The perception of Chinese speech sounds in masking noise and frequency distortion
Bao Zi-wei, Rongjue Wei, Y. T. Wei · The Journal of the Acoustical Society of America · 1979
The intelligibility tests of Chinese speech sounds were taken under five masking conditions, namely white noise, pink noise, speech noise, meaningful speech interference, and reverberation masking in auditorium, as well as under a quiet studio condition. To simulate the actual communication circumstances, the noise was introduced to input and output ends, respectively. The signal to noise ratios were varied with a fixed speech level about 80 dB at 1 m from the loudspeaker. In addition, the speech and noise were processed with high-pass, low-pass, or band-pass filtering except in reverberation. A set of simplified but rather sensitive word lists which were based on varying the initial consonants because of the higher sensitivity of initials to masking than finals were used. The effects of masking and frequency distortion on the perception of individual Chinese speech sounds will be presented in this report, with the conclusion that monosyllabic Chinese words or “characters” should be considered as an integral part of sound unit. The tones (or inflection of pitch) with patterns preserved are demonstrated by tape recordings.