Most renditions of the Cat's Paw Nebula emphasise the "paw print", but I wanted to show the beautiful dark ripples throughout the nebula and the interesting but faint background.
Here's my attempt with Ha data taken in my back yard this past month:
This a 2x2 mosaic with each panel consisting of 20 x 60 min subs (bottom-left panel has 26 x 60 min) for a total of 86 hours, processed using PixInSight. Data was captured across 17 nights of August, averaging 5.1 hours/night (ranging from 1 to 7 hours/night).
Equipment: GSO RC8 scope, SX AO-LF + Lodestar guiding, SBIG STF-8300M camera binned 2x2, Astrodon 3 nm Ha filter, and Skywatcher NEQ6 mount.
Interestingly, recent research from June this year suggests that this stellar nursery could be a mini-starbust ("a concentrated area of extremely rapid star formation usually only seen in distant galaxies"). It's apparently even more prolific than the Orion Nebula - with 200,000 Suns worth of star-creating material, and tens of thousands of new stars already formed. The research shows that most of the star creation is currently happening in the dark regions between the pads of the "paw".
that is a stunning image Dave. The detail is really interesting and it is a pleasure to spend time just looking at it and imagining how it might all have come about - wow. regards Ray
that is a stunning image Dave. The detail is really interesting and it is a pleasure to spend time just looking at it and imagining how it might all have come about - wow. regards Ray
Thanks Ray!
Quote:
Originally Posted by tilbrook@rbe.ne
Wow Dave!
That's the deepest image I've seen NGC 6334, in the high res version it feels like you could swing through the tendrils of nebulosity.
Stunning work
Thanks Justin! I spent a bit of time scouring Google for deep Cat's Paw images, with the aim of going just a bit deeper There's definitely more detail in the data to pull out, so I foresee repros during my summer holidays
Quote:
Originally Posted by RickS
Fantastic achievement, Dave! That's an extremely deep and well stitched mosaic.
Thanks Rick! Cheers for the advice and patient viewing of my works-in-progress along the way too
Quote:
Originally Posted by dvj
THAT, is a lot of hours!
After a terrible start to the year, we were very lucky in August to have a great run of clear nights.
Until I can afford a larger/faster scope, or a larger / more sensitive CCD chip, time is the only dimension I have to go deep...
I'm especially impressed by your giant mosaic, now having experienced some of the pain of stitching
How did you settle on 60 min subs and why opt for 2x2 bin?
would 1x1 have provided finer details? just curious.
Cheers
Alistair
Probably due to his seeing, the KAF8300 on an RC8 has an image scale of 0.6 arc seconds per pixel. There aren't many people who will have the seeing to resolve that so binning 2x2 would bring that too 1.2 which is probably around right.
How did you settle on 60 min subs and why opt for 2x2 bin?
would 1x1 have provided finer details? just curious.
Thanks Alistair! My scope and camera are both fantastic on their own, but unfortunately they're poorly matched. Peter's point about image scale is spot on, especially as my seeing is 3-5''. Sensitivity is also an issue - according to ideal CCD exposure calculators, at 1x1 binning I'd need 30 min Luminance and 80-120 minute Ha exposures... not practical!
I starting using 2x2 binning to speed things up a bit, but then found the histogram was still bunched up on the left using 20 min subs. I tried 60 mins subs (the maximum for my camera) and found that the sky background was still extremely dark - 0.7% sky background with no moon - so I stuck with it.
Ideally, I'd pair this camera with a scope of around 700-800 mm focal length - but I haven't been able to find one that has a reasonably large aperture, is mechanically and thermally stable, has ample back focus for adaptive optics, and is cheap. For the $700 or so I paid for my GSO RC8 CF (used), I'm very happy with the image quality
Getting round stars with 60 min subs was easy using adaptive optics (0.5 & 1 sec guide exposures), even under windy conditions. I think I discarded fewer than 5 subs in total due to tracking errors to get these 86 subs?
The main downsides of 60 min subs is that you need a lot of data in order to use more advanced data rejection algorithms like sigma clip and linear fit, and you get a lot of cosmic ray hits. The stacked panels were terrible at 5 hours/panel (visible artefacts), but once I reached 10-15 subs the noise cleaned up and the really faint nebulosity started to pop.
