Here is one of my recent RGB attempts. It suffers from my usual problems with RGB:
- Stars having halos of different colours on different sides
- Colour pixels needing to be cloned out
- Lack of "colour contrast" in the galaxy - I would prefer the centre to be more golden compared the the blue arms.
Thoughts on how to improve?
29x600s clear
19x600s red
18x600s green
11x600s blue
ST8-XME, 12" SCT @ 0.84"/pixel.
More blue data would be ideal but it's been cut short a few nights now.
I'm thinking the NABG and high noise level of the ST8-XME isn't helping my colour attempts. I'm also wondering if a better optical tube would would the clarity or star bloat.
The focus is out a touch Roger as evidenced by the visibility of the secondary shadow in a number of stars.... that also might be symptomatic of a highly curved field.
Nice Roger great effort for the focal length and exposure times {tracking LX200 ?}. Do you check the focus before each exposure or is it automated? Good weather lately in Perth not to hot to get out the gear out
That's not bad Roger. Firstly that is a tough target. Those 2 galaxies are very small and you are imaging at over 2.5 metres. SCT's do tend to give larger star sizes and there would be some chromatic aberration from the corrector plate. But the red halo may be coming from registration being a bit off. I use the CCDstack plug in registration and its highly accurate. Perhaps also the red focus is a tad off or the seeing was worse when you did the red. You could lightly deconvolve the red subs to bring the star sizes in line with the other channels.
I don't think NABG has anything to do with it as that causes blooms which I don't see here.
The focus is out a touch Roger as evidenced by the visibility of the secondary shadow in a number of stars.... that also might be symptomatic of a highly curved field.
Clive if the focus was off enough to see the secondary in the stars the galaxy would be a lot more blury The stars have an artifact in the centre due to processing of blooms and such.
Nice Roger great effort for the focal length and exposure times {tracking LX200 ?}. Do you check the focus before each exposure or is it automated? Good weather lately in Perth not to hot to get out the gear out
It re-focusses every hour, and it's automated. Focus isn't always perfect but tends to be afected more by thin cloud or atmospherics than focus drifting.
Quote:
Originally Posted by gregbradley
That's not bad Roger. Firstly that is a tough target. Those 2 galaxies are very small and you are imaging at over 2.5 metres. SCT's do tend to give larger star sizes and there would be some chromatic aberration from the corrector plate. But the red halo may be coming from registration being a bit off. I use the CCDstack plug in registration and its highly accurate. Perhaps also the red focus is a tad off or the seeing was worse when you did the red. You could lightly deconvolve the red subs to bring the star sizes in line with the other channels.
I don't think NABG has anything to do with it as that causes blooms which I don't see here.
Greg.
Thanks Greg. I'm glad you say SCT's give larger stars, makes me feel slightly better. Registration was done in CCDStack but when combining the CRGB frames I found it didn't get perfect registration
I've fixed up the stars somewhat - the problem with their centre was the "Dust & Scratches" I was using in Photoshop to remove the hot pixels in the colour channels was causing artefacts in the stars. I've backed off the strength of the filter and have an improved result.
A nice little image, Roger. I reckon you could tart it up a bit with processing. I'd have a go at a slight deconvolution, add a bit of noise reduction, reduce the star sizes, tweak the colours and do a little wavelet sharpening.
A nice little image, Roger. I reckon you could tart it up a bit with processing. I'd have a go at a slight deconvolution, add a bit of noise reduction, reduce the star sizes, tweak the colours and do a little wavelet sharpening.
Cheers,
Rick.
Thanks Rick. I've never tried deconvolution. Any tips? (PixInsight or CCDStack).
I've tried reducing star size in PS before with occasional success, but not enough success to try it often. I'll have to dig tha on up.
Hi Roger,
I think you have done a pretty good job with the equipment and exposures at hand. I don't know what your local seeing conditions are but this often can dictate the achievable SNR in the small galactic features that you are trying to capture here. Galaxies are small and faint and can be that much more difficult to get a nice looking result out of.
Perhaps I can offer you a comparison image, similar object, scope (C11 f/10), exposure times, same camera (ST8-XME) in 2-3" seeing. Attached is NGC 6632 in Hercules, 12.9 mag, 3' x 1.1' diameter galaxy, that I did back in Ottawa in 2007. The full details and a linearly stretched 200% crop are on my website here: http://www.faintgalaxy.com/ngc6632.htm
I used ~5 hrs of lum and 4 hrs total of RGB, but binned 2x2. The attached image is only cropped due to frame overlap from dithering. You can see the ST8 starts to pick up the field curvature and coma inherent in the SCT. The star bloat you mention is pretty normal for average seeing at long focal length (should it really be labeled star bloat if it is normal!?!). The NABG is not a problem in any way for this kind of image with no bright foreground stars. The ST8 loves more blue data since it's red response is so strong (>85% QE), and of course blue is always more affected by the seeing, always try to get your blue frames high in the sky and more of them than red!
