The Starry Night, 2

:: home ::

              Go to 1   2   3   4   5   6   7   8   9   10   11   12   13   14   15   16   17   18   19   20   21   22   23   24
              25   26   27   28   29   30   31   32   33   34   35   36   37   38   39   40   41   42   43   44   45   46   >>   SRCH


Who has a time-lapsed lunar eclipse? (1/16/2009). My year began with an interesting request. A filmmaker working on a PBS project asked if I had images that could be used to put together a time-lapse film of a total lunar eclipse.

He'd seen my photo of the stages of total lunar eclipse from 2003 on NASA's APOD site and got in touch in the reasonable (but erroneous) belief that I probably had additional images from that event that could be melded into such a film. No, I don't, but I did have a sequence from 2007 that might work, I said. For the August 2007 eclipse, I tried to shoot a uniform series of frames 30 seconds apart. It didn't work out very well owing to clouds of varying density and because it was tough to settle on one exposure that would work for the entire event. If I ever try it again, I'll shoot a widely bracketted burst every 30 seconds rather than one frame every 30 seconds. And I'll use an intervalometer rather than my wristwatch. And I won't take the camera off the telescope until the eclipse is over no matter how convinced I am that the sky is about to close the show with fog, or clouds, or rain.

So you see what I was dealing with: images of differering exposure and orientation, with gaps owing to clouds. Here are a few sample images.


After aligning the images and rotating the Moon to the same orientation in all frames, I used a $20 piece of shareware to meld them into a short time-lapse sequence with single frame fades between successive images. The last couple of images are lingered over to let you see the totally eclipsed Moon before the clip recycles. This is not intended to do anything except demonstrate that there are enough images here to produce a smooth time lapse (and be kind of fun to watch). Software with a few more zeroes in its price (or a lot more time with the tools I have here) would make a big difference in lots of ways in turning this into broadcast quality video.

[Clips removed -- inconvenient codec used]


The 2007 eclipse ended with the sky brightening and the still-totally eclipsed Moon falling into the wooded western horizon:

[Clips removed -- damned codecs!]


I got curious about how well I could normalize the exposures by layering a master shot of the Moon made just prior to umbral contact under each exposure of a partial phase and then adjusting the curves, levels, and alpha values of the top layer. I think the finished results could be very good, but since the sequence isn't long enough in any case, I'll forego trying to do that for the time being. Here are some of the frames from what I call the normalized sequence:


Promising even if not ready for Hollywood. So. Do YOU have the images the PBS filmmaker needs? The sequence needs to run from the beginning of the partial phases, through totality, and then on to the end of the second partial phases. The finished clip will run only 6-7 seconds. At a guess, one frame a minute should be plenty. Drop me a line and I'll send along your contact information: [UPDATE: producers have decided to use an alternate sequence; too much post-production required to come up with a useful lunar eclipse animation.]





:: top ::


                   © 2010, David Cortner