Actually I think kudos go to Fred for the original idea - single best thing you can do to improve the quality of your images, regardless of gear or your sky conditions. Assuming its being run well, as yours definitely is.
Tip of the hat, Dave. Very well executed image. So much on display and to dwell over for lengthy periods of time. The mosaic certainly adds pleasing resolution to what is a well frame subject. No confusing message. You certainly made the most of the clear nights in which by the sound of things provided solid data to work with. The end result at least indicates this. Well done. More please...
There is lots to see in there and clearly a lot of work has gone into it
Now, this is not a criticism and Ivo may hammer me here ...but to me some of the detail looks like it has been kind of "created" by the processing in a similar way that the minimum filter can. Don't get me wrong it is certainly a detailed image but looking at some of those thin wisps they look like they were fatter streaks made artificially "thinner" by the software . So what you might say (and well yes, so what it may be) but looking at the spotted background I am not totally convinced that some black magic isn't going on here...at least to some degree Hope you aren't offended by these comments just an observation, it is an amazing deep nebula image
Tip of the hat, Dave. Very well executed image. So much on display and to dwell over for lengthy periods of time. The mosaic certainly adds pleasing resolution to what is a well frame subject. No confusing message. You certainly made the most of the clear nights in which by the sound of things provided solid data to work with. The end result at least indicates this. Well done. More please...
Thanks Jason. Much appreciated, especially considering your imaging experience
We definitely had a lucky run of good weather in Brisbane in August (a pleasant surprise considering the first half of the year). I certainly wasn't originally expecting to be finished before October!
Quote:
Originally Posted by strongmanmike
Wow....pretty incredible to look at Dave
There is lots to see in there and clearly a lot of work has gone into it
Now, this is not a criticism and Ivo may hammer me here ...but to me some of the detail looks like it has been kind of "created" by the processing in a similar way that the minimum filter can. Don't get me wrong it is certainly a detailed image but looking at some of those thin wisps they look like they were fatter streaks made artificially "thinner" by the software . So what you might say (and well yes, so what it may be) but looking at the spotted background I am not totally convinced that some black magic isn't going on here...at least to some degree Hope you aren't offended by these comments just an observation, it is an amazing deep nebula image
MIke
Thanks Mike! No offended at all, if anything I'm actually rather curious... I think I know the types of false-signal artefacts you're referring to; they're particularly infamous amongst the planetary imagers.
To the best of my knowledge, all the structures that I can see in the final result are visible in the raw screen-stretched-only stack (so they're "real" as far as I'm aware).
I've attached two crops with no processing. The first image is of a very high SnR region that has been linearly stretched (0, 0.5, 0.33) in PI, followed by 200% resampling using nearest neighbour for clarity. The second is of a very low SnR region in the background - linearly stretched (0, 0.5, 0.0125) and 200% resampled again. In both cases, there was *no* other processing - no wavelets, no multiscale processing, no sharpening, no deconvolution, etc applied.
When I compare the raw stretches to the final result - I can see the same fine structures in both. The "spotty background" is just noise from the depths of my CCD due to the extreme stretching (20 subs isn't really enough to smooth out the background given the amount of thermal noise I have with 60 min subs @ -10 deg C). At 86 hours, I'm definitely at the "diminishing returns" point of further data collection
I think the "black magic" you refer to comes from PI's amazing ability to compress dynamic range. There's a huge amount of subtle detail in the Cat's Paw, but I think it tends to get swamped by the dynamic range. I'm very much a PI newbie though so it's sledge hammer strokes rather than fine tooth comb at this stage
Anyway, thanks for your honest opinion. I'm curious what you think (or anyone else) after seeing the unprocessed crops?
Quote:
Originally Posted by astronobob
A Truely remarkable piece of Work Dave Just shows what Mega Data can do, 100% credit to Ya
Thanks Bob! This is pretty much the only astro thing I've been working on since we last caught up in Leyburn... definitely a labour of love
...but to me some of the detail looks like it has been kind of "created" by the processing 
in a similar way that the minimum filter can. Don't get me wrong it is certainly a detailed image but looking at some of those thin wisps they look like they were fatter streaks made artificially "thinner" by the software 
. So what you might say (and well yes, so what it may be
Just shows what Mega Data can do, 100% credit to Ya 
