Adhering to proper Internet protocol, the world of online surfers flipped their tops and became obsessed with a new magnificent mystery, an audio clip of a single word. What word? Well, supposedly it was Laurel… or was it Yanny? Everyone from conspiracy theorists to arm-chair quarterbacks weighed in, but it was the audio experts who gleefully separated the chaos from the science.
Each word we speak has unique timbre and envelope properties. Some words like "stick" sound short and sibilant, while others like "umbrella" sound long and full. Words that have the same number of syllables and similar vowels lose obvious distinction, though. It's not at all strange to accept that many of the words we say sound alike—that easily accounts for many of those times you misheard a word in a song or conversation. Natural occurrences such as nasal congestion change the intelligibility of spoken words. For example, Bob and Mom aren't that different when you're deep in a cold! When you modify the dynamics and frequency content of a vocalized word, it can shift further from recognition of its original state and being to take on similarities to other words. Such modifications can easily come from speakers, microphones, processors, and audio codecs because each has different tonal and dynamic response characteristics.
In the debate of Laurel vs Yanny, the original word (Laurel) was played through a speaker and recorded through a microphone—two stages of potential manipulation right away. The source audio was a website that utilizes data compression for optimized streaming. That's another stage of change. The word "Laurel" starts to sound more like "Yanny" as its unique soundprint is warped. When viewed on a spectrogram in audio restoration software, "Laurel" and "Yanny" have similar patterns, though "Laurel" contains stronger low frequencies and "Yanny" contains more highs. Yet another factor is dialect; people in different regions pronounce and recognize words differently. Even without tonal tweaking, a word such as "car" spoken by a Bostonian will be less recognizable to a Californian.
Use the Laurel vs Yanny craze as a valuable lesson. If you want to maintain clarity and intelligibility for dialog and vocals, take care when recording, editing, mixing, mastering, and posting your audio. A well-recorded audiobook or podcast exported as a 64 kb/s MP3 can quickly take on warbles, garbles, and artifacts that can make listeners question what they're hearing. Whether you heard "Laurel" or "Yanny," life goes on and you've got work to do. So, get out there and do it well!
0 Comments