The first image of a black hole: how they did it
A telescope the size of the planet, and data too large to send over the internet.
4 min read · Published 27 August 2026
The short answer
The Event Horizon Telescope linked radio observatories across the globe into a single instrument using very long baseline interferometry, giving an effective aperture the size of Earth. Each site recorded data with atomic-clock timestamps onto physical drives, which were flown to central facilities and correlated. The M87 image was released in 2019 and the Sagittarius A* image in 2022.
Why an Earth-sized telescope was necessary
Resolution depends on aperture relative to wavelength. The target — the shadow of a black hole's event horizon — is extraordinarily small in angular terms, comparable to resolving an orange on the Moon.
No single telescope could do it. The technique used is very long baseline interferometry: multiple telescopes observe simultaneously, and combining their signals produces resolution equivalent to a dish as wide as the distance between them.
Spread across continents, that separation is roughly Earth's diameter — the largest baseline available without going to space.
The steps
What the process required
- Simultaneous observation. Telescopes across several continents observing the same target in the same window, requiring clear weather at all of them at once.
- Atomic clocks. Each site timestamped its data with hydrogen maser clocks so signals could be aligned to fractions of a nanosecond afterwards.
- Physical transport. Drives flown to correlation facilities.
- Correlation. Combining the recordings to reconstruct what a single Earth-sized dish would have measured.
- Image reconstruction. The array samples only a sparse fraction of the possible aperture, so an image must be reconstructed from incomplete data using algorithms that fill in gaps.
- Independent verification. Four separate teams worked with different methods and were kept apart until each had a result, specifically to prevent them converging on a shared expectation.
Why the blind analysis mattered
Reconstructing an image from sparse data involves choices, and the expected answer — a bright ring — was known in advance from theory. That is a textbook setup for unconsciously producing what you expect.
The response was to split into independent teams using different algorithms and prohibit communication until each had converged. When they compared, all four had produced a ring of similar size.
This is worth highlighting because it is exactly the discipline that astrology consistently lacks. The researchers knew what they hoped to see, recognised the bias that creates, and built a procedure to defeat it. The result is credible because of the precaution, not despite it — the same principle set out in is astrology a science.
What the image shows
Not the black hole, which emits nothing. The bright ring is superheated material in the accretion flow, and the dark centre is the shadow — the region where light paths are captured rather than reaching us.
The shadow is larger than the event horizon itself, by roughly a factor of two and a half, because gravity bends light paths around the hole. Its size and shape are predicted by general relativity, and the measurements matched.
The M87 black hole is around 6.5 billion solar masses. Sagittarius A*, at our own galactic centre, is roughly four million — far smaller but much closer, and harder to image because material orbits it in minutes rather than days, changing the picture during the observation.
Common questions
- How was the first black hole image taken?
- By linking radio telescopes across several continents using very long baseline interferometry, producing an effective aperture the size of Earth. Each site recorded data with atomic-clock timestamps onto drives that were physically flown to correlation facilities.
- What does the black hole image actually show?
- The bright ring is superheated material in the accretion flow. The dark centre is the shadow — the region where light paths are captured rather than reaching us. It is about two and a half times larger than the event horizon because gravity bends light around the hole.
- Why were four separate teams used?
- Because reconstructing an image from sparse data involves choices, and the expected result was known in advance. Four teams using different algorithms were kept apart until each converged, specifically to prevent them unconsciously producing the expected answer.
- Why was the data flown on aeroplanes?
- Each observing site recorded petabytes, which was impractical to transmit over networks. Drives were shipped physically, and data from the South Pole station had to wait for the Antarctic winter to end before flights could collect it.