M83 - Southern Pinwheel Galaxy

[jase]

Hi All,

This image represents the pinnacle in one of my imaging chapters – image scale amalgamation i.e. successfully combining data acquired at different focal lengths. Most importantly, this project prompted the redevelopment of some processing techniques that I had thought about, but never acted on until now to deliver a pleasing end result. I have explained parts of the process below. So, without further ado, I’m pleased to present M83 – Southern Pinwheel Galaxy (http://www.cosmicphotos.com/gallery/image.php?fld_image_id=145&fld_album_id=12).

The Southern Pinwheel Galaxy (M83) is located in the constellation Hydra and is a member of the Centaurus group of galaxies which includes peculiar radio galaxy Centaurus A (NGC5128). M83 is classified as a between a normal and barred spiral galaxy based on its characteristics. The immense spiral structure of the arms contain various blue and red knots of star clusters and diffused gaseous (HII) regions respectively. M83 is the southernmost galaxy in Messier’s catalogue and resides approximately 15,000 light years distant.

About the image (for those interested);

This image is an LLRGB composite totalling 6.1 hours (Luminance:100min;RGB:90min each). The luminance data was acquired from GRAS008 – 12.5” F/9 RC operating at ~2800mm. The Chrominance (RGB) data was acquired using the FSQ-106ED 4” F/5 Petzval APO operating at 530mm. As one could image, we are dealing with vastly different image scales. The chrominance data was solely to be used for colour information on the high resolution luminance data. Calibrated all frames – usual process. Firstly, I worked on the luminance. Excellent data was obtained from the RC (10x10min subs) shooting the target while it was +80 degrees (no AOL used here!). Very slight gradient present which was removed using GradientXTerminator, then continue with a multi strength deconvolution blend to bring out the details. Minor background noise removal in NeatImage. On to the chrominance - from past experience, upscaling of any data requires good quality data to ensure noise isn’t introduced into the image. I dedicated an imaging run solely for RGB, collecting 90min in each channel (6x15min subs). In the past, I’ve found that simply upscaling the chrominance to match the luminance gave strange results, in particular with the stars (due to the considerably large difference in image scales). So to counter act this, I first downsized the hi-res luminance to match the low-res RGB chrominance. The two were blended using lighten mode, blurred, boosted in colour saturation and flattened. This new image had become the superRGB chrominance frame which I then upscaled to match the luminance frame for final composition. The end result provided a greater improvement on star matching and colour correction. The luminance was layered twice so I could bring back the colour hues while keep detail. Minor noise reduction performed on the chrominance layer. Star halos subtly accentuated using colour masking. Image flattened, before final selective noise reduction. Seasoned to taste. The shear number of faint fuzzy galaxies that dot the background is impressive in the full res version.

I’ve also imaged its northern counterpart M101, which can be viewed here (http://www.cosmicphotos.com/gallery/image.php?fld_image_id=133&fld_album_id=12). Though I think this version needs some colour balance work considering it was processed on a old laptop, not a calibrated monitor (which I now have access to).

The full size image presented (2400x1800) is a 75% image size reduction of the total image. The 100% is remarkable – will use this for prints. The crop shows M83 slightly offset to show the brilliant blue star (HIP 66549 – spectral class F2III/IV). I’ve attached TheSky imaging plan to provide an insight into the composition planning along with some background PGC galaxies. I composition is actually incorrect as the instrument rotator should have been at 168’, instead of 186’ shown. However, I still found a guide star that allowed for 7sec exposures!

I hope you enjoy it as much as I did processing it! :) As always, all comments welcome.

[renormalised]

I'm going to recommend you for a position as an observer once JCU gets its 1m scope and observatory up and running. For an amateur, that is a remarkable piece of work!!!!. Well done!!!!.

You should be doing a higher degree in Astronomy.

[EzyStyles]

i have seen alot of images of M83. That is seriously the BEST i have EVER seen!. Jase, takeover the Malin award.

The Hi Res version is stunning. i stilll have abit of saliva around my mouth.

[glenc]

Jase, a magnificent image of a magnificent object. :eyepop:
M83 (mag 7.1) was the third galaxy to be discovered with a telescope, after M33 and M32.
http://seds.lpl.arizona.edu/messier/xtra/history/dis-tab.html

[prova]

Spectacular!

