IC443 - Jellyfish Nebula

[jase]

Hi All,
Still processing data from last new moon. I've got a couple more to go, but seem to not have had the time to work on them. Anyway, here's my latest effort; IC443 - Jellyfish Nebula (http://www.cosmicphotos.com/gallery/image.php?fld_image_id=168&fld_album_id=11)

About the target;
IC443 is a supernova remnant located in the constellation Gemini and is believed to have formed approximately 30,000 years ago due to a violent explosion of the star G189.6+3.3. The nebula's shape represents an expanding shell of gas scattered by strong interstellar winds from the explosion. IC443 is located 5,000 light-years away.

About the image;
Personally, I think this area is best suited to shorter focal lengths to pick up the long tendrils of the jellyfish, but always pleasing to see something different. The image is a LLRGB composite comprising of 4.5hrs (L:60min;R:70min;G:70min;B:70min). Not a huge amount of data, but of reasonable quality despite a few hiccups. The image is a hybrid containing data from both Lightbuckets 20" and 24" RCOS. The luminance was aquired on the 24", while RGB on the 20". The reasoning behind the hybrid was the 24" doesn't have the FOV of the STL11k attached to the 20". The 24" has the Apogee U42 which is an ultra sensitivity 2048x2048 sensor. So I used the RGB from the STL11k for the larger FOV and blended the 24" luminance over the top. Sounds relatively straight forward, but there where a few problems. Uncertain as to my mistake, but the MOAG missed the planned guide star on the 20" so the RGB is unguided. Huh? unguided at 4115mm FL is a testament to the PME's capabilities. This did however lead to sligh elongation of stars (as one would expect at such a focal length). I originally thought that this would destroy the data considering my original intention was to also use the RGB as a synthetic lum. However to circumvent the slight elongation I performed minor star shaping using a kernel filter in MaximDL. Elongation is still present but is considerably reduced. Some stars don't line up precisely result in colour finges, but these didn't both me too much so left them be. All other processing was conducted in CCDStack and PS4, both of which are relatively new to me, so will take me some time to get up to speed with them. I stuggled to balance the background ADU between RGB subs - something so simple, yet so hard when you're unfamiliar with the tools. So ended up using pixel math... I'm also still at a loss as the neatimage plug-in doesn't work with PS along with a few other tools like Registar. Yes, I know, I shouldn't have installed Vista64, but I wasn't willing waste the 8gb ram I've installed in the system. CCDStack sure is a pleasure to use, as is PS being 64-bit. Anyway digress, Synthetic lum and 24" lum created and passed through 30 iterations of positive constraint deconvolution to tight things up a little. RGB created and pushed out as an unscaled tiff along with the two lum images before being imported into PS. Levels, curves, the usual drill. I layered the two lums first to blend them in seamlessly - subtle brightness/contrast used to achieve the task, then flattened. Introduced this layer to the RGB and stretched. Created super RGB through flattening and stretch yet again. HPF over entire image, then used colour range tool to select highlights before deleting them to return stars to a natural form. Minor saturation boost to bring out the vibrant hydrogen alpha reds and star colours. As I don't have a noise plug-in I used a Gaussian blur applied via an inverted mask to control the results. Not one of my best, but pleasing none the less. I good test for my new processing system anyway.

Thanks for looking. Hope you like it.:)

[monoxide]

very nice!
unguided at just over 4m fl :eyepop: the elongation is barely noticable

like you said, a pleasing image with a load of fine detail but mabee a bit too much focal length for it to really shine but saying that... its still an excellent image.

[jase]

Cheers TJ! Yes, stupid mistake to not have a guide star selected by the MOAG. Should have double checked. :screwy: The framing of this target took some planning. Its quite a large target so framing was critical given the small field of view. I thought I had it all organised in TheSky so that a guide star was selected...alas come to execute the plan, it all fell apart...so I made the most of the data anyway.

The slight elongation was "partly" resolved using a kernel filter. For those interested, more information can be found here - http://www.princeton.edu/~rvdb/images/deconv/deconv_MaximDL.html

[Hagar]

Great image Jase considering your trials and tribulations. After the initial read I expected a lot less image than the image you have produced.
Your processing skills are a world above most of us.

BEAUTIFULL I wish we could see more of this.

[strongmanmike]

Some excellent detail showing there Jase, I spent some time tracing and scrutinising the tiny fillaments... and eventually became a bit mesmerised....

Looks quite balanced and natural, nice job :thumbsup:

Mike

[iceman]

It's a brain! Lovely image, Jase. Top stuff.

