PDA

View Full Version here: : Spare any change?


jase
03-10-2007, 12:51 PM
$ Here's a coin $...

Hi All,

Well it’s been a while between new posts in the DS forum. OS work commitments and other tasks have consumed much of my time so haven’t been as active in the imaging scene as I’d like. Still keeping tabs on the imaging work submitted by others which acts as a catalyst for getting out there and collecting photons.

Anyway, without further ado, I’m pleased to present NGC 253 (http://www.cosmicphotos.com/gallery/image.php?fld_image_id=115&fld_album_id=12), dubbed “The Silver Coin Galaxy”.

The barred spiral galaxy NGC 253 is the brightest member of the Sculptor group of galaxies. The group is estimated as being nearest to the Local Group of galaxies which includes our Milky Way galaxy. NGC 253 is one of the dustiest galaxies known. Dark dust patches conceal much of its spiral structure, also masking the HII regions. The galaxy appears elongated as we see it from an edge-on perspective. In spite of the dust, two spiral arms are subtly visible towards the edge of the galaxy with numerous bluish clusters of young stars. NGC 253 is approximately 10 million light-years distant.

About the image;
This is an LRGB composite consisting of 5.5 hours of data (L:120min; R:70min;G:70min;B:70min) taken over a few nights to beat the rising moon light contaminating the chrominance data (RGB). This isn’t much data to work with for such a slow telescope (F/9) and a dim target such as NGC 253. I wrestled for hours chasing colours. I haven’t got a good handle on the Astrodon filters and the colour balance characteristics they produce. Even after manual normalisation of the individual RGB channels using pixel math, I still couldn’t get the right balance with a 1:1:1 ratio (as per filter specs.) so opted for a more suitable balance. Kept a natural feel. The exact problem occurred with the Helix nebula image. The RGB data used 10min subs binned 2x2 to give a good S/N ratio. The RGB data is actually quite good, but on reflection, I should have also taken some shorter 5min subs to bring the star colours back. 10min subs was ideal for the galaxy chrominance, but pushed the stars too far (thus lost their colour information). Overall, I feel I’ve captured the “essence” of the galaxy reasonably well with handful of background galaxies dotting the scene. NGC 253 is quite high late in the evening this time of the season so lends itself well to getting good resolution as it crosses the meridian – darn those GEM flips!

Image processing;
All subs calibrated/reduced (dark, bias, flat), registered and Sigma-Reject combined in MaximDL. Luminance deconvolution performed in CCDSharp – two iterations. Colour combined chrominance (RGB) in MaximDL to achieve corrective colour balance ratios. Luminance initially stretched in MaximDL using DDP – no sharpening filters applied. Both images loaded into PS for further processing. Chrominance image stretched using shadow/highlights tool. Both images (luminance and chrominance) had gradients, but nothing considerably hard to deal with - created basic subtraction masks to remove these. Then moved on to cleaning both luminance and chrominance layers with the self-healing brush – removed colour inconsistencies and minor dead pixels that were not clean up in the combine function. Applied two selective layer masks, one to reduce noise in the dim areas and the other to highlight features of interest. Actually, I was rather slack with the noise reduction. Could have been a pushed the settings hard to better correct some colour noise.

I will try to make sure there is not such a long delay between posts. Though, I’m happy to produce a few quality images a year, than pump out a crap one every fortnight!

I hear the FSQ calling - back to wide fields is in order.
Well, that’s my couple of cents worth <pun intended>;), thanks for looking and hope you enjoy it.

Cheers

Dennis
03-10-2007, 01:09 PM
That is one of the best images I have seen of NGC253. In fact, I cannot recollect ever seeing the dust lanes so well defined, with so little noise, especially around the fainter edges of the arms.

A stunning portrait of one of my favourite galaxies – a fantastic effort in acquiring, processing and presenting this cosmic masterpiece. I am inspired and now have something to set my sights on.

Cheers

Dennis

EzyStyles
03-10-2007, 02:53 PM
:eyepop: jase, so crisp and detailed! very very nice could even see some of the nebs around the dust lanes.

The core might be abit too bright.

RB
03-10-2007, 03:06 PM
No wonder the $AU is up, look at the quality of this coin !!!

Well done Jase !!!

