My first Galaxy, M83 and friends

[Crushellon (Tim)]

My first attempt at a galaxy. Very happy with the results, especially as it was done with a dobsonian. I'm very surprised at the level of detail I was able to achieve with such short exposures and the galaxies in the background are pretty impressive too. If I can get these results with hardware that isn't even designed for imaging, then I cant wait to see the results that I will get when my Neq6 gets here on Monday it's just too bad I'm going away for 5 weeks on Thursday... typical

Canon 550D
SW Synscan 10" Dob
97 x 10' Lights
30 x Darks
30 x Bias

Hi Rez: http://www.astrobin.com/full/247897/

[vlazg (George)]

Tim, that is very, very good.

[Atmos (Colin)]

That has turned out really well with 10s subs on a dob! When your NEQ6 arrives, what are you going to be imaging with?

[Crushellon (Tim)]

Originally Posted by Atmos

That has turned out really well with 10s subs on a dob! When your NEQ6 arrives, what are you going to be imaging with?

I'll just be starting off with my camera lenses while I learn polar alignment etc. But after I've got it down I think something along the lines of a 8 or 10" reflector. Which size depends on what the mount can handle reliably, from what I've read about half of the capacity which leaves around 10kgs I believe. The hardest part will be deciding whether to go a longer F5 or a shorter F4. I'm thinking the longer length and cooling the 550D, with a homemade guider to come.

[glend (Glen)]

Good effort considering the equipment your using. Coma of course is evident in the sides. If your going to go to an EQ mounted newt then invest in a good coma corrector, the Baader MPCC is ideal for newt photos.

Re the scope size choice, try to limit the scope net weight to around 15kg on a NEQ6. I have the same mount, albeit on a pier, but imaging capacity should be kept under 18kg max despite the stated visual payload capacity. This gives you room for your camera and a guidescope setup ( which you will need at some point). A 10" production newt with a steel tube might be a little to heavy with all the extras but i am sure many people are using them. I have a 10" f5 newt i use on mine but it is a custom lightweight build using carbon fibre struts and weighs less than 15kg. Just ask about the OTA net weight before you purchase. A carbon fibre tube newt might be a good choice to reduce weight and maximise aperture but they are more expensive.

[Somnium (Aidan)]

simply fantastic for a) a first attempt and b) the equipment you are using. keep it up!

[Atmos (Colin)]

If you don't mind the tighter tolerances and the bit of extra fiddling that goes into get a near perfectly running system, go the F/4. Being that little bit shorter it may also be marginally lighter. Go full carbon fibre if you can, with the EQ6 you want to limit your weight as much as possible with a 10" system. With an 8" it won't be too important as it is significantly lights to being with.

[raymo]

It will handle a 10" f/4 or f/5 easily, with about 3.5 and 3kg respectively,
available for imaging gear.
raymo

[RickS (Rick)]

Nice result, Tim!

[Mickoid (Michael)]

Can't believe you managed that result with 10 second subs. I have always been puzzled by how short exposures producing a low signal response can accumulate as much detail with heaps of subs as someone shooting 5 min subs with far less of them.
With accurate polar alignment being a key issue in obtaining quality images with round stars, why would you not shoot more shorter subs to reduce alignment errors and thus still obtain a great quality image such as you've done here.
What is the advantage of shooting longer subs? Saving space on your memory card perhaps! If there is no advantage then do your 10 sec X 97 subs equal 4 min X 4 subs?

[Atmos (Colin)]

Originally Posted by Mickoid

Can't believe you managed that result with 10 second subs. I have always been puzzled by how short exposures producing a low signal response can accumulate as much detail with heaps of subs as someone shooting 5 min subs with far less of them.
With accurate polar alignment being a key issue in obtaining quality images with round stars, why would you not shoot more shorter subs to reduce alignment errors and thus still obtain a great quality image such as you've done here.
What is the advantage of shooting longer subs? Saving space on your memory card perhaps! If there is no advantage then do your 10 sec X 97 subs equal 4 min X 4 subs?

The main reason for going with longer subs is for the fainter stuff. Short exposures with high ISO is okay for brighter areas but fainter regions will remain buried under sensor electrical noise.

[Mickoid (Michael)]

Colin, Tim has managed to pick up faint background galaxies in his shot as well, with short subs.

[raymo]

I'd like to know how this image was achieved with 10 sec subs.
My best M83 with my 8" f/5 was 52x 50 secs @ ISO 1600, and 10x 50secs
@ ISO 3200,at prime focus, which normally gives superior results to afocal, and it shows a little less faint stuff than this image does.
raymo

[Crushellon (Tim)]

This was shot at prime focus.

I'll post a comprehensive list of the processes I went through to capture and process tonight when I get home and I'll host some files of subs and stacked and background extraction images. I dare say that APT and pixinsight helped a lot in getting a good result.

[raymo]

For some reason I was thinking that this and the adjacent Mars image
were by the same person, and the Mars image was afocal.
raymo

[rustigsmed (Russell)]

nice work tim,
you should be able to push the subs out to 20 seconds no worries on the dob. although now with the NEQ6 you won't need to!

good work, goto dobs can do some good stuff on the brighter deep sky objects.

cheers

russ

[Crushellon (Tim)]

So for the sake of possibly helping anyone get better results using a dobsonian or similarly not optimal setup I am going to go through exactly what I did to get this image result.

tools used:
Skywatcher synscan 10" dobsonian
Canon 550D
Basic laptop
Apt Astrophotography tool
Pilades Pixinsight

Firstly I picked an object that is in a part of the sky that won't rotate very much. I decided on M83. I was also inspired by the image of another user on IIS not to long ago, jcimera who got a very similar result as this using 20sec subs with a 12"goto dob. thread: http://www.iceinspace.com.au/forum/s...d.php?t=144869

Setup on the dob wasn't overly accurate but the result was good enough that M83 came up in the field of view straight away with a goto command.