[multiweb]

Very cool. :) Looks like a veil nebula on steroids. Heaps of details. Love the colours. Did you mention tracking problems? Can't see a thing. Top work. :thumbsup:

[jase]

Thanks Doug. A little hit and miss with this one. Didn't mean to talk the image down, but I've certainly acquired better data on these instruments. I think for once, I'm not entirely sure what I'm doing with processing. After upgrading CCDStack from 1.14 to 1.4 and also PS CS2 to CS4, both considerable differences from the previous version. I resisted heading back into software packages I'm familiar with with exception to MaximDL for the star shaping. With more data to process, it should give me sufficient exposure to make improvements. Thanks again.



Cheers Mike. It is an interesting target. Something different for long focal lengths. It fits the frame of the 20" with STL11k exceptionally well. Pleased you liked it.



Thanks Mike. Yes, does have brain like characteristics. Not as much as NGC6888 though - which is on my list for next season. Thanks for checking it out and making comment.



Cheers Marc. Not as large as the veil, but rather interesting none the less. The elongation is present - the colour fringes around some of the smaller stars makes this evident. Thanks for the comments.

====
Thanks all.:)

[Paramount]

Hi Jase
To be honest I have to look really close to see any evidence of star elongation and unless it is so obvious that it jumps out and flicks my ears then I don't look for it as it would mean I am not looking at the most important part, ie the image itself and I think sometimes we are all guilty of looking at images to find out the faults instead of looking at the aesthetic points (as far as I am concerned there is no such thing as a faultless image). This is a cracker with loads of detail in the whispy parts of the nebula,lovely colour balance as well. As you said unguided for over 4m focal length and a scope that is actually above the recommended maximum load capacity for the Paramount ME (when you add on the weight of guider, cameras, accessories, etc. It really speaks wonders about the capabilities of the Paramount (that's why I've got one)
Best wishes
Gordon:thumbsup:

[Bassnut]

An excellent image Jase, you cleaned up elongation nicely, masterfull processing, and compostion is pleaseing. Interesting you are finally useing CCD stack, you might be interested in the one click align plugin included in CCD inspector, very powerfull, give it a try. I quite like the slight blue star halo, gives them a vibrant touch. Neb detail is very good too, withstands much zooming :-).

CCD stack deconvolute is powerfull tool in CCD stack, it works well with megadata, as you have found.

[jase]

Cheers Gordon. I guess I'm more concious of the image problems considering I spent a few days working the data over.

Perhaps its time to do your mount justice and ditch that wide field TMB115 and give the C14 a run. Lets see what you can do! Narrowband, narrowfields are always a special treat to image. I think the only time I'd mount a wide field scope on a PME would be if I was accessing it a few hundred kilometres away in a autonomous robotic fashion. I hope you'll take up the challenge.



Thanks Fred. Yes, I've gone the full loop. I used CCDStack many moons ago (as previously mentioned version 1.14). I found the software limiting. Visiting AIC2008 changed that. Saw some really cool features in 1.4 that I thoroughly enjoyed. I'll blame Adam Block as he told me I should use it.:P In all seriousness however, I'm rather out of my depth with this software. I know MaximDL almost like the back of my hand, but CCDStack presents a different method/flow of processing. It also lead me to getting a PC upgrade so I could blast through mega stacks of subs. Yes, I also purchased the CCDIS/P for image alignment. I'm using the 64-bit version, not that I think its that memory intensive. The plug-in is sweet, but still does not handle huge image scale differences. I'm working on an image at the moment where I wanted to also align some RGB data from my FSQ with the longer focal length of the RC - *kapow* wouldn't accept it. Needed to manually align the images. This is where Registar really comes into its own. Shame I can't run it anymore - doesn't support Vista64. eeeek!:( Will push the developers an email I think considering 64-bit computing will inevitiably become mainstream. PS CS4 is a real treat in 64-bit mode very smooth and feature rich. Still trying to work out all the cool new "toys" such as tricky adjustment layers etc. Haven't played around too much with CCDStacks deconvolution. I did in this image, but was rather tame with it. Didn't push it hard (as I had No.F.Idea what I was doing;)) I'm having some real problems in balancing the individual RGB masters. Driving me nuts. I balance the background ok, but stars get a deep magenta tone to them. Its happened to two images I've been "attempting" to process now. I've got to be doing something wrong or the data is wacked and I need to acquire more. You're a CCDStack guru, so may need to zap your brains offline if I don't get any joy. Thanks again for your support.:)

[Bassnut]

Yes, the work flow in CCD stack is very different to DL, I found CCD stack painfully manual to start with.. The image manager is more powerfull than DL tho. DL is faster, but CCD stack becomes more intuative with use. Being able to see the rejected data before combine allows powerfull subjective analysis (and selection of the top (reference) image, not mentioned in the manual, but counts, which image is selected to start influences many CCD stack processes). Colour balance I also found to be a pain, oddly though, just sometimes. When it was, I colour combined in CCD stack, converted to TIFF and import to PS (as you know, fits liberator cant convert colour FITS, a real pain). Otherwise, how you import mono fits into PS with liberator makes a huge difference to colour balance, the less fiddling (relatively)between RGB subs the better.

