What I Don't Know About Image Processing...

... can fill a warehouse.

I revisited my earlier M42, M3 and M51 image sets to give them another go around in software.

As it turns out, there's a lot more detail in there which I simply don't know how to extract.

These images are the results of my experiments.  I included both the old and new for comparison.

M42

First pass

Second pass

M3

First pass

Second pass

M51

First pass

Second pass

For each of them, more noise is apparent.  This isn't surprising giving how the image data is being stretched.  These image sets have bias frames applied against them, but no flat frames.

In each of them, I had to brush out dust mote imprints and subtract vignetting in Photoshop as a postprocessing step.  I'd like to try minimize this by pre-processing the raw frames with flat field images. I haven't taken any of those yet.

The two biggest changes I made in processing these images was to:

1. Not overdo the levels stretching.  Watch that histogram closely.  It's more important to get a nice bell-ish curve going (depending on the image) than to try make your background deep and your stars pop. The moment any "blackening" or "popping" occurs, you've probably lost data.
2. Curves!!!  Holy smokes, this made a big difference!  Being able to recognize what you want to draw out, and then manipulate the curves to reflect it, has had the biggest impact.  The same rules with stretching the levels applies.  Avoid any clipping (if reasonable) so cores aren't blown out and faint stars aren't lost.

M3 and M51 with some new equipment!

Tonight was the first night I got to actually use my new Canon 1100D/T3 and SSAG!

The night before was filled with trial and error as I practiced setting up the C6-N with the new equipment.  I had to figure out issues like balance and alignment with and without the camera and autoguider. These two gadgets added about three or four pounds to the mount which really needs consideration.

I also had new software options to figure out now that my new camera is actually well supported by Nebulosity!

Here's what I learned during the prior evening of tribulation:

(Please comment if you know of better/alternate ways.)

Balancing

This started out as a point of confusion for me.  I didn't want to balance the scope without the camera and autoguider, but I also didn't want to align the mount/scope with the camera and autoguider.  I figured I had to try compensate for the weight difference between the two so I could balance and align the scope as though the camera and autoguider were attached.

Luckily, my autoguider's guidescope is the same 9x50 finderscope as the one used on my Orion XT10i.  The difference between the two scopes is that one has the autoguider hanging off the end while the other has the correct-image right-angle eyepiece thing.  So, instead of using the C6-N's lightweight 6x30 finderscope to balance and align with, I used the finderscope from the my XT10i.  The weight's weren't exact, but they were close (enough?).

As a side note for the C6, I generally find that the straight-through 6x30 finderscope is much easier to use than the "correct image" right-angle finderscope.  Being able to use both eyes to direct a star into the middle of the cross hairs makes things go much faster, especially if it's later in the evening and there are more stars out (making it more difficult to distinguish between the alignment star and the others).  For that matter, I found my Meade ETX125's reflex/red dot finderscope to be the easiest of them all.  The first time I took my XT10i out to the relatively dark skies of the Catskills in New York, I had a really difficult time aligning the scope.  I couldn't tell the alignment star from all the rest that suddenly appeared in the finderscope.  From suburban Philly I'd never had that problem before.  (My green laser pointer really helped me here).

I then looked at how to match the DSLR's weight in the eye piece.  My heaviest eyepiece, a 1.25" 32MM isn't the greatest to use during alignment, and I preferred to use my 20mm illuminated centering eyepiece from Orion (ICE).  Neither of them matched the weight of the camera.  So, I rigged a crude weighty necklace made from washers and zip ties to drap around the ICE to bring it up to scale.

I balanced the scope on the CG-5, and was good to go!

Alignment

After I managed to balance the scope as though it had the imaging equipment on it, alignment was pretty ordinary except I spent a bit more time doing the All-Star Polar Align routine.  Autoguider or not, a bad polar alignment will give you drift/field rotation in your exposures.

I'd still like to pick up a polar axis finderscope to see how well the polar alignment routine works with the CG-5 (or to make it easier), but it seems to work pretty darn well!  Turning the azimuth knobs gets difficult, though, the further you have to push the mount over.

I still find it fascinating to watch the mount get progressively more accurate during the addition of calibration stars.  By the time I get to the third or fourth star it, more often than not, gets dumped virtually in the center of ICE.

When I eventually switch out the eyepiece and finderscope for the camera and autoguider, it's nice to know I won't have to hunt for the target.

Software Test Run

After I had spent roughly 3 hours getting all this done, I eventually got to sit down with Nebulosity and the camera.  I was looking forward to trying Nebulosity's frame and focusing routines.  Apart from learning the various quirks of the software and it's interface, it's all relatively straight forward.  I did discover the following points though:

• Live view on the DSLR (pulled into Nebulosity) is very useful for framing.  Not as much for focusing.  I think I should give this another try though.  I take a ton of frames during this procedure, and maybe live view would help extend the mechanical life of the flip mirror enough that I should be using it more.
• Nebulosity's fine focus routine is where the live view shines.  The instant feedback is tremendous.  However, I still use my bahtinov mask during this process.  I've come to heavily rely on observing the diffraction spikes and it hasn't failed me (yet).

After I had framed and focused on my target DSO, I engaged PHD.  Sadly, my 10 year old Celeron-based Compaq Presario's battery threw in the towel.  A few seconds after PHD started calibrating, everything went dark.  Frustrating, but not entirely unexpected.  I wasted a LOT of time figuring out the issues, and the Compaq's battery only lasts about 2 hours.

