. As usual I struggled with controlling the stars. Any pointers or tips to improve my imaging or processing would be greatly appreciated.
Anyway image details:
40x300sec @ ISO800 for each panel, making a total of 80x300sec.
I've probably got no advice for you, but I like the image.
By controlling the stars, you mean not overexposing them? If so, maybe you could take some shorter exposures for the stars and process them into the longer exposures?
It looks clipped a tiny bit on the black end. I think your stars are pretty good. Perhaps I would try combining linear starless lights for the RGB image to reduce colour halos around stars. There is also deconvolution, but I am yet to learn properly how it works.
You certainly have good data, so several revisions over time will allow for making the most out of the data
Nicely stitched, Rex. Bloated stars are always a challenge. You can control them partly during stretching, e.g. with MaskedStretch in PixInsight, and you can shrink them in a variety of ways (deconvolution as suggested by Slawomir, or PI's MorphologicalTransformation.) Chris's suggestion of HDR techniques is worth trying too. There's no single magic bullet but several techniques worth trying... possibly even combinations of them.
Getting adventurous with the mosaic Rex! Kudos for giving it a go. Some good feedback has already been provided. I like the composition. Now you've cut your teeth, perhaps add a few more mosaic panels.
I've probably got no advice for you, but I like the image.
By controlling the stars, you mean not overexposing them? If so, maybe you could take some shorter exposures for the stars and process them into the longer exposures?
Thanks Chris. Yeah kind of mean overexposure, although I think the problem comes more when I stretch the data, they tend to get really bright and overwhelm the nebula. I have tried star masks etc while stretching but havent tried the HDR approach yet. Thanks for the tip', I'll give it a try and play around to see what I can do.
Quote:
Originally Posted by Slawomir
Nice image and well stitched together
It looks clipped a tiny bit on the black end. I think your stars are pretty good. Perhaps I would try combining linear starless lights for the RGB image to reduce colour halos around stars. There is also deconvolution, but I am yet to learn properly how it works.
You certainly have good data, so several revisions over time will allow for making the most out of the data
Thanks Slawomir. Yeah it is a fraction clipped on the second version (on astrobin), The first version seemed a little flat so I clipped it a little more. I was trying to get a little more contrast in and looks like I went a tad too far. Thanks for the tips re the stars, I hear you about deconvolution. I have actually used it on this image but I am still trying to work out the settings as I can never get the stars right without overdoing the nebula. I usually can only get about 10 iterations even with varying settings without getting artifacts. Anyway I will work it out eventually. Thanks again for the feedback.
Quote:
Originally Posted by DJT
Looking good Rex, lots of subtle tones in there.
Thanks David. I feel I am improving, although it's hard to tell for sure as the seeing is pretty bad this year up this way.
Quote:
Originally Posted by multiweb
Wow! That's tops Rex! Well done.
Thanks Marc, I'm just glad to get some light on the chip.
Quote:
Originally Posted by RickS
Nicely stitched, Rex. Bloated stars are always a challenge. You can control them partly during stretching, e.g. with MaskedStretch in PixInsight, and you can shrink them in a variety of ways (deconvolution as suggested by Slawomir, or PI's MorphologicalTransformation.) Chris's suggestion of HDR techniques is worth trying too. There's no single magic bullet but several techniques worth trying... possibly even combinations of them.
Cheers,
Rick.
Thanks Rick for the comment and pointers. I actually masked the stars manually, did a half stretch then removed the mask and finished it off when I did this one. I also used the Morph transform on it as well and some decon. I think it will just come down to practice makes perfect. There is so many different settings with all the techniques that it is hard to find the time to play around and see what each setting actually does. But steady steady I am getting an understanding of some things and I do feel I am improving, so that's always a positive. Thanks again for the feedback.
Quote:
Originally Posted by jase
Getting adventurous with the mosaic Rex! Kudos for giving it a go. Some good feedback has already been provided. I like the composition. Now you've cut your teeth, perhaps add a few more mosaic panels.
Thanks Jase. I am actually doing another mosaic project at the moment, but I will probably come back to this one next new moon and get at least one more panel up the right hand side.
A few pointers in my opinion. Possibly a tad too red. I know it has a lot of Ha but as a general rule of thumb of my own I like to try to get a range of colours in an image and not too much of any one colour (of course objects have a colour but when you add Ha or have a Ha boosted DSLR which tend to red bias it becomes a bit harder).
I have no attention on the stars when I view the image apart from the occasional red bias. The dust ridge near the bright area could be selectively sharpened as it really is the subject of the image and thus it could be made to impact more. Selective sharpening means a sharpened layer and a mask and then reveal only those areas of interest through the mask. I usually use Photoshop but I am sure it can be done in PI.
The mosaic seaming is perfect, I cant tell where one panel ends and the other starts and that's very hard to achieve in mosaics.
This is actually a pretty good image Rex and not that far from the winning deep sky shot at last nights DM awards really some colour dynamics are missing as it is a bit red/white but for me the good things are that there are no obvious signs of deconvolution, wavelets or minimum filter ie no spotty backgrounds, little bright worms or detail all turned to perfect little dots. No flattened look from heavy handed shadow highlights either, so they're all good things!... a gold star from me
A few pointers in my opinion. Possibly a tad too red. I know it has a lot of Ha but as a general rule of thumb of my own I like to try to get a range of colours in an image and not too much of any one colour (of course objects have a colour but when you add Ha or have a Ha boosted DSLR which tend to red bias it becomes a bit harder).