Another huge difference between CCD stack and DL is, DL will accept a much wider range of differences between subs with data reject, and just goes with what ever. Sometimes CCD stack requires many normalise itterations or bad sub rejects before allowing data reject.

[jase]

Funny you mention that Fred. I found to deliver great results with mediocre data i.e. not mega data, you really need to keep an eye on the weights. It got to stage where I began tweaking these manually to choose the better sub. Rather extreme as the normalisation should have address this. Will see how long it takes me before I convert back to what I'm use too. To no surprise, but there isn't one software package the does it all really well. I often find myself diving in and out of different tools, paying close attention to preserve the bit space as I go. Things aint simply any more if you want reasonable results I guess.:shrug:

[jase]

...and here's a good reason why you should check your subs before you perform any data rejection...
May take a little while to load - ~760k:D
IC443 - Interstellar Loper (http://www.cosmicphotos.com/reprocessed/IC443-Interstellar-loper.gif)

[Bassnut]

lol, thats cool !

[multiweb]

I usually do a quick mean combine after a sigma reject 2% top then look for "streaks". Then I zoom in and run the blink to see what's moving. If the moving bit is going through stars I paint it with the free hand draw tool, if it's in a relatively dark area I draw a small selection box encompassing the streak then do a reject range until it paints it and "apply all", then blink to double check the stack. The different rejections are cumulative. That's why I run Sigma first always using the best frame I aligned from and normalised from at the beginning (I check it in CCD Inspector prior to go in CCD Stack). I get a few asteroids but a lot of my streaks are hot pixels that escaped dithering.

[Ric]

Fantastic image Jase, it certainly has the appearance of a jellyfish.

The folds and swirls of nebula is amazing.

A fine image.

[Garyh]

I`m a bit behind looking at the DSO images lately but I just have to say this is a beautiful image Jase!
Your processing is impeccable like always!
:thumbsup:

[niko]

stunning...just stunning!!!

[jase]

Thanks Ric. I feel the jellyfish is more pronounced in wider field images of this target considering you can also visualise the tendrils below. Thanks for checking out the image and making comment.



Cheers Gary. Pleased you liked it.:thumbsup:



Thanks Niko!

===========
For reference, the animated object is (37864)1998 FJ10 (mag 18). No new discoveries I'm afraid, but there were two other objects in the data acquired. Tim Luce, a regular user of the Lightbuckets facility, worked the data over and submitted a MPC report. Excellent work performed by Tim.

ACK MPCReport file updated 2008.12.15 22:49:22
NET USNO-B1.0
K03E18B C2008 12 04.25753 06 18 32.11 +22 39 18.6 19.8 R H11
K03E18B C2008 12 04.26521 06 18 31.73 +22 39 19.4 19.4 R H11
K03E18B C2008 12 04.26899 06 18 31.36 +22 39 19.9 19.8 R H11
K03E18B C2008 12 04.27659 06 18 31.10 +22 39 21.7 20.1 R H11
K03E18B C2008 12 04.28047 06 18 30.73 +22 39 19.5 19.6 R H11
37864 C2008 12 04.26899 06 18 04.14 +22 55 42.3 18.8 R H11
37864 C2008 12 04.27281 06 18 03.90 +22 55 42.3 18.9 R H11
37864 C2008 12 04.27659 06 18 03.67 +22 55 42.3 18.4 R H11
37864 C2008 12 04.28047 06 18 03.44 +22 55 42.2 18.5 R H11
37864 C2008 12 04.28047 06 18 03.44 +22 55 42.2 18.5 R H11
77783 C2008 12 04.26140 06 17 58.41 +22 41 24.6 19.3 R H11
77783 C2008 12 04.26521 06 17 58.12 +22 41 24.8 19.1 R H11


=============
Anyway, thanks again all for your support - appreciated.:D

[richardo]

Hi Jase,
always good to get something different, and from the other Hem no less...
New Mexico boast some pretty clear skies and with a RC scope, cookin'!

I like this object as I've seen so many over the years from our budding Northern brethren.

I think this view with the narrower fov is pretty nice, showing nicely the structure of the delicate remnant threads.
Looks as though both evenings were lacking a little in clarity judging by the softness of the stars, still takes nothing away from the overall image.

Good to see the bad weather on your side of Oz doesn't stop you getting some imaging in :P :D

All the best
Rich