Why Researchers Built the World's Most Expensive Karaoke Machine

The MRI study captured high-speed images of the muscles used to produce sound.

ByABC News
April 22, 2015, 12:33 PM

— -- Researchers at the Beckman Institute in Illinois have built the world’s most expensive karaoke machine in order to capture the movement of the muscles used in singing.

Belting out a tune involves more than 100 muscles in the chest, neck, jaw, tongue and lips, researchers noted. Using a super-quick imaging technique, Aaron Johnson, a faculty member who moonlights as a chorus singer, lay inside an MRI scanner and sang into a noise-canceling microphone as the machine snapped pictures at over 100 frames per seconds.

“The words were projected onto a screen at the top of the machine as the machine banged loudly in the background,” said Brad Sutton, technical director of the lab who collaborated on the study published in this month’s Magnetic Resonance in Medicine.

Typical imaging techniques only take about 10 frames per second, which isn’t fast enough to catch the complex interaction between the tip of the tongue and a fleshy flap of tissue in the back of the throat known as the velum, Sutton explained.

“What we demonstrate with this technique is how various muscles are involved in different speech samples and how we transition from one type of sound to the next,” Sutton said.

The team previously recorded a Johnson singing a Christmas carol to study vocal coordination. This time around they chose the song “If I Only Had a Brain” from the Wizard of Oz because it was less seasonal, Johnson said.

“Even though we weren’t studying what happens in the brain during song, it’s obviously a very prominent feature on the video,” he said.

The study was more than just an overpriced karaoke night, the researchers stressed. It had an important practical purpose. Johnson, who studies speech in older people, said the information can be used to understand how people can keep their voice muscles in top condition as they age.

“We were able to measure in real time the different speeds of the various muscles used in speech,” he said. “This can show us what we can do behaviorally to improve the voice later in life.”