Using APT Astrophotography tool, the object was framed and I started at testing for optimal capture settings. I was hoping for 20 sec subs @3200iso but was getting oval stars so getting shorter and shorter while testing I decided on 10 sec subs. I also decided on iso3200 to try and get more detail (I have always used iso1600 so far on deep sky, but they have all been bright emission nebulae or star fields)

Focus was done by eye using live view in APT, took a bit of fiddling with the single speed crayford but I was happy enough with my setting so I moved on.

I set a plan in APT using the settings of: 3200iso, 3sec anti vibration pause, 10 sec capture, 5 sec pause (ended up at about 10 seconds with the "busy" time after the capture), with a count of 40 so that I could check focus at that interval.

APT creates and shows a thumbnail of each sub after it is captured so at about every 10 images I was adjusting the framing of the object during the pause portion of the plan (I had the speed of the hand controller set at 4 and was entirely guessing at how long I would move it for, after doing it for a while you get a fairly good idea of how far the scope will move if you hold the button for 2 seconds)

Checked focus every 40 images.

I ended up capturing about 160 subs. I started imaging just after M83 passed zenith, but as it was lowering more and more in the sky my subs became more bad than good. By the end of the session I started only getting one good image out of four where I was mostly getting good stuff when I started. So I called it there.

A single sub (untouched original CR2): https://onedrive.live.com/redir?resi...nt=photo%2cCR2

All of my darks were captured after the imaging session. Sensor temps had gone from 27deg at the start down to 17deg when I finished. So a little trick I use to simulate the temps of earlier in the night is to put the camera on live view for about 5 mins before I start capturing darks. Then when I start capturing, the darks will start off at a higher temperature and gradually go down to the temps that I ended up with at the end of my imaging session. Obviously this trick will only really work if you're using short subs because with long subs, using live view for 5mins is no different than what your doing to capture the dark frame in the first place, rendering the exercise a useless gesture)

It was my first time using 3200iso so I also captured 30 bias frames at 3200iso. It ended up being easier to just do this without APT as it took time for the software to process each image whereas, with the custom timer function on the 550D the whole process of capturing the bias frames took all of a minute or two.

Visually inspecting the images on my computer afterwards, I binned anything in which the stars where not circular in the center of the image. Even if it was slightly oval, it was gone. Being that imaging galaxies was the thing that interested me most and is the reason why I started doing astrophotography in the first place, I wasn't going to use any subs that might lower the quality of my first galaxy image.

All of the processing was done in Pixinsight. I am just going to list the steps I used to processed the image and I will link references at the end to the (main) guides and tutorials that I base all of my processing on.

Convert dark and bias frames from CR2 to xisf so that the files are compatible with multi layer integration.

Bias integration.

Dark integration. I don't calibrate the darks individually to the bias frames, after some testing I found no difference between calibrating them individually or simply clicking the calibrate button on the darks tab when calibrating the subs.

Calibrate subs with dark and bias stacks.

Debayer.

Register and generate drizzle data.

Integrate subs (drizzle data must also be entered in this process)

Drizzle integration. (scale 2)

And there you have your stacked and calibrated image.

Cropped low signal areas.

Dynamic background extraction (division)

Dynamic background extraction (subtraction)

***I attribute a lot of my final result to this very powerful tool, below I have linked two images, one of my calibrated and stacked image and one that has had the low signal areas cropped and DBE applied. The results speak for themselves.. (Linked images have had an autostretch permanently applied and were saved as JPG as the actual file sizes were 500mb and 800mb)

Calibrated and stacked image: https://onedrive.live.com/redir?resi...nt=photo%2cjpg

After low signal crop and Dynamic Background Extraction: https://onedrive.live.com/redir?resi...nt=photo%2cjpg
Background neutralization

Color calibration

Build various masks (****see M42 pixinsight tutorial at bottom of post)

Attempted deconvolution (I still havn't succeeded in this and I ended up skipping it after a dozen attempts)

TGV denoise RGB/K mode

Multiscale Median Transform to luminance x2, then to chrominance, luminance again and chrominance again

SCNR to green

Histogram transformation

More SCNR to green

HDR Multiscale Transform (very good tool for bringing down the bright ares while bringing out the low signal areas, it worked extremely well with this image)

Local histogram transformation x2

Curves to stars, less color.

Unsharp Mask on galaxy using a range mask. (brought out good detail in the galaxy but also made a fair amount of noise which you can see in the final image. I am getting fairly competent at removing background noise but as for noise in the bright parts, I still have no idea.)

Curves to background

Curves to galaxy

Final Image.



I have been doing astrophotography for about two months and have been able to achieve great results with budget gear and limited knowledge. The internet is truly an amazing thing. None of this would be possible without all of the hard work of all the people that have come before me. There was nothing special about the way that I captured these images, they were done simply using the basic methods of astrophotography. My results (I believe) are entirely the result of powerful software in pixinsight and the very hard work of the people who have made great tutorials for using this amazing software. Below I have linked (most) of the tutorials which form the basis of my processing in pixinsight, and I believe the credit for my work goes to them.

Image calibration & integration:
https://www.youtube.com/watch?v=zU5jJgjKuQQ
http://www.lightvortexastronomy.com/...ixinsight.html


Dynamic background extraction:
https://www.youtube.com/watch?v=UjUrEXqvNtU

Image processing:
https://www.youtube.com/watch?v=6HUOe4R5Hng
****M42 pixinsight tutorial http://trappedphotons.com/blog/?p=642
https://pixinsight.com/forum/index.p...0&topic=5283.0