Experiments with the Canon T1i continue, this time with a 400 mm telephoto lens. On Wednesday night, November 3, 2010 (November 4 UT) I "piggy-backed" the T1i with the 400 mm lens attached to the 10-in LX-200 at Broemmelsiek Park and pointed the assembly at M31, the Andromeda Galaxy. This combination gives a field of view of some 3.17x2.11 degrees which is well suited for M31. (We might note that the diagonal of this field is about 3.8 degrees and how much the galaxy extends along it.) A few test images suggested that the telescope would track well for at least 30 second exposures at this focal length and a setting of f/16 on the lens would minimize the effects of coma in the corners (although, after processing, I believe that was extreme and f/8 would probably be acceptable). I then set up the camera control software to take a series of 30 second exposures in the RAW mode (some 19MB per image). I ended up with 54 images of 32 second exposure each (even though I had set up the software to take 30 sec images). At the end of the exposure run I capped the lens and took 16 dark exposures of the same exposure each.
I processed the stack of images in Deep Sky Stacker 3.3.2 and still have a lot to learn about this program. In particular, i could only get it to stack 19 of the images (608 seconds of effective exposure). There must be some setting(s) that I can adjust so will continue studying the matter.. I like the fact that DSS applies a master dark derived from the 16 dark images to each of the stacked frames. When the stacking is completed (takes a long time), the RAW images are decoded into R,G and B layers and a histogram shows the distribution of each color. By adjusting the sliders in the display I (more-or-less) aligned each of the histograms on top of each other which seemed to achieve a decent color balance for the final image. I saved the final image as a 16b/ch TIFF image and imported that into MaximDL where I did a lognormal histogram stretch and tweaked the final histogram slightly to get maximum spread of the pixel values in the image. I then converted the image to monochrome (not too interested in what color there may be in this object), did a screen stretch at 8b/ch and saved the image as a jpeg. The final result is shown below (click on the image for full size).
Note that the 3x3 binning yields pixel sizes of 7.2x7.2 arcseconds. The result is still pleasing for a "pretty picture" and any coma in the original image is well masked. The original image stars were some 32 arcseconds FWHM.
I finally got DSS to stack 53 of the 54 images. The trick? I used "Superpixels" wherein the Bayer matrix is not interpolated for each pixel but a 2x2 "superpixel" containing two Gs and one R and B each is taken for the color components. This results in a lower resolution (which we too much of anyway for this setup) but each image is 1/4 the size of the full image - hence four times faster to process. Final stacked image was loaded into MaximDL, stretched to use full 16 bit values, converted to monochrome, histogram stretched using lognormal function, binned 2x2, contrast adjusted to darken background, screen stretched to 8 bits and saved as JPEG. Total exposure = 1,696 seconds.
A new image taken 7 November 2010 using same setup as above but this time I opened the lens aperture to f/6.3 (widest possible) and set the ISO to 800 (lower noise). I got 120 32-second exposures, stacked them in Deep Sky Stacker using the "superpixel" mode and aligned the three color planes at the end. Post-processed in MaximDL with a bit of stretching optimization and then a lognormal stretch. Did a screen stretch to 8 bits for this jpeg image. If you click on the image to show it at full scale, you might note the bright stars have hexagonal (octagonal?) diffraction patterns due to the shape of the iris in the lens.