Imaging the Solar Corona Outside of an Eclipse

Viewing or imaging the solar corona outside of a total solar eclipse has been elusive. It was first achieved by Bernard Lyot in the 1930s when he invented the coronagraph. Since then, professional solar astronomers have built coronagraphs on mountaintops and in satellites. As far as we know, the first amateur to see and image the corona outside of an eclipse was Klaus Hartkorn of New York in 2020, using a homemade coronagraph atop Mount Mitchell, NC. I reached out to Klaus, and we soon became friends. Under his guidance, I converted my 30-year-old Baader prominence viewer into a coronagraph and first saw and imaged the corona in 2021 from Utsayantha Mountain, NY (see Astronomical League’s Reflector at

Since then, I built a new coronagraph from scratch with better performance, and here I report on its use for imaging. See also Solar Astronomy by Christian Viladrich with a subsection by Klaus on coronagraphs.

The corona has several components. The K-corona is what we mainly notice during an eclipse, and it is due to sunlight from the photosphere (the bright surface of the Sun that we normally see) being scattered by free electronics in the million-degree upper solar atmosphere. The E-corona is due to sunlight being absorbed and emitted in narrow wavelengths by those free electrons; it is up to 100 times dimmer than the K-corona. The F-corona is further out from the Sun, and it is due to sunlight scattering on interplanetary dust, a continuation of Zodiacal light.

The challenge of capturing the corona is that it is a million times dimmer than the Sun’s photosphere. In a coronagraph, you create an artificial eclipse by blocking the Sun with some occulting object in the light path. The hard part is that the brilliant Sun manages to sneak its way down the light path and obscure the corona. As sunlight travels through the air, the air itself scatters the light; at sea level, this effect alone is brighter than the corona near the Sun. Anything floating in the air like dust or pollen scatters light even more. Viewing from a mountaintop reduces the amount of air above the telescope and may reduce the dust and pollen. The season of the year has a large effect, too, based on the pollen. You can tell how the sky is by putting your finger in front of the Sun; the sky should look deep blue up to the blocked Sun.

Next the sunlight hits the objective lens of the telescope, which in addition to focusing light scatters light in all directions, diffracts light outside of the focused solar image, and reflects light outside of the solar image. Lyot’s genius was in addressing these sources of scattered light. To reduce the scatter from the objective lens, I use a 50mm f/20 uncoated singlet lens that is polished far smoother than typical astronomical lenses; it is made for high-powered lasers. A coating reduces reflection but increases scatter so much that you are better off without it. The objective lens in held in the body of an old Celestron 80mm FirstScope with its internal baffles removed.

Next comes a 9.2mm metal cone that produces the artificial eclipse. Mine came from my old Baader prominence viewer. The cone is held by a screw that is epoxied to a “field” lens. Between the cone and the field lens is a field stop, which is a 19mm aperture that helps keep the reflected sunlight from getting down the tube. The 38mm-diameter 100mm-focal-length field lens focuses an image of the objective lens to a point behind it. At this exact point follows the Lyot stop and Lyot spot. The stop is circular aperture 4.2mm in diameter that blocks sunlight that has been diffracted around the perimeter of the objective lens. The Lyot spot is a 0.4mm centered spot that blocks light that has been reflected within the objective lens. The spot is painted onto the next lens, a 25mm-diameter 125mm-focal-length “relay” lens that sits up against the Lyot stop. Next comes a second relay lens, 38mm diameter and 125mm focal length. The light coming out of the second relay lens is telecentric and optimal for a narrowband filter, which requires parallel light to stay on band. The internal walls of the system are painted dark black using Musou black paint from Japan, which scatters ten times less light than typical flat black paint. The lenses and stops are held in an inexpensive set of T2 optical tubes.

At this point, the photosphere has been attenuated to well less than one-thousandth. Yet this is still not enough to see the corona, whose brightest part is a few millionths of the brightness of the photosphere. The E-corona emits light in narrow bands and can be captured with a filter. My coronagraph uses a double-stacked narrowband filter with a bandpass of 1.2 Angstroms at 5303A from Andover Corporation; it is the most expensive component. The narrow bandwidth requires tight temperature control at 32C. I built an oven out of inexpensive parts: a T2 tube, a 1” tube, two thermistors, adhesive strip heaters, adapters, and a pair of controllers. The filter improves the signal-to-noise ratio by 600 times, effectively dropping the scattered light to a fraction of a millionth of the original photosphere and making the dim E-corona visible.

For a camera, I use a ZWO ASI1600MM Pro with a reducer lens in front. I use ASI Capture obtaining 30 seconds of 0.5 second exposures at 250 gain. I then use AstroSurface to stack the frames and Photoshop to remove artifacts.

The instrument needs to be well aligned so that the objective’s diffraction ring and central reflection project precisely onto the Lyot stop and spot, about 0.2mm accuracy. Despite the tolerances, I managed to build it using standard hand tools. It took me about a year build this coronagraph from scratch. Much of the time was spent debugging sources of scattered light using a test bench made from a bright LED and a cardboard box light trap to mimic the photosphere-corona contrast. Any dust on the objective destroys the coronal view. Even a single visible piece of dust can obliterate the corona. Much of my time is spent cleaning the lenses, especially the objective. It can take an hour to clean a lens with a series of solvents and wipes.

Using this instrument, I have imaged the corona from Utsayantha Mountain, Cherry Springs State Park, and even the Peconic Bay (sea level) on a very clear fall day. The main remaining challenge is that the 5303 filters create a low-intensity fringe pattern that has to be removed via image processing. I also use the coronagraph to image prominences in hydrogen beta, hydrogen gamma, helium D3, etc. using not-so-expensive filters.

While I have looked at the corona through my coronagraph, my description does not cover visual use. For visual use, you need to assess the brightness when the Sun peeks around the occulting cone and how you are blocking light from ultraviolet, visible, near infrared, and far infrared, as well as how you are double filtering light in case you accidentally leave one filter off. It would therefore require extensive study of eye safety and additional filters and design changes.