A team of researchers has developed a video reconstruction algorithm that transforms scattered very-long-baseline interferometry (VLBI) observations of black holes into continuous time-resolved movies, resolving a fundamental gap in how astronomers image the most dynamic objects in the universe.
The algorithm, called kine, uses neural fields to interpolate between sparse VLBI snapshots, capturing the temporal variability of supermassive black hole accretion and jet formation that static imaging methods miss entirely. Applied to existing Event Horizon Telescope data, kine could reveal how plasma flows around M87* and Sgr A* change over hours and days, showing the wobble of hot spots, the launch of jet components, and the twisting of magnetic field structures in real time.
That matters because the fastest known star orbiting a black hole already hints at the spin of Sgr A*, and dynamic imaging could confirm whether that spin drives the variability seen in flares. "Supermassive black hole accretion and the ejection of collimated, relativistic jets are intrinsically dynamic processes," the authors write in their Nature paper, noting that previous algorithms could only produce single-time images. The method arrives as new observatories like Vera C. Rubin begin delivering unprecedented cosmic imaging data, and could extend to any astronomical source observed interferometrically.