The ‘Great American Eclipse’ of 2017 – Revisited

The total solar eclipse of 21 August 2017 was a great experience. Not only was it a perfect eclipse weather-wise, taking place in a cloudless blue sky in mid-morning (as seen from Jackson Hole, Wyoming) and featuring a great display of shadowbands at both the start and end of totality, but as it occurred during summer vacation time I was able to take my son and daughter on the trip, allowing them to witness their first total eclipse.

I recorded the eclipse on both video and still cameras. For the video (which I have not yet published on this site) I disabled the camera’s automatic light levels to capture the dramatic drop in ambient light as totality approached, then receded. With my still camera I shot several series of bracketed exposures to capture the detail of the corona from the brightest part near the lunar limb extending to the dimmer parts several lunar radii away.

At the time I processed the images to compress the high dynamic range (i.e. the difference between the brightest and faintest parts) of the corona using a technique published by George Pellet (Sky & Telescope January 1998) and was pleased with the result which I published on this website here. My only regret was forgetting to zoom the camera out to capture the full extent of the corona, which was a pity because at that eclipse the coronal streamers were very long and the longest is cut off in my images.

But my interest in these images was re-awakened when I read Sean Walker’s excellent article “Revealing Totality in HDR” in the March 2020 edition of Sky & Telescope. Since 2017 I had brought my version of Photoshop up to date and I now had access to some of the more advanced tools described by Sean. I felt that there was more detail in my original images of the coronal streamers and loops close-in to the sun, and I also wanted to show the faint lunar surface detail illuminated during totality by Earthlight.

So I pulled out my 2017 images and processed them again from scratch. I had six groups of nine bracketed images so it took some time to align, stack and enhance each group before combining them into a final stacked set. In the end I rejected the sixth and final group because the end of totality was fast approaching and the sky near the moon’s upper limb was getting too bright. The result is shown below.

Comparing this result with the earlier post a lot more detail can be seen in the inner corona. Thanks to Sean Walker and the improved Photoshop CC!

The other goal was to display the moon’s surface features. They had only been revealed in my longest exposures (½ second) so I aligned and stacked these, enhanced them with the PS Camera Raw filter, and cut and pasted the circular disk of the moon onto my finished image of the corona.

Processing the moon is not straightforward: during totality the moon is of course in constant motion across the face of the sun, revealing first more of one limb, then the opposite one until the sun emerges to form the brilliant diamond ring effect.

For corona photography we align our images on the coronal streamers, so the moon is blurred in the stacked result. This is why a bright ring shows around the edge of the moon above, as the blurred disk is removed during processing.

Equipment: Nikon D810A DSLR and Nikkor zoom lens at 800mm focal length. Processed in Photoshop CC.