Laurel, Yanny, and Speech Analytics

To settle the debate we ran the viral audio file through the VoizTrail speech analytics platform

Last week the internet was set ablaze by the great “Laurel versus Yanny” debate. The audio file, which originated from the vocabulary.com site entry for “Laurel,” was believed to have first spread via Reddit before being picked up by the YouTube sensation Chloe Feldman, who shared the file on Twitter effectively sparking a viral debate. The debate continued to rage over the course of several days, moving across social media platforms before gaining traction on television, with honorable mentions on shows like The Daily Show and Ellen. Even new age singer/songwriter Yanni weighed in on the debate (casting his vote for “Yanni” of course). The clip provoked a slew of confusion and arguments over which was word was correct before its original origins were finally revealed.

The fact that so many people were divided by their impression of the audio ignited a variety of theories; from the conspiracy that there were two files circulating, one stating “Yanny” and the other “Laurel,” to the idea that people were hearing the word based upon their expectation of what it would be- literally hearing what they wanted. Eventually, science provided an explanation: the file is an “ambiguous figure” meaning the audio input can be organized in two ways. That is, what you hear when you listen to the audio file depends greatly on the frequency in which it is played. Higher frequencies sound more like “Yanny” and lower “Laurel.” This explains why some people could hear both based on what devices they listened with (headphones versus speakers, etc.). It is a matter of acoustics. This got us thinking. Our VoizTrail® Communication and Compliance Suite utilizes advanced speech analytics technology to process call audio. A major component of the processing is reliant upon acoustical data…

So what would VoizTrail® hear?

We ran the audio through the platform and… drumroll please… the results read “Laurel.” That’s right, VoizTrail’s advanced speech analytics technology translated the audio file with 100% accuracy, settling the debate and hearing it better than (in many instances) the human ear.  Sorry, Yanni.

How accurate is your call auditing? Discover how to put VoizTrail’s technology to work for your business by visiting our VoizTrialⓇ page today.