The impending digital-TV transition has a forgotten victim: the big bang. You can tune an analog set between broadcast channels and see static, part of which is energy left over from the hot primordial universe. This static is known as the cosmic microwave background radiation, and its discovery in the 1960s proved the big bang theory. But on a digital TV, the best you can do is "The Big Bang Theory".

Last week at the American Astronomical Society's meeting, astronomers announced the detection of a second type of radio static from the heavens, and although it may not come from an era quite as ancient as TV snow does, it may probe the period immediately afterward—an equally mysterious time when the first stars and black holes were lighting up. That is, if the signal proves real.

The team, led by Alan Kogut of the NASA Goddard Space Flight Center, took measurements with a radio antenna named ARCADE that dangled from a high-altitude balloon over eastern Texas in July 2006. At radio frequencies greater than 10 gigahertz the radio emission matched that of the microwave background, but at lower frequencies it was several times stronger. Such an excess first emerged in the late 1960s and was mapped in 1981 by Glyn Haslam of the Max Planck Institute for Radio Astronomy in Bonn, Germany, but few astronomers thought much of it until now. "Maybe we should have taken it seriously all this time," says David Spergel, a Princeton University astrophysicist.

The reason for their skepticism was that background measurements are notoriously tricky. A background, by definition, is what's left after you account for everything you already know. Astronomers measure the total radiation coming from the sky and subtract off the radiation generated by the instrument itself, by our own planet and civilization, and by known celestial bodies.

As anyone who has balanced a checkbook knows, the act of subtraction has to be done with the utmost precision or you may erroneously believe you have some leftover light (or cash). Many "discoveries" of background radiation evaporated as data became more precise, and the ones that endured faced a hard slog to prove themselves.

"Measurements of the extragalactic background radiation are always hard to get, because this signal is very faint and, as a result, its detection is strongly dependent on how well one can remove the sources of contamination," says Angelica De Oliveira-Costa, an expert on cosmic background observations at the Massachusetts Institute of Technology.

The surviving signals include not only the microwave background but also three others that peak at different wavelengths: infrared light, visible light, and x-rays.

ARCADE's detection of a cosmic radio background must now run the gauntlet, too. "It's a tough measurement," says Mark Devlin of the University of Pennsylvania, an astronomer who has done similar work. "Unfortunately, most of the things that could go wrong could cause an excess."

He says that although the team used an innovative antenna design and took pains to correct for noise, errors may still have occurred. The device used to calibrate the radio receiver was less precise at low frequencies, where the putative signal appeared; the antenna had very low resolution, some 12 degrees on the sky, which could create a spurious signal at low frequencies; and the instrument took only about two and a half hours of data and covered only a small portion of the sky. "I'd like to see it done one more time, at least," Devlin says.

Another astronomer with experience making background observations, Judd Bowman of the California Institute of Technology, concurs: "The authors have made a careful analysis, but at this early stage there may still be room for unanticipated effects to conspire."

Even if the signal is real, astrophysicist Doug Finkbeiner of Harvard University warns that it may not be a truly cosmic signal—that is, one that originates in the distant universe as opposed to our own Milky Way Galaxy. He says the team subtracted the galaxy's contribution to the signal using a highly idealized model. A slight infidelity in this model might mimic an extragalactic signal. Both galactic and extragalactic processes produce light in the same way—namely, by electrons spiraling around magnetic field lines—so the galactic subtraction is especially error-prone. "You're subtracting two things that are almost exactly the same across frequencies," Finkbeiner says.

If the signal is real, where might it come from? Kogut's team speculates it may be a curtain of light emitted by the earliest stars to form in the universe—so-called Population III stars. That would explain why the radio background does not match up with the infrared background, as it would if the radio sources were, like most celestial bodies, surrounded by dust. Dust is produced in stars and litters space only upon their demise, so the first stars lived in dust-free environs. But the team acknowledges that evidence is thin for this hypothesis. Astronomers have yet to do the calculations to verify whether these stars could indeed account for the emission.

"There's a lot of invoking of Population III sources for various observations that people make," says astronomer Michael Hauser of the Space Telescope Science Institute. "I think that's jumping to conclusions."

Another culprit might be active galactic nuclei (AGN), which are blazingly strong light sources powered by black holes. Kogut and his colleagues argue that there are not enough AGN to produce the signal they see. Others, though, are not ready to rule out this possibility.

"They're extrapolating from the known AGN sources, and there could be more," Spergel says. The infrared, visible and x-ray backgrounds turned out to come mostly from hidden populations of AGN.

Finkbeiner speculates the source may be electrons given off by dark matter in our galaxy or extraneous emission that accompanied the release of the microwave background in the primordial universe. Whatever the answer, astronomers are clearly relishing a brand new phenomenon to explore.

"This is a fascinating observation that will take some time to decipher properly," says Michael Vogeley of Drexel University in Philadelphia. Stay tuned, even if you have digital TV.