Abstract

Vibration within the vocal tract during speech is transmitted through body tissue to the skin surface and can be used to transmit speech. Achieving quality speech signals using skin vibration is desirable but problematic, primarily due to the different sources of sound during speech. The objective of this study was to characterize the frequency content of speech signals at various locations on the head and neck. Signals were recorded using accelerometers attached to 15 locations on the heads and necks of 14 males and 10 females as well as a microphone to record audible speech. The subjects produced several isolated phonemes and one phrase. The power spectral densities (PSDs) of the phonemes were used to determine a quality ranking for each location and sound. A spectrogram of the phrase was used to compare the response at selected locations. A perceptual listening test was conducted and compared to the PSD rankings. The PSD rankings were also calculated for signals recorded with background noise in order to identify locations that are least sensitive to external noise. With background noise, the frequency response of the skin was also used to study how the skin itself responds to external noise. The signal-to-noise ratio (SNR) was found for various sounds and locations with and without the presence of background noise. The frequency response of a concentrated area of the neck was also studied. Notably, while high frequency content was found to be attenuated at locations on the throat near the thyroid cartilage, it was detectable at some other locations. The best locations for speech transmission were found to be generally common to males and females in quiet environments but varied with background noise. During speech in the presence of background noise, the accelerometers performed better than the microphone when compared to the PSD of a clean microphone recording of the same sound. All SNR of all locations were influenced somewhat by external noise, some considerably more than others. Some neck locations may be better suited for contact microphone placement other than directly over the thyroid cartilage (where many commercial contact microphones are currently worn) if the neck is the preferred location for contact microphone placement.

Degree

MS

College and Department

Ira A. Fulton College of Engineering and Technology; Mechanical Engineering

Rights

http://lib.byu.edu/about/copyright/

Date Submitted

2009-05-28

Document Type

Thesis

Handle

http://hdl.lib.byu.edu/1877/etd2936

Keywords

contact microphones, throat microphones, skin vibration, frequency response, noise

Share

COinS