[jase]

Thanks Renormalised.:) It would be my absolute pleasure to get my hands on some quality data from a 1m class scope for image processing. Even if you simply feed me quality monochrome luminance data. I'll find away of processing it! Seriously though, if you're looking to perform some collaborative work, just let me know.



Cheers Eric. Malin award!:lol: Actually, this image is unlikely to be in the panel I'll present this year. Of course, I can't give too much away ;)



Thanks Glen. Pleased you like it.:thumbsup:



Cheers Prova.

[Tamtarn]

Absolutely brilliant Jase. The clarity, detail, colour balance, smoothness and the number of background galaxies clearly defined makes this an absolutely brilliant image Jase.

By combining the data from the different focal lengths you certainly reached that pinnacle.

[iceman]

OMG that is absolutely magnificent.

Truly stunning, Jase. What an inspiration. :bowdown:

[Peter Ward]

Jase,

A benchmark image for sure. Nothing not to like here! Well done :thumbsup:

Cheers
Peter

[bloodhound31]

Truly Beautiful Jase, gives us all something to aspire to.

Blessings,

Baz.

[jase]

Thanks Barb and David. I certainly put the hard yards into this image. Will be honest, the stars still caused me some grief with the upscaling of RGB data so took some work to bring back the aesthetics. I feel I still managed to keep the image looking natural. Thanks again.



Cheers Mike. Appreciate the kind words. As they say, once bitten by the imaging bug, you're poisoned for life - there is no return from the dark side.:lol: Seriously, I did find the data scaling challenging especially at the focal length difference we are talking about. Seems to have come together quite well.



Thanks Peter. Any chance of getting some time on that 14" RC of yours? Not sure if I'm happy dealing with the light pollution gradients though ;) (such a waste of quality gear - I have a dark sky site in Gippsland if you want to ship your gear):P Actually, I may soon have access to a 14" RC in WA, 20" and 24" RC's in the US from lightbuckets.com - Steve is still bringing things online. I'm getting edgy to nail some targets with either the 20" or 24".:D Should boost my imaging output and experience. Surely there are clear night skies somewhere in the world!:lol:



Cheers Baz. Keep at it mate, we all have goals to work towards. Make them realistic and you'll succeed. Thanks again.

[Matty P]

Definitely the best M83 I have ever seen!

Amazing!

[bluescope]

Looks great Jase !

:thumbsup:

[gbeal]

What can I say? It is a real credit to you. The actual imaging is always secondary to the result, processing is where it happens.
Well done.

[dugnsuz]

The large version is stunning jase - well done, a fantastic result.
Doug

[jase]

Thanks Matty. Pleased you liked it.:)



Cheers Steve!:thumbsup:



Darn right Gary. The workload is 20%-30% acquiring the data, 70%-80% processing it. Once you've got the data, you need to know what do to with it. Mask the defects and highlight the good detail. Sounds simple, but in practise not always easy. Don't be afraid to try something different. Once you've got a few routines that work well for you, stick with them - but then add some tasks to extend and/or challenge yourself. Thanks again for your comments.



Thanks Doug.:thumbsup: The large version is a 75% reduction! The full frame (3900x2600 - 100%) is a visual treat.

[Alchemy]

Stunning image Jase. Beautiful detail in the dust lanes had a look at the large image .... and youve got an even bigger one:cool:

[skwinty]

Hi Jase,
Absolutely awesome! Definately worthy of an award and world wide publication.
Well done and make sure we see lots more like that!:thumbsup:

[Zuts]

Hi Jase

Fantastic image Jase. Even just reading the notes lets me learn something re being able to mix FL's. Just looking at it encourages me to get out there and capture some of my own images.

Paul

[Bassnut]

Hi Jase

Tastefull colour balance on the M83, a pleasing view. The M101 is a tad blue, and looks like some blue fringing on the brighter stars. I think youll find G8 now has external guide. Ive taken flat runs at Morrook on as my task, so if you have a problem with them please advise.

