How to align sub-frames recorded at low elevation angle?

[MLParkinson]

Here is an image of the Andromeda galaxy recorded at Wiruna, Central Tablelands, NSW. This spectacular object is always low on the horizon when observed from the latitude of the Central Tablelands (33 South).

In order to pull-off an LRGB image of M31, one must record the sub-frames as the object rises and sets through a large cross section of atmosphere which is highly refractive at low elevation angles. I recorded the red frames first, then the green, blue and then finally the luminosity frames when the object was close to the meridian passage – a standard strategy.

The second image shows a close-up of the top-right hand corner of the image which was located closest to the horizon. I use Nebulosity to align and stack the sub-frames. The transformations used in this program are all linear – shift, rotate and equal scaling in orthogonal directions. It is impossible to bring all the stars into alignment across the entire image without using non-linear transformations which compensate for the effects of strong atmospheric refraction.

What software do you use to align sub-frames and will it cope with this demanding situation? Does it employ non-linear transformations to align the sub-frames?

My hunch is that if the sub-frames can be scaled independently in orthogonal directions then I will succeed in aligning the sub-frames through several re-alignment cycles.

Murray

[RickS]

RegiStar would probably deal with this: http://www.aurigaimaging.com/

PixInsight StarAlignment also handles distortion and might be worth a try if you have it (or grab a trial copy): http://pixinsight.com/forum/index.php?topic=5800.0

[MLParkinson]

Thanks for the advice Rick. How well would MaxIm DL handle image distortion? MaxIm DL is "meant" to be the best. If I am going to buy a license for more software, then I want the best. Cheers.

[RickS]

Murray,

I use Maxim myself. It is a very good capture program but, in my opinion, adequate but not great for processing. The alignment algorithms only appear to deal with translation, rotation and linear scaling. There is a plugin that might do what you want: http://winfij.homeip.net/maximdl/undistort.html

RegiStar is pretty clunky to use but is very good at image registration which is just about all it does.

PixInsight can be challenging to learn and doesn't suit everybody but it is a very powerful image processing package.

There may be other options so perhaps others will chime in...

Cheers,
Rick.

[Ross G]

Still a nice looking M31 Murray.

Good luck with the alignment.

I photographed M31 from Wiruna in August and it was difficult enough with just an OSC ccd. You're lucky to get 3 hours of decent imaging time in a night.

Ross.

[MLParkinson]

Thanks again for more advice Rick. It really does help. I visited the PixInsight website and my impression was that the software is continuously under development by a team of very gifted Spainish mathematicians and software engineers. I am truly impressed by how clever some astro-imagers are. I will let you know what I eventually decide to do.

[IanP]

This issue can be fixed in PixInsight using DynamicStarAlignment.
Go for PixInsight, you will not regret it>
:hi:

[multiweb]

+1 for registar.

[Shiraz]

+1 for Pixinsight - it has a very flexible registration module which can accommodate variable distortion. http://pixinsight.com/doc/tools/StarAlignment/StarAlignment.html

[MLParkinson]

I couldn't figure out how to successfully install Registar under my version of Windows 7 Professional and I suspect the program needs a re-build using a modern compiler like Microsoft Visual Studio 2012, or something like that anyway. I'm not a software engineer so I am just guessing.

I am so impressed by the serious talent behind PixInsight, I have ordered a trial license and suspect that I will end up buying a full license. The knowledge gained will be priceless. The PixInsight guys are pros.

[MLParkinson]

Thank you very much to those of you who recommended I try PixInsight. The software is brilliant, one of the finest s/w packages I have ever used in my life. PixInsight aligned my atmospherically distorted images on my first attempt, and with breathtaking speed because it utilised all 8-cores of my Windows 7 machine. I did not have to adjust any of the default settings. The people who develop this s/w are very gifted. I think that I will still be exploring the capabilities of this s/w in two year’s time.

[DavidNg]

Wonderful image Murray. Thumb up for PI too from a beginner.

[RickS]

Very nice, Murray!

[LewisM]

GORGEOUS result! Worth correcting it!

I think CCDStack will align it too - you have a choice of 5 registration routines, from using the FITS header data/CCDInspector plugin, to Starmatch, Star snap, FFT and manual, with varying transform algorithms available with that.

DSS failed to register my first ever LRGB from low elevation targets (Rosette, rising) - bottom corner seriously "trailing" the rest of the image. An auto align in CCDStack, and the whole lot lined up first time, not issue.

Best software money I have spent yet. And it does it all SO quickly and accurately, with a LOT of adjustment parameters along the way. Expensive, but worth every cent.