The morning space digest brought a brief note: the Chandra X-ray Observatory found "mysterious objects" in the Pinwheel Galaxy — faint in X-rays but suspiciously bright in ultraviolet. The phrasing "could solve two long-standing mysteries" sounded like typical press release fanfare, and that's exactly why it was worth checking what lay behind it. After digging in, I discovered a story far more interesting than "scientists baffled": this is a tale of how entire populations of cosmic objects spent half a century falling into a blind spot of observational astronomy — and how they were extracted from already existing data without spending a single second of new observation time.
The discovery in brief. In September, Mustafa Mujibulla's group from the University of Alabama published a paper in Nature Astronomy about a new class of sources — hypersoft X-ray sources (HSS). A total of 84 such objects were found in six galaxies, including Andromeda and the Pinwheel. Their peculiarity: they emit X-rays only at abnormally low energies — below 0.3 kiloelectronvolts, where most X-ray telescopes barely look — and their main energy output, apparently, falls entirely in the extreme ultraviolet. The brightest of them shine with a power of about 10³⁸ erg/s — millions of times brighter than the Sun. They're among the most energetic sources in their galaxies. And until recently we simply didn't know about them.
The Pinwheel (M101) isn't accidental in this story. It's one of the most studied spiral galaxies in the sky: bright, facing us "head-on," at a distance of about 21 million light-years. That's precisely why Chandra has imaged it dozens of times over all the mission years, and it's in the archive of these images that part of the new objects were found — on the composite image published with the paper, seven such sources are marked with circles. Further on, data from Andromeda (M31) and four elliptical galaxies were added.
A nice detail: HSS were found both in regions of active star formation and where old stellar populations live. So this isn't a side effect of some single environment, but apparently a ubiquitous population. 84 objects in six randomly chosen galaxies — that's a hint that throughout the nearby Universe there are tens of thousands of them.
What exactly makes these objects strange? An ordinary X-ray binary — a neutron star or black hole pulling matter from a companion — shines mainly above 0.3 keV. HSS are the opposite: in the hard part of the spectrum they disappear, all their X-rays are supersoft. And supersoft X-rays on the electromagnetic spectrum border energetic ultraviolet, from which follows the conclusion the team made: the real luminosity of these objects is even higher than measured, and most of it pours out into extreme UV.
Here it's worth honestly saying: ultraviolet radiation in this work is not a measurement but a conclusion. Nobody photographed HSS in the EUV range; the team reasons from spectral models and position on the scale. This is normal for astronomy, but the boundary between "found" and "postulated" here runs exactly along this line, and NASA's press release sometimes blurs it.
So why hasn't anyone seen such a bright class before? There are two reasons, and they reinforce each other. Soft X-rays are hard to detect — Chandra needed long exposures to recognize these faint sources. And energetic ultraviolet, where HSS shine brightest, physically doesn't reach our telescopes: hydrogen and helium in the interstellar medium absorb it, forming what the authors call an "almost impenetrable barrier." A double wall — absorption from outside, weak signal from inside. Each individual survey failed both filters.
Astronomers have roughly known the family of "soft" sources since the 1990s: ROSAT back in 1991 found supersoft sources in the Large Magellanic Cloud, and in 2003 Rosanne Di Stefano and Albert Kong counted 149 "very soft" objects in four galaxies, 72 of which fell into a separate quasisoft category with temperature below 300 eV. Back then these sources even had a moment as candidates for intermediate-mass black holes, after which the excitement died down.
What changed now? HSS are the next step down the same ladder toward lower energy: detection below 0.3 keV, luminosities up to ~10³⁸ erg/s, and crucially — a systematic conclusion that the source reaches its emission peak already beyond the X-ray range, in EUV. Notably, Rosanne Di Stefano is a co-author of both that 2003 work and the current one. Someone methodically dug the same section of spectral backwater for twenty years, and sooner or later they had to dig up something.
The first is prophetically old. Type Ia supernovae serve as "standard candles" by which we measure the Universe's expansion; it was on them that accelerating expansion was discovered in 1998, earning a Nobel Prize. Yet no one has reliably seen a Type Ia supernova progenitor: we study explosions after the fact and still argue about what exactly explodes — a white dwarf pulling matter from an ordinary star, or a merging pair of two dwarfs. A 2024 review honestly calls this "a long-standing puzzle." The HSS team suggests that some hypersoft sources are accreting white dwarfs and post-nova systems, meaning potential progenitor candidates. If this idea gathers confirmation, astronomers will for the first time have a way to find future Type Ia supernovae before they explode.
The second mystery is ionization. Spectra of star-forming galaxies show lines of ionized helium (HeII), requiring photons with energy above 54 electronvolts. Ordinary hot stars can't provide such photons, and where to get the missing ones is a question that since the mid-2020s has been directly called the "ionization budget crisis." A 2024 work already showed: supersoft sources with temperatures of 10–100 eV precisely close this gap in magnitude. Hypersoft sources, with their extreme UV, fit this mechanism perfectly — and their undetected population means that the ionization balance of galaxies has been using an underestimated figure all along.
A separate irony of the story: no new telescope was needed for this discovery. The data had lain for years in Chandra's open archive, and all the work came down to methodically sifting through it at the lowest-energy edge of the detector. "We removed what was previously a blind spot of the telescopes," Di Stefano said in the press release. A picky reader will ask, isn't this an artifact: Chandra itself in its advanced age has acquired a layer of contamination on the filter that cuts sensitivity precisely in soft X-rays, and the team is still calibrating this degradation — a fresh 2026 paper is devoted to exactly this. But proper calibration is part of the method, not its undermining; it's what allows trusting the conclusion that the sources are genuinely soft, not "dimmed" by the instrument.
Others will be able to check the authors relatively quickly. If HSS are accreting binaries, they should show variability; if they're future supernovae — they'll evolve. And the main final touch is rather philosophical: periodically there's talk that "the low-hanging sky has already been picked" and fundamental discoveries in observational astronomy are over. And here are 84 objects, each millions of times brighter than the Sun, that lay in a publicly accessible archive for a quarter century waiting for someone to look at the data from a different angle. The sky is not exhausted. The familiar viewing angles are exhausted.