The Making of "Lunar Eclipse over the Continental Divide"
Updated: Feb 15
From a photographer’s perspective, the moon is tiny. True, it’s 2,159 miles in diameter, but it’s nearly a quarter of a million miles away, so its angular diameter is just 0.5 degrees. To make it a meaningful part of your composition, you either need a long telephoto lens or you need a lot of moons.
That’s why I decided to shoot the May 15, 2022, total lunar eclipse as a “string of pearls” image. I’d lock the camera down on a tripod and make one exposure every five minutes from the time the partial eclipse began until the partial eclipse ended. Then I’d place all of the moons into one image that showed every phase of the eclipse.
This type of shot is simple in concept, but tricky to get right. First, I’d need to know where the moon would be at the beginning of the eclipse, both its direction (azimuth) and distance above the horizon in degrees (altitude). I’d need the same information for the moon’s position at the end of the eclipse. From that information, I could calculate the angle of view I needed both horizontally and vertically to include the moon’s journey through Earth’s shadow. The Pro version of Photo Ephemeris Web, an app which runs in a web browser ($30/year), had all the information I needed. During the May 15, 2022, eclipse as viewed from Rocky Mountain National Park, the moon’s azimuth would range from 119 to 162 degrees – a span of 43 degrees. Its altitude would climb from 3 to 27 degrees – a span of 24 degrees.
With that information in hand, I opened the field-of-view calculator in a mobile app called PhotoPills. A little guess-and-check showed that a 30mm lens on my full-frame Sony a7R IVa had a vertical angle of view of 43 degrees – enough to include the moon’s path through the eclipse plus some land below and a comfortable gap between the highest moon and the top of the frame. The field of view would be more than wide enough to include the complete eclipse sequence. Once the eclipse ended, I would need to continue making images every five minutes until the moon exited the frame.
With the basic parameters set, I next needed to decide where to go. The moon would be only 3 degrees above the horizon when the eclipse began. That meant I either needed to go somewhere flat, like the plains of eastern Colorado (too boring), or go somewhere high, like the summit of a mountain (much more interesting). If I was down in a valley, the moon would be out of sight behind a ridge when the eclipse began. In addition to going high, I needed to be looking southeast at something interesting.
I went back to Photo Ephemeris Web, the best photographer’s planning app that will run on a desktop or laptop computer. Maps are most useful when they’re big. Photo Ephemeris Web lets me plan shots on my 27-inch NEC MultiSync 272W monitor. With the app open in Chrome, I began exploring possibilities.
What would be the easiest way to get up high? Trail Ridge Road, which goes to 12,000 feet, was still closed by snow. Loveland Pass was open and equally high, but the moon would be over the Arapaho Basin ski area for much of the eclipse, and the image would lack a wilderness feel. Hallett Peak, in Rocky Mountain National Park, emerged as the logical choice. The summit offers a great view looking southeast toward Longs Peak and the Continental Divide. Photo Ephemeris Web showed that the partial eclipse would begin with the moon to the left of Longs Peak and end with the moon a little to the right of Powell Peak. I’d hiked the trail to Hallett Peak many times with my wife Cora, so I knew the terrain. I’d even snapped a study frame of the skyline from Longs Peak to Taylor Peak from the summit of Hallett. Now I could refer to that study frame to confirm that the image I imagined when plotting angles on the map would look good in reality.
The final puzzle was figuring out how to point the camera up at the right angle to include the full path of the moon through the frame. Compositionally, I thought it would look best if the moon exited the right side of the frame rather than the top. The moon would transit – reach its highest point in the sky, 29 degrees – at 1:05 a.m. when it was just to the right of Taylor Peak. That’s when it would exit the right side of the frame. I decided I needed six moon diameters – 3 degrees – between the highest moon and the top of the frame. If I leveled the camera front to back with a 30mm lens (43-degree angle-of-view vertically), then my frame would include objects within a range of 21.5 degrees above a level horizon to 21.5 degrees below a level horizon. Since the moon would be 29 degrees above the horizon at its peak, and I needed an additional 3 degrees to give my string of moons room to breathe within the frame, the altitude of the top of my frame needed to be 32 degrees. In other words, I needed to point the camera up 11 degrees (32 minus 21.5, rounded off). I would have to keep the camera locked down on the tripod throughout the eclipse sequence, so I would have to position it correctly long before the moon actually reached its high point in the sky. My solution was to bring a lightweight multi-row panorama setup, Really Right Stuff’s PG-01, which has a degree scale on the pitch control. (Pitch is the number of degrees the camera is pointing up or down.)
