Part 4: Novice Webcam Imaging Tests - Transit of Mercury

Fred Espanek's article Transits of Mercury indicates how rare these events are: "During the seven century period 1601 CE to 2300 CE, Earth experiences 94 transits of Mercury across the Sun." 3.5 to 13 years may pass between them; but they're not always visible from every place on earth, of course. There have only been two since the start of the 21st century, and I can't remember the event of 2003 at all. I know I have never witnessed one before, so it seemed useful to try to document it with our telescope and webcam.

But, have you heard of "Murphy's Law"? It did not disappoint!

To make sure that things would go smoothly during the event, I set up our equipment as thoroughly as I thought necessary, on Tuesday, 7 November. Our garage has two permanent desktop PCs set up near our workbench. We must run a 20 foot USB 2.0 extension cord from the machine, out to the place in the driveway where our telescope is set up. But, because of the time of year, Sol was low in altitude, and there was absolutely no chance of imaging or even seeing second contact, when Mercury's transit ended, from our property. I used a planetarium program to plot the exact position of the sun each hour during the transit, obtaining its altitude above my horizon. My measurements indicated that I had a window from 11:19am, time of first contact, to about 2pm, when the Sun would go down behind our garage roof. I could not move the scope further away and still take a picture. I ran a dry-run, taking a picture at infinity of power lines a couple of miles from our house, on the ridge of a distant hill: fuzzy, but discernible. Now I knew what the focus should be, and had the entire system working, saving pictures.

Below, the setup of webcam on Sky View 120: though this picture was taken later, it is almost exactly the setup for the Mercury transit. Since the star diagonal has absolutely no visually degrading effect through my cheap webcam, for ease of use I leave it in place.

Murphy's Law event no. 1: the weather. While I was not completely unlucky, "first contact" at 11:19am on Wednesday was lost: it was completely overcast. I sighed with relief that I had not succumbed to the temptation, at the start of the week, to plunk down nearly $100 for a single axis drive corrector for the scope, JUST for the Mercury transit! I had figured that if I did purchase it, I wouldn't be able to see the transit at all, nor use it. But if I did not get it, this would guarantee that I could indeed see, and photograph, the transit. It turns out that this logic was quite correct, for by 1pm, most of the sky was clear -- except for a stubborn white fluffy patch of clouds directly over the face of the Sun!

I went ahead, discouraged but stubborn, and set up the gear. YES: finally, full clearing! I quickly began to try to focus; but unlike the previous day, when I had done the tests later in the afternoon, this time the Sun was right in my eyes when I looked toward the PC monitor that is facing the driveway. I have to use binoculars to see it, for critical focusing; but the bright sunlight washed out the view. I called Regina out to help me, just before her first student of the day, and the two of us managed to get the telescope focused properly, in time to start snapping stills and taking short videos.

Murphy's Law episode no. 2: XP Pro refused to open properly the Logitech camera control panel, which had been working correctly 24 hours earlier. The application would come up without the "radio buttons" that operate the functions. Closed; reopened: same thing. Rebooted. Ditto. Rebooted AGAIN. The same: sigh...

That meant that we could not use the "good" PC, but had to fall back on the old one next to it, running Win ME. Frantically, I booted it up; changed the USB cable; and gave a big sigh of relief when the Logitech control panel opened properly. We started imaging.

There was a remarkable, huge sunspot at the limb, with the black shimmering dot of Mercury's disk well away from it, requiring careful framing to get the two in the same field in the webcam image. The raw feed on the PC monitor looked greatly inferior to the live eye view, but I reckoned that processing could bring up latent detail. Sadly, I did not have access to a better, high resolution planetary imager like a Meade LSI, which Charlie Chew and I used for our solar system experiments discussed in the early entries in this series. The Logitech webcam image is comparatively soft, lacking some dynamic range.

Here is the best of the still images, a 1/750th second exposure with the 120mm Orion f/8 achromat and the Logitech modified webcam. Some processing was added by Arcsoft PhotoStudio 5: a bit of unsharp masking, some sharpening, and tone gradation shading.

After we had taken some usable images with the refractor, the Sun was grazing our roof in its field of view, so I departed from my plan and got out Regina's little Orion "Starblast" 4-inch reflector scope, which costs all of $179. The Newtonian system, at f/4, would have a wider field of view. But, not being on a German equatorial mount, the difficulty of keeping the telescope pointed long enough to get back to the computer and snap the picture -- or even to start making a video -- was huge. I'd adjust, and run ... go back adjust, run to the PC ... and so forth. One video came out fine, as did two stills.

Still image taken with Regina's little f/4 Newtonian telescope with 4" mirror: an Orion Starblast, which costs less than $200. Exposure duration is 1/500th-second, with a bit less processing than used in the higher-resolution, larger scale image done with the refractor.

Murphy's Law episode no. 3: wouldn't you know it? The USB extension cord, which was supported on a ladder, worked its way loose despite being held down at several points by clamps. As soon as the camera was disconnected, Win ME locked up, leaving one picture frozen on the screen of the Logitech camera software program: it had not yet been saved to disk, and we lost it (a darned good one, too!) when the machine was rebooted.

Murphy episode 4: by now, I was pretty exhausted. I was just about to make a final careful adjustment to reposition the scope and do one more AVI file. But a utility worker who had parked down the street was puzzled by the strange activities at our house, and walked up right behind me, clearing his throat: startled, I suddenly JUMPED. The scope view of the Sun went skittering away offscreen. I have to admit that I was inconsiderate enough, not having more than a couple of minutes of time left before the sun disappeared, to growl curtly at the guy. "Sorry: I can't talk; I'm trying to do something important here."

He stroked his chin and gave me a VERY strange look. "Important, eh?" he said, with a very suspicious and doubtful tone of voice. Frankly, I had to agree with him!

That seems to have broken the spell and from then on, nothing went right and I got no further usable pictures. Murphy triumphs!

But, at least we have some raw data to work with. I opened up one of the short video files I had taken, using a 25 second AVI file made with the Logitech webcam on the refractor. I set Registax 3 to "throw away" the poorest frames with least detail, and ended up with 63 good frames. Since the telescope did not have a clock drive, and was fixed in position, the image drifted quickly across the field. I was not able to register the entire width of the frame, which covered about 15 to 18 arcminutes, but was able to "lock" onto the small discrete image of Mercury's disk. The stacking process greatly improved the sharpness of the image, as did final wavelet processing.

63 frames of best steadiness from a 25 second AVI file, taken with Logitech webcam and Orion 120mm undriven refractor, processed with Registax 3.

That's actually about the way that Mercury's disk looked by eye, at fairly high magnification, during the steadiest moments of seeing! These pictures are certainly not world-class amateur images, especially when compared to something like this large scale, detailed picture done by Matthew Ota with a 10" SCT and a camera with much higher res than our lowly webcam. Nevertheless, our little experiment turned out results that surpassed our expectations. If you'd like to see what the raw video looked like, click here for a Windows Media version file of 107 kB (note that the mike in the webcam picked up the noisy utility truck parked down the street!) The amount of improvement wrought by throwing away fuzzy frames and registering, stacking, and processing the best ones is simply staggering, thanks to Registax. Murphy may have had his fun, but he left a little entertainment untainted for Stephen and Regina to enjoy, and share.

Addendum, 14 July 2007: For another amusing little "Murphy's Law" anecdote involving a solar photograph, read Jerry Lodriguss' account of imaging the Internation Space Station against the disk of the Sun, found here on Jerry's blog.