A noisy train station at rush hour. Thousands of people hurry past, voices merging into a dull roar. Somewhere in the crowd, two people try to shake hands with a special, barely visible gesture. Spotting such a handshake is nearly impossible. That's the kind of challenge physicists faced at the Large Hadron Collider when they set out to search for new particles in the mass range from 20 to 60 GeV (for comparison, the proton mass is about 1 GeV).

This range used to be a 'blind spot': signals from light particles drown in the background like a whisper at a rock concert. To hear the whisper, the researchers built a special 'hearing aid'—an algorithm that picks out the tracks of tau leptons from the data mess. The tau lepton is a heavy cousin of the electron that lives only for an instant and decays into a shower of other particles. Catching such a shower, especially when the tau lepton is born almost at rest, is a task bordering on the fantastic.

The new approach worked: physicists not only heard the whisper but also made out the words. They managed to register the production of the well-known Upsilon mesons, which decay into pairs of tau leptons. The signal was so clean that its significance above the background reached 5.8 sigma: that's like flipping a coin and getting heads 20 times in a row. But no new, uninvited guests were found in this mass range.

The absence of a discovery is also a result. It maps out a boundary beyond which new particles cannot hide: their production followed by decay into two tau leptons must occur less often than 40–400 events per hundred billion collisions. Now we know where not to look, and we can focus on other hiding places. And the noisy train station has become a bit clearer, and this is just the beginning.