Difference between Mandarin Chinese and English speech masking caused by signal-driven processes

Speech intelligibility in competing speech is worse in native masking than in non-native masking, which could be caused by signal-driven processes and by knowledge-driven processes. To evaluate the contribution of signal-driven processes, the speech intelligibility of Mandarin Chinese was measured on native Chinese listeners in four types of maskers for both Mandarin and English masking: (1) 1-talker’s normal speech (NOR); (2) amplitude-modulated noise with speech envelope (AM); (3) speech-like signals synthesized according to the speech F0 modulation with steady temporal envelope (FM); (4) speech-like signals with speech envelope (a combination of AM and FM, AFM). The results show that speech reception threshold in Chinese masking was about 4 dB higher than that in English, in which about 2 dB was caused by the AM, −1 dB was caused by FM, and 1 dB was caused by AFM. The results revealed the signal-driven processes might play minor role in the difference of speech masking between Mandarin Chinese and Engl...