The funny colors on different sides of the stars I think are due to small collimation shifts and varying atmospheric dispersion from capturing at different altitudes and orientations. I have them to a lesser extent in most of my SCT images also, it is an endemic problem at long focal length.
I'm also attaching the processing log for this one, lots of retries at the processing after going down blind alleys. I threw out a lot of frames here (14 L and 50 RGB) for cosmetic and seeing conditions.
One thing you could try is an Lab combine, I have always found it works a lot better than straight RGB combine, especially when the color frames are noisy but the L channel is pretty good, which I think is your case here. Note I have less color data but it is binned 2x2, which is big boon for f/10 + galaxy + ST8 + avg seeing. The 15e- read noise with color filters and long focal ratio needs very long subs to get sky limited. Binning 2x2 in avg skies really has no downside and gets you sky limited in a reasonable sub length in this case.
Afterwards, I often use PixelMath to blend the deconvolved version and the original to reduce the strength of the effect - especially if there's any hint of Mike Sidonio's dreaded worms.
Cheers Very interesting. Will take some experimenting ...
Quote:
Originally Posted by ericwbenson
Hi Roger,
I think you have done a pretty good job with the equipment and exposures at hand. I don't know what your local seeing conditions are but this often can dictate the achievable SNR in the small galactic features that you are trying to capture here. Galaxies are small and faint and can be that much more difficult to get a nice looking result out of.
Perhaps I can offer you a comparison image, similar object, scope (C11 f/10), exposure times, same camera (ST8-XME) in 2-3" seeing. Attached is NGC 6632 in Hercules, 12.9 mag, 3' x 1.1' diameter galaxy, that I did back in Ottawa in 2007. The full details and a linearly stretched 200% crop are on my website here: http://www.faintgalaxy.com/ngc6632.htm
I used ~5 hrs of lum and 4 hrs total of RGB, but binned 2x2. The attached image is only cropped due to frame overlap from dithering. You can see the ST8 starts to pick up the field curvature and coma inherent in the SCT. The star bloat you mention is pretty normal for average seeing at long focal length (should it really be labeled star bloat if it is normal!?!). The NABG is not a problem in any way for this kind of image with no bright foreground stars. The ST8 loves more blue data since it's red response is so strong (>85% QE), and of course blue is always more affected by the seeing, always try to get your blue frames high in the sky and more of them than red!
The funny colors on different sides of the stars I think are due to small collimation shifts and varying atmospheric dispersion from capturing at different altitudes and orientations. I have them to a lesser extent in most of my SCT images also, it is an endemic problem at long focal length.
I'm also attaching the processing log for this one, lots of retries at the processing after going down blind alleys. I threw out a lot of frames here (14 L and 50 RGB) for cosmetic and seeing conditions.
One thing you could try is an Lab combine, I have always found it works a lot better than straight RGB combine, especially when the color frames are noisy but the L channel is pretty good, which I think is your case here. Note I have less color data but it is binned 2x2, which is big boon for f/10 + galaxy + ST8 + avg seeing. The 15e- read noise with color filters and long focal ratio needs very long subs to get sky limited. Binning 2x2 in avg skies really has no downside and gets you sky limited in a reasonable sub length in this case.
Best,
EB
Thank you Eric for your great feedback Very interesting to have such detailed information for comparison.
Certainly the collimation of my SCT varies greatly from one side of the sky to the other. It makes sense this would have some impact on the resulting image other than simply loss of clarity. It makes me wonder about interleving the LRGB frames more, so doing a cycle of 1,1,1,1 repeated. Thereby having a more even spread of the colour frames and perhaps helping the overall result.
I'll give LAB another shot perhaps. I tried it once a while ago but haven't used it since.
Afterwards, I often use PixelMath to blend the deconvolved version and the original to reduce the strength of the effect - especially if there's any hint of Mike Sidonio's dreaded worms.
Way to go Roger! A great galaxy pair this one and it looks pretty good to me is this the image that will be on display somewhere..?
Mike
I was reminded of in my search for a merger for the exhibition at work, but this one is a bit small. I'm hoping the Antenae galaxies will come up nice once they reise a bit earlier, as they're a better size for printing and having in an exhibition
great effort for the focal length and exposure times {tracking LX200 ?}. Do you check the focus before each exposure or is it automated? Good weather lately in Perth not to hot to get out the gear out 
