It only takes four syllables to identify the voice of a close friend or relative, according to a new study preformed by scientists. At least the excuse for not hearing your mom still holds up.

The study used forty-four Canadian-French speakers between the ages of eighteen and sixty-five and the close friends and family of those being tested. Each listening participant heard the recorded voice of their friend mixed in with similar voices—a sort of ‘voice line up’. The voices sounded similar based on pitch, accent, and other similar speech qualities. The speakers would say simple words and phrases, between one and eighteen syllables each. While there little to no ability to distinguish voices with short, one-syllable words, the rate of success sky rocketed at four. The phrase in this case that got the listeners to recognize the voice, in this case, was ‘Merci beaucoup,’ (Thank you very much) though any four-syllable word or phrase should do the trick.

The study’s data was extremely impressive: when the voice of a friend or family member was put into a mix of similarly sounding voices and said made than four syllables worth of speech, participants could distinguish their friend’s voice with a 99.9% accuracy. This rate of accuracy even beats out computers, including those built for vocal recognition: these have about a 92% when identifying voices.

Though the data is impressive, it is not necessarily shocking to audio scientists. From a young age, babies are able to distinguish the voice of their mothers from the voices of others, and this, then, carries over into childhood and adolescence. And now is proven to continue into late adulthood as well.

There are, of course, some exceptions: past studies have found that those with autism were less likely to successfully identify a familiar voice, and several other studies found that those with dyslexia found it harder to recognize voices in their own language than those without dyslexia.