The Imaging Session

Tonight, the set up went much better thanks to my experiences from the previous night.  At the same time, I used my mid-2009 Macbook instead of the 10 year old Compaq.  After 3.5 hours of imaging with it, it still had 60% battery left.  Plus it's a vastly more capable machine, power-wise.  It didn't stutter or stall at any time during the evening.

My first target of the evening was M3.  I started with ten 25 second exposures at 800 ISO just to get the hang of the camera control in Nebulosity with the guiding of PHD.

The seeing was apparently not great as PHD would occasionally warn me that the guide star's mass had changed.  Luckily, it always appeared to stabilize a second or so later and PHD never actually lost its lock on the star.

The one thing I learned from this?  Make sure you take off the bahtinov mask before you start imaging.

Next time, I'd like to try image M3 with a greater exposure time to try resolve more of those stars.

Messier 3 (Globular cluster in the north constellation of Canes Venatici)

My second target was M81 and M82 together.  For some reason, PHD had a lot of trouble with the telescope in this orientation.  It was unable to detect movement in the right ascension.  Switching the guide star didn't help at all.

I gave up on Bode and the Cigar after taking a few unguided 2 minute exposures.  I didn't want to waste my time taking shots that would probably be unusable anyway.

I then shifted downwards to M51, the Whirlpool galaxy.  After framing it through Neb, and re-calibrating PHD, I started a run of 10 exposures at 5 minutes each at ISO 800.

Messier 51 (Whirlpool galaxy in the northern constellation of Canes Venatici)

A few learned facts from this set:

• Doggone skyglow.  I've gotta get me one of these clip filters.  I've heard good things about them, and heck, it's my only real option with this equipment.
• Take more exposures than you think.  I lost half of my ten subs from a surprise breeze, satellites, a plane and a late night visit from from a neighborhood dog.
• Bring something else to occupy your time during a long shoot.  Apart from occasionally glancing at PHD to make sure everything is going okay, there's not much to do.

I decided to also try image M101 (Pinwheel galaxy), but it wound up being rushed and just a waste of time.  And overall, the biggest thing I learned:

• Choose one, maybe two, targets to image in a given night.  Bias frames are fast and easy to take, but if you want to also shoot flats, that'll add a great deal more time to the process.

The C6-NGT and Astrophotography Take 2

On this night I decided to try image two targets.  One that I could clearly see with the telescope and another that I couldn't.

This round, two things were different:

1. I spent the time ensuring I had a better alignment.  It took a great deal longer, but the results blew away my first shoddy attempts at taking pictures with this gear.  Instead of doing a simple two-star alignment with an additional calibration star, I did a two-star with several calibration stars AND I performed Celestron's All-Star Polar Alignment routine.
   
   Assuming the All-Star routine really works (which is evidently does), it's really amazing how far off my mount was in pointing to the north celestial pole.  I don't have a polar axis scope, and given other amateur astronomers feedback on the All-Star routine it looks like I may not really need it!
   
   On a good night, the alignment routine will take me 15 minutes.  This is a far cry from the 45min+ needed for a good drift alignment.  I also assume it's not as accurate as a carefully applied drift alignment.  Regardless, the All-Star routine gave me one minute exposures with virtually no trailing.  Maybe I was just lucky!
2. I bought a bahtinov mask from Farpoint Astro.  This took all the guesswork out of focusing and as far as I'm concerned, is a cheap no-brainer accessory to add to your imaging toolkit.  I couldn't imagine focusing without it.  I first tried to make my own, but I eventually decided that $18 bought me a piece of equipment vastly superior to what I could make.

The Easy Target

M92, a globular cluster in Hercules is nice and bright in the eyepiece.  You can easily see it in a short exposure and is therefore easy to locate and center.  You don't have to spend hours imaging and processing to discover that your target is being clipped accidentally.

M92 - Globular cluster in Hercules.  4 x 30"

The alignment is definitely better, with little to no trailing.  However, evidence of coma clearly shows around the edges of the image.  Darn newt.  Well, maybe when I get better at this I can look into a coma corrector.

The Hard Target (for me)

Caldwell 34 -- the Veil Nebula in Cygnus -- is a DSO I can't see from my home with my telescopes, although I have seen it in my 10" newt at a dark(er) site.

I only took one sub of the Western Veil (30 seconds at ISO800) and at first glance I could see nothing...

C34 - Veil Nebula in Cygnus

... but it's there!  This was my first face-to-face with a DSO that would need more subs, a longer exposure and heavier processing to draw out the details.

Here's a Photoshopped version of the same image to help draw out the nebula:

The LX200 and MallinCam Experiment - M3

My first foray into astrophotography began with a Meade LX200-ACF with a MallinCam Jr.  The LX200 is great catadioptric telescope that sits on a rock solid alt-az mount.  My alma mater made this equipment available for use to help promote astronomy at the school and in the community.

The MallinCam is a very sensitive CCD video camera that features frame integration, allowing a live view of deep space objects (within reach of this telescope) on an external display.  This is a terrific way to show groups of people what the telescope is looking at without having to form a queue at the eye piece.

This particular MallinCam is the entry-level version.  The frame integration is insufficient for capturing color so everything comes through as grayscale.  My very first photograph was of M3 the huge globular cluster in the northern constellation of Canes Venatici.

It was terrible:

Messier 3 (NGC 5272) - Globular Cluster

I took 2.5 minutes of footage and processed it using Registax.  This was an experience in and of itself.  The end result was a low resolution, coma-stricken, posterized shot of M3.

What an awful picture... but what a headrush to have captured something 199 quadrillion miles from earth!