MxWulf 🐺🏍

??? Lost Cosmonauts: A Whisper from the Void That Never Faded Long before Yuri Gagarin smiled down at the world from orbit, dark rumors were already crawling across Europe. They claimed the Soviet Union had tried more than once to push a human beyond the atmosphere — and that every attempt had gone wrong. Officially, nothing was said. Unofficially, people whispered. The legend was sparked by two amateur radio brothers in Turin. At the end of 1960 they claimed to have caught a faint, almost ghostly signal from space: SOS in Morse code. The sound was thin, distant, and slowly fading — as if its source were drifting farther and farther from Earth into infinity. From that single transmission grew the most chilling version of the story: somewhere beyond the atmosphere, a Soviet spacecraft was adrift with a living person aboard, already beyond any hope of rescue. More than sixty years have passed. The tale has gathered names, details, “proofs,” and rebuttals. Yet not one reliable confirmation has ever appeared. All that remains is that fading SOS — and the question the Cold War never quite let go: were they real, or merely the echo of fear reflecting off the stars?

The Ring Was Real: Independent Reanalysis Confirms the Famous Black Hole Image In April 2019, the Event Horizon Telescope (EHT) collaboration released the first-ever image of a black hole — a glowing orange ring surrounding a dark shadow at the center of the galaxy M87, about 55 million light-years away. It wasn’t a conventional photograph. It was a reconstruction built from sparse radio signals collected by telescopes scattered across the planet. Hot gas spiraling into the supermassive black hole shines in millimeter radio waves, while the black hole’s gravity bends light paths, forming a bright ring around the event horizon. The image became one of the strongest observational tests of Einstein’s general relativity in extreme gravity. Years later, doubt appeared. Independent teams, notably Japanese researchers led by Makoto Miyoshi, reprocessed the publicly available EHT data in 2022 and 2024. By changing processing settings and widening the field of view, the ring vanished. In its place they recovered an elongated bright spot and scattered clumps. Critics suggested the iconic ring might have been an artifact — the result of algorithms that were given a preferred size and shape from the start, plus a narrow search window that suppressed other possibilities. Why a Black Hole “Photo” Is Always a Reconstruction M87* is tiny on the sky: roughly 40 microarcseconds across. Observing it at the 1.3 mm wavelength used by the EHT would require a single telescope the size of Earth. Instead, astronomers link existing radio observatories on different continents (Chile, Mexico, Hawaii, Spain, and later others) into one virtual Earth-sized telescope. Each pair of antennas captures only a narrow slice of information — specific spatial frequencies. With so few telescopes, most of the information about the object’s appearance is missing. Atmospheric water vapor further distorts the signals. From the same sparse, noisy measurements it is mathematically possible to reconstruct a clean ring, an elongated blob, or streaks. Classical imaging algorithms start with a template (a fuzzy blob of the expected size) and iteratively adjust both the image and atmospheric corrections. Change the template size or the field of view and the result can shift dramatically. That sensitivity is exactly what the critics exploited. How the Australian Team Settled the Debate Researchers from CSIRO and ICRAR (Samuel Lai, Nithyanandan Thyagarajan, Ivy Wong, Foivos Diakogiannis and colleagues) developed a fully independent imaging pipeline. Their results appeared in *Nature*. The first key step was to abandon ordinary pairwise measurements and switch to **closure invariants**. By grouping the telescopes into closed triangles, atmospheric delays and instrumental gain errors cancel out mathematically. What remains is information about the object’s shape that is protected from the usual calibration uncertainties. They then fed these cleaned measurements into a neural network called **GenDIReCT**. It is a diffusion model: it starts from pure noise and gradually “reveals” an image, constrained at every step by the real radio data. Crucially, the network was trained mainly on ordinary everyday photographs (animals, vehicles, household objects from the CIFAR-10 dataset) plus simple geometric shapes — not on simulated black holes. The goal was to avoid teaching the model “black holes look like rings.” A second, purely mathematical Bayesian geometric method (no neural network) produced matching results. What the New Analysis Found - Across all eight observing days in 2017, 2018, and 2021, both methods recovered a compact ring with an average diameter of **42.2 ± 1.5 microarcseconds**. - The ring remains stable even when the field of view is expanded far beyond the point where earlier critical analyses saw it disappear. - The critics’ ring-free images are explained by incomplete or noisy data on certain days (for example, 5 April 2017) and by the temporary loss of a key telescope (the large LMT in Mexico on 25 April 2018). On the best days the ring is robust regardless of field-of-view size. The measured diameter implies a black-hole mass of **6.5 ± 0.5 billion solar masses**, consistent with stellar-dynamics estimates and ruling out the significantly lower mass previously inferred from gas kinematics. The ring is also asymmetric — brighter on the southern and southwestern side — as expected from the relativistic Doppler boost of plasma orbiting at near-light speed.