[Ric]

Hi Jase, If I can achieve an image to 1/10th of what you produce I will be a very happy chappy indeed.:D

Your images are stunning and M83 is no exception, a truly exceptional image, I found it to be the type of image to sit back with a medium dry red and just enjoy.:)

I eagerly await your next image and wish you all the best with your Malin entries

Cheers

[peter_4059]

stunning.

[jase]

Thanks Clive. Indeed the larger image is a corker!



Cheers Steve. I'm not one for big noting myself. This is the only forum I post on at the moment. I use to post on the SBIG yahoo forums sometime back, but I couldn't keep up with all the posts there.:lol: May return someday. I'm not APOD calibre like the big boys.;)



Thanks Paul. This comes back to a recent post I think I replied to you on regarding image registration (or I think Mike (Iceman Mike that is) mentioned it :confuse3:)Anyway - one word - RegiStar (http://www.aurigaimaging.com/). Its the best US$150 you'll spend. It opens many avenues like what I've presented. Thanks again for your comments.



Hi Fred, Indeed M101 is a tad blue (as I mentioned in the original post - colour balance was out). Was not aware G8 had external guide capabilities now. This must be very recent as imaging through the Ha filter a couple of weeks ago, I saw large guide exposures, in some cases resulted in elongated stars. Strange you mention external guiding...was in contact with a few RC owners, one a 14" and 20" - both doing external guiding. Was rather surprised that 20" could be guided externally considering the long FL. I guess if you used something around the 1000-1500mm mark you'd be ok - considering sub pixel guiding algorithms available today. Haven't given it much thought. I hope the PIR remains in the imaging train so users can continue to define the composition. Will there come a time that you'll log a users imaging run and produce flats at the same PIR angle for accuracy? I do have some qualms with the calibration library. In some cases, the calibrated data looked black clipped which gives the impression of a bad S/N ratio with the raw data. Close examination of the FITS headers looked fine. The correct calibration files were applied. I'll send you some Ha light frames of NGC6164 for your opinion. Thanks again for your comments.



Thanks Ric. Once in a while I hit the right note, this would be one of them. Its rather strange in that I've enjoyed my recent work, but I keep going back to my original FSQ images and get the most pleasure viewing them. As much as I like long focal length work, there has always been a keen interesting in wide field views such as sweeping shots across the galactic heart or trying to cram as many object pairings into a single frame. I'll always come back to wide field imaging. It is fun...and hey, when you're imaging at +3 arcsec/pixel, it doesn't matter how crap the seeing is...you'll still get a killer image.:lol:



Cheers Peter. Pleased you enjoyed it.:)

[montewilson]

Train spotters will have noticed a small but obvious error in the description Jase has written for the web page of this astro-image.

No one will however have failed to notice how this is really a first class image.

Don't make me tell you all what the error is............

Well done buddy!

[dcalleja]

Jase
A brilliant image. I lost count of the background galaxies. Thanks so much as well for the detail around your processing - it really gives me some more clues as to how to improve my own.

[GrahamL]

This ..must.. be the image you mentioned last night glen :)

wow thats a cracker jase:thumbsup:

[tornado33]

Excellent shot, the high res version is stunning
Scott

[Garyh]

Absolutely gorgeous Jase!
So much detail in the dust lanes and the little Ha regions throughout.
Love the colors!
Tried looking at the high res version but I could only get half a image :(
darn dialup for ya..
cheers Gary

[seeker372011]

always enjoy your images-and also your processing notes-thanks for sharing
narayan

[spearo]

Very nice Jase
now I know what its supposed to look like!
I'll be reprocessing mine a bit i think
frank

[jase]

Thanks Monte!:thumbsup:



Cheers Dan. :)I hope others will try this processing technique of matching different focal lengths. Considering the RGB data was only used for colouring the luminance, there is no need for it be high resolution.



Thanks Graham!



Thanks Scott. Pleased you like it.



Cheers Gary. Its gratifying when everything comes together. Happens on rare occasions. Thanks again.:)



Thanks Narayan.:thumbsup:



Cheers Frank. Pleased you like it.

