First-time eclipse viewers are usually advised not to worry about photos and to instead take in the show. But I am a shutterbug, and by the time the eclipse rolled around, I had been plugging away at low-budget astrophotography (by astrophotography standards) for close to half my life. I was not going to simply avoid taking pictures of an event I had looked forward to for most of my life. But that said I also understood that solar eclipses, especially for a first-time viewer, can draw unexpected reactions. My goal, then, was to automate as much of the process as possible and hope muscle memory could handle the rest.
I had two DSLR cameras available to me, a seven year old Nikon and a Canon that was six months old. Having two DSLRs were good, because I had two totally separate astrophotography goals. One was to capture a landscape time-lapse spanning the entire eclipse. The other was to do some mobile landscape photography and capture some closer images of the eclipse during totality. The two cameras meant I could do both.
I set the Nikon on landscape duty. Although I knew my way around the Nikon much better than the Canon, I wasn't totally confident in its ability to handle the lower-light conditions during totality. Having previously worked with the camera's raw format for astrophotography, I knew the photos would be noisier and that there would be much less room to recover from mistakes in setting exposure length. I was also more confident in my ability to set up an hours-long timelapse with the camera. I figured the best way to use it was to set it on a tripod, set it to aperature mode, and simply let it snap away.
That left the Canon for mobile photography and the main goal of capturing photos of the Sun during totality. Relying on the Canon for primary duty was iffy. Although it had much more advanced methods for automating imaging sequences, I was still learning my way around how to use them. But, programmed correctly, it would dramatically simplify my ability to get good pictures during totality. I could program three separate sequence types into the camera, and then switch between those sequences by simply rotating the camera's mode dial. All I would need to do was remove the lens cap prior to totality, and then rotate the mode dial by feel before capturing different sequences. The only effort on my end would be figuring out which kinds of pictures I wanted to take during the eclipse, and then getting an appropriate sequence programmed in.
Early on I decided to try for a mid-field view of the Sun and corona. This plan went against the typical types of photos taken during an eclipse, which try to zoom in as much as possible to capture as much detail as possible in the corona, or widefield photos that capture the Moon's shadow projected onto the sky. In part, my decision to do a more unusual type of photography was due to being an experimentalist at heart. I definitely wanted to find a different angle on eclipse photography. But there were very real constraints also driving the decision as well. A wide-field photo was definitely out. Trees were packed in tightly around the creek, preventing me from seeing enough of the sky to actually allow the Moon's shadow to be recognizable in a photo of it. The reality of my telephoto lens was that it would be incredibly difficult to get a good photo of the eclipse too. At maximum zoom, I would have to make adjustments to make sure the corona was properly framed. The telephoto was also borderline too heavy for my star-tracking mount, and attempting to make fiddly adjustments to it would more than likely eat up most of totality.
The mid-field photo let me make the most of my equipment, and didn't feel like a totally random choice. I was hoping to get a few stars in my photo. I had seen very few photos of an eclipse with visible stars, even though I had heard that eclipses got dark enough to see some of the brighter stars in the sky. Mars, Jupiter, and Regulus would all be relatively close to the Sun, which made for a tempting target. I had previous experience of grabbing photos of stars and coments low in the sunset, so I figured I had a good chance at getting an HDR photo with stars. I also remembered attempts to image sungrazing coments in the low light of totality. Although the two minutes I had in the Moon's shadow wouldn't be enough to actually spot movement (and actually report a discovery), I thought it would be cool to spot a comet just beyond the watchful eye of the SOHO spacecraft. With luck I could push my processing to a lower limiting magnitude than the spacecraft.
With all this in mind, I developed the following plan. In the hours before eclipse, I would carry the Canon around and capture the changing landscape as totality approached. About 10 minutes before totality, I would post up the Canon on my tracking mount and get it ready for totality imaging. The mid-field view would use a focal length of 85 mm with a smaller adjustable zoom lens, which would provide me with a 10x7 degree field of view around the Sun. I would keep the lens at a fixed focal length throughout this phase of the eclipse. Although this meant losing out on the best possible photos of the diamond ring effect that I could get with this telephoto, it minimized the amount of effort on my part during totality. Changing the focal length brought the risk of accidentally knocking the camera out of focus, and I would be relying on manual focus to avoid autofocus problems in low-light conditions. I didn't want to lose precious time in the Moon's shadow attempting to get the focus right.
As totality approached, I would grab a short time-lapse of the diamond ring entering totality. This sequence consisted of 15 frames taken at one-second intervals, using and exposure length of 1/1000s at f/10 ISO 160. Shortly after the start of totality, I would rotate the mode dial clockwise to switch to an HDR sequence. Then, shortly before the end of totality, I would rotate the mode dial clockwise again to switch to a copy of the time-lapse sequence to capture the diamond ring at the end of totality. With luck, I would spend only need to spend at most a few seconds looking away from the sky.
The HDR sequence involved a bit of guesswork. The programmable HDR sequences were limited to six photos separated by up to three exposure values. On one end, I wanted the last exposure to be as long as possible to get good signal in the skies around the edges of the frame. However, to simplify the HDR processing, I needed to minimize the jumps in exposure values between images as too large a jump would be difficult to assemble into a seamless photo. I eventually settled on exposures of 1/2500, 1/500, 1/100, 1/20, 1/4, 1.3, and 6s exposures (or jumps of 2 1/3 EV between photos) at f/6.3 ISO 640. This differed quite a bit from exposure guides published by other eclipse photographers, but my understanding of those guides was that they were developed for narrow-field photos of the corona, not the skies further from the sun. this departure from recommended settings was an attempt to account for the fact that these exposure times were developed to capture detail in the corona.
As fate would have it, I spent so long dithering on what to actually do with the HDR photo that I didn't actually sit down to program it into the camera until the evening before the eclipse. I'll admit that procrastination also had a little to do with waiting so long. Putting it off this long was a bad idea, but I think I was so excited and nervous about the coming events that I don't think I would have been sleeping anyway. Sometime around 11 pm, I got the test sequences working properly, and I added in another 45 minutes of practice switching between the different sequences and testing my timers to build the muscle memory and shake out any bugs that might be lurking in the programming. Once satisfied I settled in for a short night of sleep.
Written: May 28, 2020
Edited: August 14, 2021