:eyepop:

h0ughy
03-10-2007, 03:26 PM
LOL more than a doller spent on this one - as the others said - well done

leon
03-10-2007, 03:32 PM
They may be few and far between Jase, but this is certainly worth waiting for, what a ripper of a shot. :eyepop:

Leon :thumbsup:

jase
03-10-2007, 03:47 PM
Thanks Dennis. Pleased you liked it. It is certainly a real gem to image, but rather difficult to get the spiral structure right. Thanks again.



Cheers Eric. Indeed, the core maybe a little on the bright side. If you stare directly at the core, your peripheral vision will notice the two barred spiral arms. My goal was to bring these out as I have not seen many images portray this effectively. I found a dim core lost this effect.



Thanks RB. Will not take credit for the AU economy though.;)



Cheers H0ughy. Appreciate the kind words.



Pleased you liked it Leon. Will try to step up my output. Will never compete against the volume of you DSLR guys put out. But then again, quality over quantity! Thanks again.:thumbsup:

Comet Hunter
03-10-2007, 05:52 PM
Another ripper Jase. Good to see you back at it.:)

BluMoon
03-10-2007, 06:06 PM
:party2:sorry its taken me this long to post a comment but I was busy bowing in front of my LT!!! well done great effort.

jase
03-10-2007, 06:19 PM
Cheers Andrew. Appreciate the kind words. Hope you get up and imaging soon mate.:thumbsup:



Thanks Blu Moon.



As I listen to all critiques, I pondered on this further Eric. Having three iterations of this image in different forms, all have a similar core brightness levels. Attached is a "unprocessed" luminance. Unprocessed in that it has gone through registration, hot-pixel removal and sigma-reject then stretched using DDP to present the expected dynamic range. No other processing was performed. When looking at the final image posted and this image, the core has nominal values and as such I don't believe I've lost much (if any) data due to this (white clipping). I may reprocess a few more renditions, but for now I'm satisfied with the presented image. Thanks again.

Garyh
03-10-2007, 07:10 PM
Beautiful Jase! :eyepop::eyepop::eyepop:
I have had 2 looks at it so far today and its mind blowing!!
Makes me want to use it for my desktop so I can study it some more!
Absolutely well done!!:thumbsup:
look forwards to your next post!!
cheers Gary

jase
03-10-2007, 08:17 PM
Thanks for your comments Gary. Much appreciated.:)

Phil
03-10-2007, 08:25 PM
nice one Jase like all your shot stunning. Well done
Phil

iceman
03-10-2007, 08:43 PM
Wow, how am I going to post mine now!?!? Brilliant!

Bassnut
03-10-2007, 08:50 PM
Excellent Jase. The mark of a good NGC253 is how well defined the dust lanes are, I have found them difficult to resolve. You have certainly done them justice, and the colour balance is superb, also difficult to get right.

most enjoyable

Lee
03-10-2007, 10:02 PM
Exactly what I was thinking Mike! :shrug:

Jase - can you please post a list of what you intend to point your scope at over the next 12 months - I won't bother imaging them then!

Really good shot too......

Ric
03-10-2007, 10:53 PM
A wonderful image Jase, so rich and detailed.

You have done a great job with this image.

Cheers

Tamtarn
03-10-2007, 11:16 PM
Another beautiful image Jase The full resolution is AMAZING

DavidU
03-10-2007, 11:36 PM
Awesome ! great detail......:astron:

jase
04-10-2007, 05:50 AM
Cheers Phil.:) Its pleasing when tasks come together. I had my doubts about the resolution of the luminance, but after a little deconvolution it worked out well. Thanks again.



Thanks Mike. :thumbsup: Please post your rendition. Don't be deterred by my image. All images are special in their own right.



Cheers Fred. :D I focused on the dust lanes in the processing as I tried to emphasis the barred spiral structure. As can be seen comparing the NIR and visible wavelength images, its not easy to achieve (at visible wavelengths). The final colour balance I used was 1:0.7:1.3. This follows the characteristics/sensitivity of the KAI-11000M chip. Seems to have worked quite well. Thanks again.