=======
Also, thanks to the folks at JCU for the emails. Much appreciated.
=======

Once again, thank you all for taking the time to view and make comment on this image. Back to capturing more photons as I've ran out of data to process!:lol:

[marc4darkskies]

A belated well done Jase - a magnificent rendition in every respect. :thumbsup::thumbsup::thumbsup:

I'm working on my version of this now - won't be able to beat the res of an RC though :(

Cheers, Marcus

[Peter Ward]

Correct me If I'm wrong, the the res in Jase's image seems to vary. On one hand the Binary on the outer arm (at 2.00 o'clock) is not cleanly split, yet the stars in the core show remarkable depth and resolution (that I have not seen in anything less than a 20") Must be those dark Gippsland skies? Fascinating !

[sjastro]

Wonderful image. Jase. I'm interested in your comments on processing/imaging breakdown. If anything the ratio is an even split in my case (although it is hard to quantify). I throw out up to 50% of my raw data images if the results fall below a particular FWHM value. That value depends on the type of object I am imaging.

Regards

Steven
http://users.westconnect.com.au/~sjastro/small (http://users.westconnect.com.au/%7Esjastro/small)

[jase]

Thanks Marcus. :) Look forward to seeing your efforts!



That's the downside of not having a smooth transition in your processing routine Peter. How did I get such resolution I hear you say. I teased it out - Multi Stength Decon Blend (MSDB) layers from none other than Ken Crawford - http://www.rdelsol.com/Presentations.html. The transition masks could have certainly been smoother. I do blur and alter the opacity of all masks to assist in the transition, but perhaps not enough this time. You sure you don't want to give me remote access to that 14" RC of yours.



Cheers Steven. A 50% cut on your subs is pretty harsh. Certainly a low efficiency rate. I'd be trying to increase the output. Whether using AO or carefully selecting the nights where seeing is great to go deep. Its not always a bad thing to be a wide field woose and image even when seeing is woeful. Thanks for your comments.

[Peter Ward]

I suspected as much ;) Tidy effort just the same! :)

Cheers
Peter

[Dennis]

Hi Jase

Well, I’m on 3rd cup of Earl Grey tea and 4th biscuit and I’m still mesmerised by this extraordinary image of M83, filling my 1600x1200 display. Actually, image is a somewhat inadequate word to use in describing this masterpiece; this is easily the best M83 portrait I have ever seen, simply breathtaking in its portrayal of this island universe.

I think that this is the only high resolution, large image scale M83 where I haven’t been aware of noise in the faint, outer regions. And the detail in the core – wow! The dust lanes, knots of stars and nebulosity are so strong and clear – no signs of any over processing. You have been very honest with the data and it clearly shows.

Move over Rob Gendler, there is a new kid on the block.

Cheers

Dennis

[sjastro]

It ultimately depends on what I am imaging. Fine details in galaxy images requires a greater degree of discrimination on the quality of the raw images than say some indistinct very rarely imaged object where I might utilize all the raw data.

I don't put as much emphasis on image processing as others.

Regards

Steven

[Bassnut]

Jase. On a second read, the info you provide on combining different FLs is very interesting indeed, and the Ken Crawford tutes are just awesome, extremely usefull and well presented. Most usefull afternoon spent just on reading all this, best waste of time ive had in months, beats work ;-).

[strongmanmike]

Fantastic Jase!!! very well compiled image undeed. I can just see your face as this was coming together :thumbsup:

Your M83 looks quite like RJ's actually but even though RJ's is showing some remarkable structures, especially in the outer arms compared to yours, your over-all look is more natural to me and I also like your slightly wider more interesting FOV. I find as fantastic as RJ's images are they always have a science art painting look to them that I can't quite warm to 100% (bit like narrow band images actually :P)

http://www.cosmotography.com/images/small_m83.html

I've found combining data from different scopes rewarding for some time now. I use Astroart for much of my image processing and it's pretty easy to combine data from different scopes IMO and I am surprised more people don't do it :shrug:

This was compiled from data taken with three scopes - an ED80, 12" SCT and my 6" Starfire:

http://www.pbase.com/strongmanmike2002/image/64128103/original

Again, a truly excellent M83 :thumbsup::thumbsup::thumbsup:

Mike

[RB]

Absolutely jaw dropping image Jase, stunning work my friend !!

I keep coming back to this image to have another look.

Also thank you for the link to the MSDB tut.