Pulsar PSR J0614-3329 spins 318 times per second. A point on its equator moves at about 10% of the speed of light. From a distance, the star looks like a lighthouse spinning so fast that its beam blurs into a continuous ring. But the NICER telescope aboard the International Space Station catches individual X-ray flashes and, from their pattern, reconstructs the shape of the invisible object.

Hot spots on the star's surface are regions heated to millions of degrees by streams of particles. When a spot turns toward us, we see a bright flash; when it turns away, the light fades. But light from the far side of the star doesn't just travel in a straight line: gravity bends its path like a lens. So a flash from a spot that has already slipped behind the horizon can still reach us. The more compact the star, the stronger this effect.

The scientists tested models with two circular spots, two oval spots, and three circular spots. The three-spot model explained the observations best: its support was tens of thousands of times higher than that of its competitors. For each photon, they traced its path through curved spacetime. The resulting radius is 9.88–12.77 kilometers with 68% confidence. This is smaller than some earlier estimates and shifts the allowed stiffness of matter toward the softer side.

Previously, the radius of this pulsar was known only approximately, and the equation of state for neutron matter allowed a wide range, from "spring" to "putty." Now the range has narrowed. Combining the new result with measurements of other pulsars, the maximum mass of a non-rotating neutron star drops from 2.20 to 2.15 solar masses. The radius of a typical star with a mass of 1.4 solar masses shrinks from 12.56 to 12.44 kilometers. Centimeters, it seems, but for nuclear physics this is a noticeable shift.

When a lighthouse flashes, a sailor recognizes it by its rhythm and estimates distance by its brightness. Here it's the opposite: the distance is known, and from the brightness and rhythm of the flashes, the size of the lighthouse is reconstructed. The more such lighthouses we can measure, the more accurate our map of superdense matter will be. And if the lighthouse turns out to be smaller than expected, we'll have to rethink what it's made of.