Newswise — The human voice routinely produces sounds at frequencies above 5000 hertz, but it has long been assumed that this treble range, which includes high sounds such as cricket chirps, is superfluous to the understanding of human speech. Now researchers from the University of Arizona and the University of Utah are taking a second look at this underappreciated range. Their research reveals that people can glean a large amount of information, including the identification of familiar songs or phrases, from just the higher frequencies. The work, which the team will present at the 162nd Acoustical Society of America Meeting in San Diego, may prompt a re-evaluation of how much spectrum is in fact necessary to carry the full meaning of the spoken word.
Although current cell phone technology only transmits frequencies between 300 and 3400 hertz, musicians have known for decades that an unbalanced or cut-off treble range can ruin the quality of vocals at concerts. Brian Monson, now a research scientist at the National Center for Voice and Speech at the University of Utah, started out his career as a sound engineer and experienced first-hand that too much treble can make singing voices sound shrill, while too little treble can make them sound dull. “People are very sensitive to a change in the treble range,” Monson says. “They can tell immediately that something is different.” From previous research, Monson also knew that singing carries more energy in the treble range than normal speech.
Curious about whether the treble range was indeed more important for singing than for speech, Monson teamed up with researchers from the Department of Speech, Language, and Hearing Science at the University of Arizona to test what sort of information lies within the treble range for both spoken and sung words. The researchers recorded both male and female voices singing and speaking the words to the U.S. national anthem, and then removed all the frequencies below 5700 hertz. When volunteers listened to the high-frequency-only recordings, they were able to identify the sex of the voice, the familiar passages from the “Star-Spangled Banner,” and whether the voice was singing or speaking the words.
“This was definitely an unexpected result,” says Monson, who notes that the traditional view of speech would not have predicted that such specific information would be carried in frequencies above 5000 hertz. The results may help explain why it can be especially difficult to understand cell phone conversations in trains or at cocktail parties. In loud environments, the low-frequency range can become cluttered with noise and the higher frequency signals that might serve as a back-up in face-to-face communication are cut off by the cell phone transmission. The scientists’ studies may also have implications for the design of hearing-aids and may help guide the quest to produce synthesized speech that sounds natural. “This started off as a scientific study of the art form of singing, but the results can tell us surprising information about human communication in general,” Monson says.
The paper 5aSCb3, “Perceptually relevant information in energy above 5 kilohertz for speech and singing,” will be presented in a poster session Friday, Nov. 4, from 8-10 a.m.
# # #
Main meeting website: http://acousticalsociety.org/
Searchable index: http://asa.aip.org/asasearch.html
Hotel site: http://www.towncountry.com/index.cfm
Webcast registration and viewing: http://www.aipwebcasting.com
WORLD WIDE PRESS ROOM
In the week before the meeting, the ASA's World Wide Press Room (www.acoustics.org/press) will be updated with lay-language papers, which are 300-1200 word summaries of presentations written by scientists for a general audience and accompanied by photos, audio, and video.
The Acoustical Society will grant free registration to credentialed full-time journalists and professional freelance journalists working on assignment for major news outlets. If you are a reporter and would like to attend, contact Charles E. Blue (firstname.lastname@example.org, 301-209-3091), who can also help with setting up interviews and obtaining images, sound clips, or background information.
ABOUT THE ACOUSTICAL SOCIETY OF AMERICA
The Acoustical Society of America (ASA) is the premier international scientific society in acoustics devoted to the science and technology of sound. Its 7,000 members worldwide represent a broad spectrum of the study of acoustics. ASA publications include The Journal of the Acoustical Society of America (the world's leading journal on acoustics), Acoustics Today magazine, books, and standards on acoustics. The society also holds two major scientific meetings each year. For more information about ASA, visit our website at http://www.acousticalsociety.org