Thanks Lee. Believe it or not, but I have completed my acquisition listing for the next three months. Sounds like a long time, but its only a handful of images as it takes time to deliver an image of good quality (especially with a mono camera and filters). I've worked out how the objects will be framed and in some cases a mosaic maybe required. Certainly not releasing the list - that would spoil the surprise. ;)



Cheers Ric. I'm pleased you enjoyed it.:thumbsup:



Thanks Barb and David. Appreciate the kind words. :)



David, thanks for your comments. :)

=========

Once again, thank you all for taking the time to look at this posting and make comment. Appreciate it.:)

strongmanmike
04-10-2007, 07:36 AM
A mighty fine effort there Jase!

The galaxy itself has a nice full body feel to it and yep there is no substitute for focal length when it comes to imaging galaxies.

I assume that was Brad's scope again?

There is always that tossup between revealing the core structure and producing a "flatter" looking disk and keeping the core more bulbous and burned out. Both approaches have their proponents and each to their own as long as the core revealing doesn't look too unnatural or the burnout too strong and saturated.

I am sure you must be annoyed by the gradients and colour spekles remaining but the galaxy itself is really quite good. With such a small field and having imaged with no moon what do you think caused the gradients? Are the spekles from dithering problems again? I remember you had the same problem with the Helix shot done on Brad's beast. Are you noticing that after dark subtraction there are still hot pixels remaining that don't correspond between subs so when conbined you end up with coloured pixels all over the image?

My ProLine had this issue initially and thus dithering the subs and median combining them was the only way around it. A reload of new firmware in the camera, courtesy of FLI, has fixed the problem though and now I don't need to dither guide anymore and simply summing or "adding" the undithered subs shows very few residual rougue unmatched pixels remaining, After dark subtraction, it's now almost as good as my SXV-H9 with its silky smooth Sony chip :). Aparently my KAI11K chip had been incorrectly "clocked" and dark current was leaking out over time producing variations in hot pixel maps from sub to sub? All cured now though.

Very nice work again :thumbsup:

Mike

jase
04-10-2007, 08:43 AM
Hey Mike, thanks!
Yes, acquired using Brad’s scope. Remote imaging – got to love it. It opens up many possibilities. Just compile the imaging plan, upload it, and then get some sleep. Planning and image processing is the fun part. IMO, acquiring the data is rather mundane/routine.

Indeed, there is always a trade off between highlighting the spiral structure and not burning out the core. DDP does a reasonable job, but I usually don't stretch to final levels using DDP. I’ll only stretch to say 60%, and then finish off with curves in photoshop. This makes it easier to manage the colours in a phased approach. I typically go with a core size similar to that of the largest star in the frame for aesthetics; however there are some exceptions with this if the galaxy core is quite diffused.

I got tired of imaging cleaning. I really could have chased them further to remove more speckles. Brad has recently redone the calibration frames which I’ve noticed a significant improvement, but still not perfect. Plenty of hot/dead pixels present making for some heavy work. RC Console cleaned quite a few with a 20% threshold, may have cleaned it better with a different value. I left ACP dithering as default – didn’t tweak it further in fear of losing the guide star on the tiny ST237 chip. I’m getting use to ACP’s style of imaging as I’m also using it for remote operation in my own observatory. A couple of sub frames weren’t guided correctly, but I included them in the combine routine. After blinking the images straight after registration, the error appeared random so the combine routine removed the outlier pixels.

Not entirely sure about the gradient source. The RC12.5 of Brad’s doesn’t have the “shorty” light shroud around the primary (this may have changed). I’m also not aware of local light pollution sources at the remote observatory site (in SA) – Fred will know. Could be a combination of things.

Interesting to hear about the dark current leaks. This would have a detrimental effect. Regardless of the scope/camera combination, I always dither – even if using a class 1 chip. I don’t know of a more effective way to remove a dead column unless of course you want to build a pixel map of the chip. The downside of using a map is its static and may not represent the current pixel correctness. i.e. some pixels die over time. I'm open to suggestions.

Thanks for your comments.:)

EDIT: I should also add, that one night, the weather was quite warm at the remote location and the chip couldn't be cooled to the set point of -15. This did not help the situation in controlling noise! Perhaps, need to persuade Brad to get an FLI.:whistle:

Dr Nick
04-10-2007, 11:27 AM
Wow, superb image! ;)

tornado33
04-10-2007, 03:52 PM
Gee the big chip camera and telescope sure bring out the detail.
Scott

jase
05-10-2007, 06:48 AM
Cheers Dr. Nick. Pleased you liked it.



Thanks Scott. Yes, the two components help. Though, image processing played the key part.