A Cosmic Collision in Corvus - (The Antennae Galaxies)
This was my major project for the last new moon, and with quite a few clear nights in a row, I decided to focus the whole time on the Antennae Galaxies - NGC 4038 & NGC 4039.
This is a fascinating pair of galaxies involved in an epic slow-motion collision. The 2 extensive tails of dust & stars ejected from their host galaxies resemble antennae, but there's also some fiery activity where the galaxies entangle and entwine causing lots of rapid star formation. This interaction has been well-studied as it may well resemble the fate of our own Milky Way when it collides with the Andromeda Galaxy.
Captured from my B7 backyard with a tad over 30hrs total integration (culled from around 35hrs total in 60sec subs). There were a few nights of excellent seeing which have helped reveal some pleasing detail. But even with the hefty integration, it was pretty hard to get those faint extents from my location, and a myriad of background galaxies are only just peeking through...
Stacking, LP removal in APP, processing in PI and final tweaks in LightRoom.
Love the full field mate. What a labour - 35 hours of 60 second subs. Wow. B7 skies certainly make the job harder than it needs to be. You've done a great job teasing the detail out of that background sky glow. How did you go dealing with the gradient during post processing?
I can see what looks to be small star forming regions in the outskirts at the very end of both antannae, and you've just managed to bring in quite a few tiny galaxies in the field too, which makes the image even more interesting IMO. Not an easy task with B7 skies. A great photo of a super interesting object.
Dave,
That’s a beautiful image with excellent resolution and detail under less than ideal skies ( even during New Moon period with short subs there’s still loads of gradient to deal with under Sydney’s light dome )
Gee a mammoth project indeed around 2000 x 60 sec exposures
To expose those tails like you’ve done is testament to your skill and experience in Galaxy imaging.
A great result
Well done !!!
Hi Dave,
excellent result -
you've done well and you've worked hard for your meal -
1800 frames stacked with Astro Pixel Processor.
I bet you could hardly see anything with a single frame?
In theory you have reduced the noise by a factor of 42 -
so you were able to capture a mag 11 target under city lights.
Love the full field mate. What a labour - 35 hours of 60 second subs. Wow. B7 skies certainly make the job harder than it needs to be. You've done a great job teasing the detail out of that background sky glow. How did you go dealing with the gradient during post processing?
I can see what looks to be small star forming regions in the outskirts at the very end of both antannae, and you've just managed to bring in quite a few tiny galaxies in the field too, which makes the image even more interesting IMO. Not an easy task with B7 skies. A great photo of a super interesting object.
Cheers Dean. Yeah B7 makes you work pretty hard but for a target like this that's so well positioned it's possible to get good hours in a night - and we were particularly blessed with a number of clear nights this new moon!
Gradients were actually pretty tricky on this one - but APP has a fantastic gradient tool, and then I did another run with the new PI gradient tool which cleaned up a few pesky spots. Glad you liked the result!
Dave,
That’s a beautiful image with excellent resolution and detail under less than ideal skies ( even during New Moon period with short subs there’s still loads of gradient to deal with under Sydney’s light dome )
Gee a mammoth project indeed around 2000 x 60 sec exposures
To expose those tails like you’ve done is testament to your skill and experience in Galaxy imaging.
A great result
Well done !!!
Martin
Thanks Martin - I was really pleased with the res in the core and can really put this down to a few of those nights with really excellent seeing. I just went back to my data sets and some of the best subs were coming in at FWHM 1.5-1.6"! Thats pretty much as good as it gets here in Padstow. Not sure if you've noticed too, but on the odd clear nights we've had here in Sydney the seeing really has been excellent.
The tails are a different matter with all the LP. I really had to push the data hard to get them to come out. Luckily theres not too much discernable detail in them so could counter with some aggressive noise reduction.
Beautiful image, David. A lot of detail to behold, and fantastic colours as well.
Your poor computer - stacking and registering that many frames!
V.
Lol thanks mate - this is the first time I FINALLY got SPCC working properly in PI and wow it made a great difference. I compared to a Hubble image and the colours were almost identical so really happy with that and much less stuffing around trying to manually adjust things..
Yeah the laptop's been working pretty hard but probably not as epic as you'd assume... My trick is that I actually treat each night as 2 sessions - one for the best data around the meridian and another set for the rest. I stack each set individually with grading to suit the quality of the data... so for this project that's 12 stacks. I then stack the stacks together for a final image. Not only does this make the task MUCH easier and less time consuming, but APP weights every sub and prioritises the best data for the final stack. So in effect, the stacks from poorer data contribute less to the final product. End result is better on all fronts I believe.
