December: Test shots with new scopes/mounts
Dec 21: TMB 80/480 Arrives!
Dec 3: AP1200 Arrives!
Nov 30: TMB 152/1200 Arrives!
This is my first attempt at an ultra-widefield image of the Milky Way, inspired, in part, by Axel Mellinger's astonishing All-Sky Milky Way Panorama to which my modest attempt hardly compares. Check out the next frame to the left, shot soon after this image was taken, and also a panorama containing both of these images. Also check out the large (712 kB) or small (113 kB) animation, showing the earth rotating undernearth the Milky Way over the course of several hours. Each frame of the animation is one of the 32 exposures that were ultimately stacked to form the image shown above.
The Canon 14mm lens captures a generous 77° × 56° swath of sky with the 350D's sensor. But that large field comes with a significant amount of distortion and chromatic aberration towards the edges, surpassed only by (here comes a technical term) a boatload of coma, as can be seen in the higher-resolution image. (These aberrations did not appear to improve by stopping the lens down to f/4.0 or to even f/5.6.)
Those of us at mid-Northern latitudes are at a significant disadvantage when trying to shoot the Milky Way's core. This is because the core is located in the Southern Celestial Hemisphere, at a Declination of -29°. Therefore, it never gets higher than roughly 23.5° above the southern horizon at my latitude. And because I have fairly tall mountains to the south, well, a road trip to a site with better horizons was in order.
So, the other member of the Saratoga Astronomical Society (aka Rich) and I loaded up the RX300, and headed to Montebello Open Space Preserve. Montebello is a reasonably dark site with very good horizons in almost all directions. I think it was well worth it, because the photo above reaches Declinations within a few degrees of what is theoretically achievable at our latitude.
Processing this photo was a significant challenge, as the usable portion of the image — i.e., that portion which was above the horizon — changed appreciably from frame to frame. I therefore wrote some code to implement what I've named Masked Kappa-Sigma Stacking. The algorithm allows only valid data to be included in the sigma computation and, ultimately, in the stack at any given pixel location. A mask image determines which data is valid and which is not (i.e., because it's at or below the horizon) for each, respective image. Of course, the natural and unavoidable consequence of such an algorithm is that the closer you get to the horizon, the noiser the image, because relatively few exposures are being stacked. In the limit, only a single image is being “stacked”. And, of course, the closer to the horizon, the more atmosphere the camera is looking through, and the worse the captured image will be regardless. A future article will describe this technique in detail.
Mouse over the image to see constellation outlines and some other annotations. The image is oriented along the galactic plane, which is to say, the Milky Way's North Galactic Pole is up. Earth's North is inclined at roughly 27° with respect to the galactic plane (63° with respect to the North Galactic Pole), as can be seen by the direction arrows in the mouse-over. A higher resolution image is also available.