I have no attention on the stars when I view the image apart from the occasional red bias. The dust ridge near the bright area could be selectively sharpened as it really is the subject of the image and thus it could be made to impact more. Selective sharpening means a sharpened layer and a mask and then reveal only those areas of interest through the mask. I usually use Photoshop but I am sure it can be done in PI.
The mosaic seaming is perfect, I cant tell where one panel ends and the other starts and that's very hard to achieve in mosaics.
Greg.
Thanks Greg. Your comments are very helpful and I understand what you are saying about the red and also the sharpenning, I will tweek this a little in a couple of days so I can look at it subjectively. I have to find some time to play a whole lot more with decon ss well so I have at least a bit of an idea of what to do with the settings instead of guessing with trial and error. As for the seeming, PI has some pretty good tools that help with that once you work out what the settings do. Thanks again for the feedback.
Quote:
Originally Posted by strongmanmike
This is actually a pretty good image Rex and not that far from the winning deep sky shot at last nights DM awards really some colour dynamics are missing as it is a bit red/white but for me the good things are that there are no obvious signs of deconvolution, wavelets or minimum filter ie no spotty backgrounds, little bright worms or detail all turned to perfect little dots. No flattened look from heavy handed shadow highlights either, so they're all good things!... a gold star from me
Mike
Wow thanks Mike for the huge compliment. I think I have a long way to go before I feel like my images are anywhere near as good as those submitted to the DM's, but thanks for the vote of confidence and the gold star regarding my processing. As I have said I do feel like I am improving so just gotta keep at it.
that is a mighty fine dragon and egg there Rex. I would be very happy with that. A single frame is hard enough and you've gone stitching, worked very well.
I don't have anything to add by way of constructive feedback, other than to say this is a fantastic shot. Star shapes look great across the frame, good colour and saturation, seemlessly blended panels and nicely processed. Top work
I really like your image. Lots of detail in there. It's definitely a target I'd like to spend a bit more time on myself.
Quote:
Originally Posted by Rex
, I hear you about deconvolution. I have actually used it on this image but I am still trying to work out the settings as I can never get the stars right without overdoing the nebula. I usually can only get about 10 iterations even with varying settings without getting artifacts. Anyway I will work it out eventually. Thanks again for the feedback.
I had a bit of a breakthrough with deconvolution yesterday.
In PI you are supposedly meant to set a linear RGB space. Basically this means using the RGBworkingspace tool with R, G, B and gamma all set to 1. Then make a starmask with the appropriate tool and a custom PSF with dynamic PSF (there's a tutorial out there on this). Use these in the deconv tool. Main thing I play with is the "global dark" deringing setting. I've had a lot of success with this process so far, since discovering the RGB working space thing.
That's pretty good Rex. Probably a little too red but that's not serious.
Cheers
Steve
Thanks Steve, the dreaded RED, lol. I always seem to have trouble with that. Damn modded camera ahahahah. I'll have another look in a few days.
Quote:
Originally Posted by Paul Haese
That's pretty good Rex. Maybe your blue stars a little cyan, but the detail looks good and the reflection nebula areas look natural.
Thanks very much Paul. I'll definitely have another look at the colour balance of this one.
Quote:
Originally Posted by rustigsmed
that is a mighty fine dragon and egg there Rex. I would be very happy with that. A single frame is hard enough and you've gone stitching, worked very well.
Cheers
Russ
Thanks Russ, I actually really enjoyed doing this one.
Quote:
Originally Posted by Octane
Rex,
This is outstanding.
As I often comment on NGC 6188 images, I have struggled for the last ten years to make a decent image of it. Perhaps next year?
Well done on a beautiful and aesthetically-pleasing image. You should print, frame and hang this one up.
H
Thanks for the huge compliment H. I may actually frame it yet once I am finished with it. Should only take another four or five tweaks lol.
Quote:
Originally Posted by rmuhlack
I don't have anything to add by way of constructive feedback, other than to say this is a fantastic shot. Star shapes look great across the frame, good colour and saturation, seemlessly blended panels and nicely processed. Top work
Thanks Richard, I appreciate it.
Quote:
Originally Posted by LewisM
Really good Rex!
Thanks Lewis.
Quote:
Originally Posted by bugeater
I really like your image. Lots of detail in there. It's definitely a target I'd like to spend a bit more time on myself.
I had a bit of a breakthrough with deconvolution yesterday.
In PI you are supposedly meant to set a linear RGB space. Basically this means using the RGBworkingspace tool with R, G, B and gamma all set to 1. Then make a starmask with the appropriate tool and a custom PSF with dynamic PSF (there's a tutorial out there on this). Use these in the deconv tool. Main thing I play with is the "global dark" deringing setting. I've had a lot of success with this process so far, since discovering the RGB working space thing.
Thanks Marty, for the comment and the tip about decon. I have been doing all the things you mentioned except the RGB workspace set to 1. I usually only do that to export the luminance channel and then reset it back to the native settings. I will definitely give that a try. Thanks again.
Thanks Marty, for the comment and the tip about decon. I have been doing all the things you mentioned except the RGB workspace set to 1.
It was specifically the RGBworkspace thing that was my breakthrough yesterday - the rest of the workflow is my old process. The old process worked great with some images and not at all on others. The RGB thing changed an image from the second category to the first yesterday, but more testing is needed
