Deep Sky Images

This is where I will put occasional deep sky images.

With the acquisition by ASEM of a Canon T1i DSLR, a new dimension to the capabilities for members to pursue advanced activities in astronomy.  I set out to characterize the use of the camera with the telescopes at Broemmelsiek Park.  My first trial was using the 10-in  LX-200.

I plugged the camera into the eyepiece tube of the telescope using a Canon to T-adapter and a T-thread "nose piece."  This set up uses the 2,500 mm focal length of the LX-200 without amplification or reduction.  At this focal length the field of view of the camera is about 1830x1219 arcsec.  That's wide enough to go all the way across the disk of the Moon.  It is also quite compatible with many deep sky objects.

I started off  by focusing on Vega using the Bahntinov mask I described in an earlier post.  Then I commanded the LX-200 to slew to M57 (the Ring Nebula).  There is still much to learn about how long an exposure we can take with the LX-200 without having smeared images and I was testing software and technique anyway.  So I settled on 10 second exposures and used the software that came with the camera to take a series of these 10-sec exposures for M57 (and later for M27 and M13).  I set the camera to deliver RAW images which are about 16 MB each.  At the end of the session, I took a series of dark images to create a master dark for later processing.

I processed with a new (to me) program - Deep Sky Stacker.  I input the list of dark frames and let DSS create the master dark frame by stacking and averaging the images.  Each of the dark images was also exposed for 10 seconds - an important factor.  After that, I loaded the master dark with the list of the light files for each of the three targets and started the process.  DSS examines each image for quality and to locate stars with which to align the frames.  Quality seems to be determined by the shape of individual star images.  This seems to be a reasonable approach as I examined the images that DSS rejected to find the star images were usually trailed.  DSS requires a minimum of 8 stars in each image to include in the final stack.  DSS will also handle and apply flat frames but I haven't tried that yet.

The files produced by the camera in RAW mode are quite large.  The images contain some 15 megapixels (4770x3177) and it can take quite a while to go through a large stack.  The final image that is produced is a 32 bit/channel (three channels for R,G and B) .tiff file and is HUGE - some 70-80 MB.  This is too big for MaximDL to process so it was necessary to save the final image as a 16 bit/channel .tiff file.

An image of 4770x3177 pixels is WAY to large to display on a typical computer screen which is the usual mode for sharing the images.  Also, at the 2500 mm focal length each pixel covers about 0.4 arcsec.  In Eastern Missouri, our chief limitation to image resolution is the seeing disk produced by the tiny motions of the atmosphere during any exposure over a few seconds (maybe milliseconds).  Typically, we are very lucky to get 4 arcsec (FWHM) seeing disks so a 0.4 arcsec resolution is overkill in the extreme.  Accordingly, the processing I settled on was to do a low pass filter in MaximDL on the full size image to smooth the data a bit, then I did a 3X binning which cut the image size to 1590x1059 (still large for many computer screens).  It also cut the size of the image file by 9X.  The 3X binning also rounds up the star images that may have a bit of trailing or coma in them.  I then saved the re-worked image as a JPEG file which also reduced the size of the file to make for easier downloads over the Internet.

Below are the final images of this process along with some cropped images.  Click on the image to see the full resolution.

I am not bragging about the quality of these images but I am very pleased with them in that they hint at what is possible with equipment available to ASEM members. Clearly longer effective exposures (more 10-sec sub frames) are necessary for great images.  We can also explore how long the sub frame exposures can be.  Other experiments would be to use a Powermate (as is commonly used with lunar and planetary imaging) or a focal reducer to give even wider fields of view while minimizing tracking errors (longer sub frames).


Stack of 13 10-sec sub frames

Final image binned 3X

Stack of 12 10-sec sub frames

Final image binned 3X

Stack of 60 10-sec sub frames

Binned 3X, jpeg mode


M13 cropped
M57 cropped
M27 cropped

M16 cropped

142 10-sec sub-frames stacked

RAW mode, Low-pass filter, binned 3X, cropped to 600x800

Stacked with Deep Sky Stacker

10-in LX200, f/10 (2500mm)

Canon T1i camera, ISO 1600

Clearly needs more exposure but "Pillars of Creation" starting to appear.

M11 Cropped

20 10-sec sub-frames stacked

10-in LX-200, f/10 (2500 mm)

Canon T1i camera, ISO 800
JPEG mode, 3X binned in camera