Thanks to Swaroop Shere for the excellent image of the Corona. The rest of the timelapse images were shot at 800mm with an 8'' f/4 astrograph and Canon t3i DSLR.
Compositing this video was a huge pain. :-)
I had about 500 individual timelapse frames, but the alignment wasn't spot on, and I had to periodically push the wandering sun back into the middle of the frame across the ~2.5 hours of the event. Because of this the images needed substantial alignment to look OK, and normal alignment software has no conception of how to deal with the aligning on a disk that is only partially visible, and no part of the disk is visible in all frames. I ended up using PIPP:
https://sites.google.com/site/astropipp/ PIPP seems to have a concept of aligning on the edge, seemingly for lunar applications...great! It took some massaging, reversing, and manually adjusting as it seemed unhappy aligning on the opposite edge for the second half. After that I manually adjusted Swaroop's Corona image to have roughly the same image scale and added a transition in between the two halves of the video.
I wasn't sure what the correct orientation between the two sources was, but in retrospect I think the corona is actually 90-ish degrees off from the actual orientation relative to the animation. (Not that it matters a lot, but I expect someone will notice!)
One other cool thing is that you can see the sunspots move a little from beginning to end. The movement should be approximately 2 degrees, as the sun rotates at the equator every ~25 days (14.4 deg/day), and this was almost 1/8th of a day (=1.8 deg). It looked bigger than a couple of degrees to the eyeball, but I carefully aligned and measured the distance and it came out to ~20 arcseconds on a point near the center of the sun; since the sun is ~1800 arcsec, that works out to 1/90th of the face (which is 180 degrees), or 2 degrees. So hey, it's consistent!
Comments