This study investigates the production of neutral(unemotional)and surprised(emotional)‘yes/no’questions in spoken Russian.The paper compares acoustic parameters across these two questions types in three-syllable one-word sentences with the terminal rise-fall tone.The study takes account of traditional intonation parameters as well as pitch event alignment data.The study demonstrates the existence of statistically significant differences between the following parameters of neutral and emotional questions:total word(=sentence)duration, pitch peak height, characteristics of the accented syllable(duration, pitch heights at the onset and offset of the stressed vowel, intervals of pitch movements).The study suggests the existence of language-specific acoustic cues in the expression of emotions.The study also shows that despite the existence of speaker-independent ways of expressing surprise, there is also a considerable degree of intra-speaker variability.The results have direct implications for speech processing and communication systems, natural language processing, speech synthesis and recognition.

View full abstract

This paper proposes an objective speech distortion measure as a substitute for human auditory systems.Simultaneous and temporal masking effects are introduced into this measure called auditory-oriented Spectral Distortion(ASD).We calculate the ASD using spectral components over masked thresholds in the same way as the Spectral Distortion(SD).We confirmed that the ASD is more compatible to subjective mean opinion score that represents distortions on auditory impression than the SD.We applied the ASD to optimize a noise reduction algorithm proposed by the authors, and confirmed that this optimized algorithm reduces noises appearing to the ear.ASD is sure to be an available guide to design noise reduction algorithms.

View full abstract

The upper frequency limit of a p-p type sound intensity probe is mostly determined by the sensitivity reduction of the probe.The sensitivity reduction is caused by obtaining the pressure from the average of the two pressure signals including the phase, and by obtaining the particle velocity from the finite difference of the two pressure signals.This is a systematic error suggesting a possibility of correction.In order to compensate or correct the sensitivity reduction, the direction of energy flow must be known.If a one-dimensional intensity probe is used, the probe direction must be varied to find the direction of the energy flow.If a three-dimensional intensity probe is used, the direction of the energy flow can be at least roughly estimated from a single measurement and then the sensitivity reduction can be corrected.This paper proposes a method of sensitivity reduction correction in the high frequency region for a three-dimensional(3-D)intensity probe.A theory of the method is described first and then numerical results that validate the theory will be given.This method works even for a probe with some amount of phase mismatch among microphones and for a sound field with reflections.

View full abstract