With the planning complete, I began contemplating the logistics of the shoot, which were daunting. It was only mid-May, so there was guaranteed to be a lot of snow on the trail. In the spring, the snowpack goes through a daily melt-freeze cycle. Travel is always easiest early in the morning after a cold night, when the snowpack is well frozen and will support your weight even without snowshoes. Unfortunately, I would be hiking up late in the afternoon, after the snowpack had thawed in the spring sun. Without snowshoes, I could easily be plunging in knee-deep and beyond. I added snowshoes to the small mountain of clothing I would need to be comfortable hanging out at 12,720 feet for many hours in the dark. I kept the camera gear to a minimum: Sony a7R IVa, Sony 24-70mm f/4 lens, Really Right Stuff Ascend-14 tripod, the PG-01 pano kit, and some extra batteries.
With the exception of a few short dry stretches, the trail was snow-packed all the way from the parking lot to the summit of Flattop Mountain. I wore Kahtoola micro-spikes where the trail was well-packed, switched to snowshoes at timberline, then lashed the snowshoes onto the pack again for the rocky traverse along the Continental Divide from Flattop Mountain to the final ascent to the summit of Hallett Peak.
I arrived on the summit around 6:00 p.m. after four and a half hours of hard labor. Dramatic cumulus clouds were billowing up over the Continental Divide. The sight was beautiful, but I hoped they would vanish quickly when the sun got lower. The partial eclipse would begin just 16 minutes after sunset. A thick band of haze near the horizon also threatened to block my view of the first minutes of the eclipse.
Fortunately, the clouds dissipated as sunset drew near, and the band of haze proved thin enough to let the moon shine through shortly before the eclipse began. I began shooting bracketed sets of images every five minutes using the correct exposure for the moon. In addition, I shot several bracketed sets of images using the correct exposure for the land and background sky. These exposures were much longer than the exposures for the moon during the partial phase of the eclipse.
The total eclipse began an hour and 17 minutes after sunset. The moon turned red and grew very dim. Many more stars became visible. My exposures became much longer. I increased the ISO as needed to keep the exposure time to no more than 8 seconds. Much longer than that, and the moon would have begun to blur due to the rotation of the earth. The moon remained totally eclipsed for an hour and 26 minutes. A few clouds drifted across the sky, so I shot whenever the moon was clearly visible through a gap in the clouds. To my great relief, the last of the clouds passed by as totality ended and the final phase of the eclipse began.
At last midnight rolled around and the partial eclipse ended. I shot one more bracketed set of the once-again full moon and packed up. The moon hadn’t yet exited the frame, but I was tired. I knew it would take at least three hours to descend to the parking lot, plus another hour and a half to drive home. Why keep shooting still more frames of the now-unchanging full moon as it slowly crept across the sky toward the edge of the frame? I could just make multiple duplicates of the final full-moon frame and position the additional moon images correctly using data from Photo Ephemeris Web. I headed down.
Three hours later, after losing the tracks once in the dark, then finding them again, I reached the parking lot and drove home. At 5 a.m. I collapsed into bed after a 23-hour day.
To assemble the shot I first selected the correctly exposed moon from each bracketed set. Then I chose a background image shot about 35 minutes after sunset, when the sky was a deep, rich blue and the peaks still had adequate detail, and loaded all the frames as layers in a single Photoshop file. With the background image on top of the layer stack, I changed the blend mode to Lighten for all layers. In this blend mode, Photoshop drills down through the layer stack and reveals the brightest pixel. That meant that all the moons suddenly showed through, along with the background sky and mountains, which were brighter than the black sky surrounding each moon. I saw immediately that I hadn’t pointed the camera up quite enough, so the line of moons seemed to be crowded against the top of the frame. The sky in the background image was a uniform blue, so I used Content-Aware Fill to add some sky at the top of the frame. The image was obviously a composite, so taking a few additional liberties to assemble the image I envisioned seemed acceptable. I hadn’t added anything that wasn’t actually present. I also duplicated the final full-moon image and placed the additional full moons in the frame, using data from Photo Ephemeris Web that told me exactly where each moon should be placed.
In retrospect I should have summoned the discipline to stay another hour and a half on the summit until the moon exited the frame. And I should have given myself a larger margin for error when I pointed the camera up to include the highest moon. Despite my disappointment at making those mistakes, I’m still pleased I was able to make a string-of-pearls image of a total lunar eclipse over the Continental Divide from the most spectacular vantage point imaginable.
