Scientists have generated the first reliable image of a black hole using data from the Event Horizon Telescope (EHT). Images of black holes were thought to be impossible since no light is emitted from them. However, when contrasted with the bright ring of illuminated gas that sometimes surrounds them, the silhouette of the event horizon pops out nicely. This particular black hole is called M87*, residing in the center of galaxy M87, and it’s 6.5 billion times more massive than our sun. The photo captured the world’s attention not only because it’s the first of its kind, but also due to the mystery that surrounds these objects.

Almost as impressive as the image is the method by which the team at the EHT produced the image. The technique is called very-long-baseline interferometry (VLBI), and it is as much computer science as it is physics. VLBI is a process whereby light is collected by multiple radio telescopes with a certain distance between them and synchronized by an atomic clock. The time-stamped data collection across the radio telescopes emulates an optical telescope with aperture equal to the maximum distance between any two radio telescopes in the array. That means the Event Horizon Telescope is not actually one telescope, but rather a series of radio telescopes organized around Earth at different points in North America, South America, Europe, Oceania, and Greenland – emulating an optical telescope with an aperture that is almost equal to the size of Earth. The distance between the telescopes in the array and the high level of precision that was used in synchronizing them gave the EHT 4,000 times the resolution capability of the famous Hubble Telescope.

But these radio telescopes are still only able to capture a very small fraction of the light that would be needed to form a picture. That’s why VLBI relies on imaging algorithms to take the small amount of data that is collected and turn it into an image of what the object “probably looks like.” Kate Bouman, the computer scientist who led the development of the computer algorithm that was used to image M87*, compared the process to reconstructing a song by only listening to a small amount of notes. To create the image of M87*, data was collected at each of the telescopes for 4 days and transported to a central location where it was compiled using the time stamps and run through Ms. Bouman’s algorithm. The resulting image reflects, to some level of statistical significance, what M87* looks like. Going forward, we should expect more and higher resolution images of M87* and other black holes. As more radio telescopes are added to the EHT’s array, more data will be available to generate the images. Indeed, the future of black hole imaging is bright.