Total Lunar Eclipse 2019 Jan 20

I've had a lot of questions about how I captured the Lunar Eclipse, so I thought I should share the whole story here. It's full of mistakes and imperfections, but there should be plenty to learn from.

Last Sunday was the total Lunar Eclipse -- for some reason I had it in my mind that it wasn't visible in this region, but when I started getting questions about it from people in the US, I looked up the info and found that I was right in the center of the region where it would be visible (middle of the US), and that the skies were predicted to be clear! I only realized this last Friday -- there's been plenty keeping me busy, but I'm not sure how I almost missed this event. I didn't get the chance to do a test setup beforehand, but as you’ll learn if you keep reading, pre-tests should always be done when possible!
I was at my in-laws that night for dinner with my family, so I brought everything with to set up and leave there since there's less light pollution at that location. I used a Fuji X-T3 with the 100-400mm lens at 400mm. This is an APS-C sensor, so this makes for a 609mm equivalent on a full-frame (4.1° FOV). I used the Dynamic Perception Sapphire pan/tilt head with the NMX controlled connected to the VIEW Intervalometer via a USB hub. The camera was also connected to the hub, which was externally powered by a USB battery pack that had USB as well as a 12 volt output that I connected to the NMX for the motors.
I forgot a power cord for the VIEW, but didn't think it mattered since I was powering the hub (the X-T3 draws power from the USB), although in retrospect this was a problem, likely because the USB voltage from the battery pack was a little lower than that from the VIEW.
Since there was a lot to setup and connect, I set it all up indoors and then brought it outside around 6:30pm to get started (this was a mistake!). I enabled the VIEW's built-in wifi access point and then set the tripod with the whole setup outside. I set the exposure (1/200, f/8, ISO 200), used AF to focus on the moon, then switched the camera to MF and went inside to do the rest from my phone connected to the VIEW's wifi.
I used the VIEW's app to move the motors to center the moon, making sure that the last direction I moved each axis was the same direction it would be moving to follow the moon (to take up backlash, which is quite significant at this focal length). I set both the pan and tilt axes to follow the moon. At this focal length, the moon would exit the frame in just a few minutes if it wasn't tracking it, whereas I needed it to stay in the frame for about 6 hours. I set it to auto ramping with the shutter limited to 1 second, a fixed interval of 20 seconds, and a length of 2000 frames (more than enough) and left it there and went back to get the kids to bed.
Back at home I was concerned about the VIEW's battery in the cold (it was -12°F/-24°C), and not knowing if the camera would be drawing extra power from it through the hub, so I drove back over to check on it and connected to the VIEW on my phone from the vehicle. The app showed the VIEW's battery at 0%! ! I quickly grabbed a charging cable and went over to it, finding it already powered down. It started back up, and connected the app again and pressed start (all the settings were still there). Once it was going I moved the tilt axis by 0.5° to catch back up and re-center the moon in the frame, and then set the exposure to increase by 3 stops over the next 20 frames since I had interrupted the ramping and needed it brighter during totality that had just begun. Making the exposure change gradually after restarting it made for a better result in post rather than changing the settings on the camera before starting. Thankfully, the slight pause to restart (probably 2 minutes between frames instead of the 20-second interval) happened right at the point before complete totality, so it was thankfully unnoticeable. I left it there just as totality was beginning and went to bed.
In the morning I went back and picked it up -- a little worried about how it had gone. All the batteries were dead, but the camera was pointing to the West, so I was assured that it had been running and tracking. Loading the images on the computer showed that it started out beautifully crisp but after the first 20 minutes, the focus was no longer very good. I attribute this to the temperature change -- the lens is basically a metal tube, and since I focused at 70°F/21°C and then moved it outside to -12°F/-24°C, it's not surprising that it would have contracted some, which would change the focus. After the initial change during the first 20 minutes, it remained stable after that, which also agrees with this theory. Despite this disappointment, the tracking and exposure was just perfect. For the tracking, the goal was simply to keep it in the frame the entire time -- and it worked. Within the frame, it did move around a bit, especially as it moved across top center and the weight shifted over the tripod (it's not a balanced setup). But it stayed in the frame, which is all I needed.

I edited the images using the Timelapse Workflow plugin for Lightroom. This allowed me to smooth out the exposure changes and blend the white balance across the transitions (different white balance for totality vs partial). I increased the sharpness and didn’t crop as much as I would have liked in order to cover up the focus issue as much as possible. Then I exported them as high-quality JPEGs and loaded them into Apple Motion.

In Motion I stabilized the footage so that the moon remained basically stationary the entire time. I had to manually align a few frames right at the point before totality since it had some trouble tracking across that transition. Then I rendered that and loaded it into Final Cut X where I added the music, "booms" and a Ken Burns effect so that there would be a natural amount of motion across the frame that somewhat aligned with the motion of the shadow.

I've had some questions about the moon rotating -- this is because with the pan/tilt tracking method, the VIEW is keeping the horizon (not visible) level, so as the moon passes overhead it will rotate on the pan axis before tilting down again (so tilt never goes beyond straight up).

In conclusion, some lessons learned:

  1. Test everything beforehand! This is an opportunity to find out what can go wrong that you didn’t expect

  2. If dealing with temperature extremes, let the equipment acclimate first. I should have left the setup outside for an hour before focusing and starting it.

  3. I wasn’t sure how well auto-ramping would perform with the eclipse, but was the best option since I wouldn’t be monitoring it and I wanted to test it anyway. In the end, I didn’t fully find out how it would have done entirely unattended since when I restarted it at the beginning of totality (due to running out of power) I changed the exposure manually. Still, it handled the transition from totality back to partial perfectly (even while I was sleeping!). If I were to have been present the whole time, I probably would have opted for manual ramping while monitoring it from the app indoors.