Just a few days ago, I was lucky enough to witness, with my own eyes (safely protected by a proper solar filter!), one of the most fascinating astronomical events visible from Earth: a solar eclipse. Eclipses are not particularly rare on a global scale, but their visibility strongly depends on geographical location. The path of a total solar eclipse covers only a relatively narrow region of the Earth’s surface, meaning that totality may return to the same location only after a very long interval. Solar eclipses have fascinated humans for thousands of years and are among the astronomical phenomena most extensively recorded by ancient civilizations. Historical observations from Babylon, China, Europe, and the Arab world are so valuable that they have even been used to investigate changes in the Earth’s rotation over the past several thousand years.

On 12 August 2026, a total solar eclipse crossed parts of the Northern Hemisphere. From the UK, where I currently live, totality was not visible, but a substantial partial eclipse could still be observed. Having a basic refractor telescope and a good observation point, I decided to try recording the entire event as a time-lapse. I invested in a suitable solar filter for my Bresser Arcturus 60/700 mm telescope. The telescope has no automatic tracking, so this meant manually following the Sun throughout the roughly two-hour event. Afterwards, I planned to use Python, computer vision, and ChatGPT to align and filter the recorded frames and reconstruct a smooth eclipse sequence.

But my interest in this event was not limited to photographic observation. I was also curious to extract quantitative information from the recorded frames and to investigate what was happening to the environment while the Moon progressively obscured the Sun. Changes in incoming solar radiation should affect quantities such as light intensity, UV radiation, and, potentially, temperature and humidity. I already had several Arduino boards and a collection of inexpensive sensors at home, so this seemed like a perfect opportunity to turn the eclipse observation into a broader experiment. The question behind the project therefore became: Could a simple telescope, a few low-cost sensors, and some Python code reveal not only the progress of a solar eclipse, but also how the environment responds to it? The answer was YES: it is possible, and it was pretty simple to set it up!

I have published an article on my Instructables website in which I outline the setup of my Environmental Solar Station (ESS), show what I measured during the eclipse, and suggest ideas for future development. In this post, I will instead give some information about the creation of the video on the YouTube channel embedded below.

The Observation

As mentioned in the introduction, I used a simple beginner-level refractor telescope to capture the eclipse photo.  I bought a suitable solar filter compatible with the 60mm tube and attached a smartphone adapter to an old iPhone 6. This adapter connected to the telescope’s 20mm (lowest magnification) objective.  For the best vantage point and longest observation time, I positioned the telescope on the upper floor of my house in front of a south-oriented window.The location provided a convenient spot to mount the ESS on the roof outside the window.  This allowed me to connect it to my laptop at a comfortable distance using a long USB cable.

For the image collection, I used Lapse It, a well-made time-lapse app that I had already used for a previous small project. Unfortunately, due to a moment of distraction, I recorded the entire sequence at the lowest resolution available: 640 × 480 pixels. Alas, sbagliando s’impara (“you learn from your mistakes”), as an Italian saying goes! Fortunately, this turned out not to be a major problem for the subsequent analysis. The resolution was sufficient to follow the progression of the eclipse and determine the relative positions of the Sun and Moon. Nevertheless, recording at a higher resolution would certainly have produced much better images and, in particular, might have allowed details such as sunspots to be resolved more clearly. As the Earth rotates, the Sun’s image doesn’t stay centred in the telescope.  Advanced telescopes have automatic tracking systems to keep the telescope pointed at the object, but I had to manually adjust the orientation to prevent the Sun from disappearing from view. I continued this until around 19:40 when the Sun’s position in the sky was setting too low and a branch of a tree in front of the house blocked its view.

Nevertheless, I had plenty of frames, and the resulting video showed the sun and moon moving around during the video and sometimes partially out of view. I needed to select frames with the sun/moon system in view and center the sun to follow the moon’s motion on the solar disc. An excellent problem that can be solved using computer vision. I used the assistance of ChatGPT. This helped me with code writing and debugging, reducing the time needed to obtain the different Python programs needed for the preparation of the video that I had in mind. The Python programme automatically identifies the good frames from the cut ones and also centres the sun’s circle in the image. This provided a sufficient number of frames. Still, there was some problem matching the data with the partial recording of the eclipse. Therefore, the missing frames at the end of the movie have been generated taking into account the predicted values from the Jet Propulsion Laboratory ephemerides of the Sun and Moon in the sky.

Fortunately, the ESS data logging functioned perfectly. It continued throughout the event until approximately 20:10. The image collection was repeated every 5 s for a total of 1400 frames, while the data was logged every second. The data logged several parameters, but the most relevant was the one from the phoresistor module, the UV sensor, and the temperature, that have been used to monitor the environmental changes occurring during the eclipse. These data have been synchronized with the video recording in the form of graphs and numerical data. This allowed us to quantitatively monitor the effect of the different phases of the eclipse on these environmental variables.

In preparing the video for my YouTube channel, I wanted to add a musical background that could evoke some of the emotion associated with such a remarkable celestial event. Since the time of Pythagoras, music has been associated with the ordered movements of the heavens through the ancient idea of the music of the spheres. Classical music therefore seemed a particularly appropriate companion to the astronomical simulation and to the images recorded during the eclipse. A solar eclipse is, in a sense, an entrance into darkness—but a darkness quite different from night. It is the shadow cast by our celestial companion, the Moon, temporarily interrupting the light of the Sun. The shadow advances, the landscape darkens, and then the light gradually re-emerges. For the Eclipse 2026 video, I wanted the music to reproduce something of this progression: the approach of the shadow, the moment of maximum obscuration, and finally the return of the light. The extraordinary transition from the third to the fourth movement of Ludwig van Beethoven’s Symphony No. 5 in C minor, Op. 67 seemed particularly fitting. Near the end of the third movement, the music becomes mysterious and expectant, almost suspended in darkness. Then, without interruption, it bursts into the triumphant C-major opening of the fourth movement. Darkness gives way to light. This musical progression seemed remarkably well suited to the sequence shown in the video of the partial solar eclipse observed from Lincoln on 12 August 2026, where the astronomical simulation was synchronized with the recorded images and the environmental observations made during the event.

The next challenge was to find a historical recording of Beethoven’s masterpiece that I could use for the soundtrack. Searching through the collections of the Internet Archive, I discovered a remarkable recording made on 14 November 1940 at New York’s renowned Liederkranz Hall, with Beethoven’s Fifth performed by the All-American Youth Orchestra, the remarkable ensemble of young musicians founded by Stokowski that same year, under the direction of the legendary conductor himself—who, also in 1940, famously collaborated with Walt Disney on Fantasia. The recording was originally made for Columbia and subsequently issued in 1941 as part of the Columbia Masterworks series on 78-rpm records. Particularly convenient for my project, the archived version presents the third and fourth movements together in a single audio file, preserving Beethoven’s uninterrupted transition from the Scherzo into the Finale.

You can enjoy the complete video using the embedded link below. I hope you enjoy watching it as much as I enjoyed putting this project together.

If you like it, please consider liking the blog post and the YouTube video—and feel free to share it with anyone who might enjoy this journey through astronomy, nature, and music!