Yeah Dave, another lovely bit of composing, we'll have to start calling you Mozart Can't add anything on what Martin said, top job
Great that such shots can be achieved from a big city too
Mike
Cheers Mike I appreciate your comments! I was really trying hard to reveal as much of the tails as you did in your image last year but will have to concede defeat on that front - oh for some dark sky!
Hi Dave,
excellent result -
you've done well and you've worked hard for your meal -
1800 frames stacked with Astro Pixel Processor.
I bet you could hardly see anything with a single frame?
In theory you have reduced the noise by a factor of 42 -
so you were able to capture a mag 11 target under city lights.
cheers
Allan
Thanks Allan yep certainly a bit of work involved but really happy with the result. Single subs did show a fuzzy blob, but yeah I think to get any fine detail through the noise from my location I've gotta go north of 20hrs.
I mentioned to Steve below, I've got a method to handle projects with lots and lots of data which involves splitting my sessions into smaller data sets and stacking each individual set first, before a final stack of all the set together for processing. APP weights every sub and prioritises the best data for the final stack. I think that the result is cleaner and better resolution, even though it may not be quite as deep, as effectively I'm allowing APP to reduce the contribution from poorer data sets.
Thanks Allan yep certainly a bit of work involved but really happy with the result. Single subs did show a fuzzy blob, but yeah I think to get any fine detail through the noise from my location I've gotta go north of 20hrs.
I mentioned to Steve below, I've got a method to handle projects with lots and lots of data which involves splitting my sessions into smaller data sets and stacking each individual set first, before a final stack of all the set together for processing. APP weights every sub and prioritises the best data for the final stack. I think that the result is cleaner and better resolution, even though it may not be quite as deep, as effectively I'm allowing APP to reduce the contribution from poorer data sets.
Is it possible to load all 1800 sub frames into one stack and check
to see if the result is better?
cheers
Allan
Hi Allan,
I believe you are correct in that it is going to be a bit of a compromise. But the reality is that I just don't have the computing power to process that much data in one go. Just for interests sake, I loaded all the subs (1989 in total after manual removal of obvious duds) and let it give me the estimated stats - it needed 256GB for processing (way more than i have to play with) and from experience, would take an enormous amount of time. And thats binned 2x2.
When I stack sets, APP actually recognises when you load 'lights' that have already been integrated and deals with those accordingly. From my experience the only real problem I've encountered is if you've got some sets with differing amounts of integration. The problem there is that the stacking program will prejudice against good data just because its in a smaller set. But this is easy to solve by ensuring all sets are roughly the same size.
All in all, I'm happy with the compromise as the testing I've done on smaller projects seem to show very little difference - but yes I will accept that with the right computer running the whole set in one go would theoretically provide a better result.
Thanks for the very interesting link. Although judging by the responses the juries probably still out - with those with APP experience saying it definitely works, and some others with valid concerns...
I believe you are correct in that it is going to be a bit of a compromise. But the reality is that I just don't have the computing power to process that much data in one go. Just for interests sake, I loaded all the subs (1989 in total after manual removal of obvious duds) and let it give me the estimated stats - it needed 256GB for processing (way more than i have to play with) and from experience, would take an enormous amount of time. And thats binned 2x2.
When I stack sets, APP actually recognises when you load 'lights' that have already been integrated and deals with those accordingly. From my experience the only real problem I've encountered is if you've got some sets with differing amounts of integration. The problem there is that the stacking program will prejudice against good data just because its in a smaller set. But this is easy to solve by ensuring all sets are roughly the same size.
All in all, I'm happy with the compromise as the testing I've done on smaller projects seem to show very little difference - but yes I will accept that with the right computer running the whole set in one go would theoretically provide a better result.
Thanks for the very interesting link. Although judging by the responses the juries probably still out - with those with APP experience saying it definitely works, and some others with valid concerns...
Thanks Dave,
256GB of RAM - that's incredible.
Most desktops don't even have room to fit
that much on their Motherboards.
I have 32 GB - I thought that was a lot.
Yes - it seems that the stacking methods are debated strongly online.
One thing that amazes me on here is seeing images that, as a kid, I saw as black & white photos in books- little blobs without much internal detail, that got their names from the outside shape.
Dave, just amazing. I can appreciate how tricky getting this target to look sharp is. I've currently got the Antannae in the pipeline myself and am struggling to get nice detail the way you have! Beautiful colour balance too. Very well done indeed.
Neville I agree it is quite amazing how astrophotography has evolved over the years. For me as a relatively newcomer to the hobby I’m still learning lots and try to make some improvements to my process each project. Always more to do but happy how it’s going.
Stephane i really look forward to seeing your efforts and judging by your recent works I’m sure it’ll come out great! A bit of aperture does help with the detail tho lol
Can't add anything on what Martin said, top job 
