Yanny or Laurel? The debate explained.

MAYA SALAM and DANIEL VICTOR

New York Times News Service,

updated at 4:32 PM

Three years ago, the internet melted down over the color of a dress. Now an audio file has friends, family members and office mates questioning one another’s hearing, and their own.

Is the robot voice saying “Yanny” or “Laurel”?

The sound jumped to prominence when Roland Szabo, an 18-year-old high school student in Lawrenceville, Georgia, posted it on Reddit.

On Wednesday, Szabo said that he was working some time ago on a school project and recorded the voice from a vocabulary website playing through the speakers on his computer. People in the room disagreed about what they were hearing. He said a friend of his first posted it on Instagram. (Wired magazine spoke to two other teenagers who also claimed credit.)

One detail may frustrate some and vindicate others: Szabo said he found the original clip on the vocabulary.com page for “laurel,” the word for a wreath worn on the head, “usually a symbol of victory.”

Sharing of the clip really took off after a Twitter post from a self-described YouTube “influencer” named Cloe Feldman.

On Tuesday evening, Feldman said in a video that she was fielding multiple interview requests and searching for the original creator.

“I did not create Yanny vs. Laurel,” she said. “I don’t know how this was made.”

More than one person online yearned for that simpler time in 2015, when no one could decide whether the mother of the bride wore white and gold or blue and black. It was a social media frenzy in which internet trends and traffic on the topic spiked so high that Wikipedia itself now has a simple entry, “The dress.”

It didn’t take long for the auditory illusion to be referred to as “black magic.” Many audio and hearing experts have weighed in.

Jody Kreiman, a principal investigator at the voice perception laboratory at the University of California, Los Angeles, helpfully guessed that “the acoustic patterns for the utterance are midway between those for the two words.”

“The energy concentrations for Ya are similar to those for La,” she said. “N is similar to r; I is close to l.”

Patricia Keating, a linguistics professor and the director of the phonetics lab at UCLA, said: “It depends on what part (what frequency range) of the signal you attend to.”

“I have no idea why some listeners attend more to the lower frequency range while others attend more to the higher frequency range,” she added. “Age? How much time they spend talking on the phone?”

Kreiman cautioned that more analysis would be required to sort out the discrepancy. That did not stop online sleuths from trying to find the answer by manipulating the bass, pitch or volume.

Some speculated, like Keating, that the differences might be related to hearing loss or the age of the listener. It is known that some sounds are audible only to people younger than 25.

“If you turn the volume very low, there will be practically no bass and you will hear Yanny,” a Reddit user wrote confidently.

Yet making those adjustments did not change the word for some.

“I literally just turned all frequencies below 1khz to negative 70 decibels and I still hear ‘laurel,’” someone said on Reddit.

In the New York Times newsroom, there seemed to be no pattern among those who heard “Yanny” or “Laurel,” although a few heard faint traces of both, and some manipulation of the bass allowed those who heard “Yanny” to hear “Laurel.”

With time, a definitive scientific explanation will probably surface, like the one given for the dress, which had much to do with lighting. Until then, baffle your friends and astound your enemies, until the next random internet phenomenon has you doubting your own senses.

Close

Get news as it happens. Sign up for Boston.com’s email news alerts.

Thanks for signing up!



Source link

Leave a Reply

Your email address will not be published. Required fields are marked *