TY - GEN
T1 - Effects of prosodic factors on spectral balance
T2 - 3rd International Conference on Speech Prosody, SP 2006
AU - Miao, Qi
AU - Niu, Xiaochuan
AU - Klabbers, Esther
AU - Van Santen, Jan
N1 - Publisher Copyright:
© 2006 Proceedings of the International Conference on Speech Prosody.
PY - 2006
Y1 - 2006
N2 - In natural speech, prosodic factors such as accent, stress, phrasal position and speaking style play important roles in controlling several acoustic features, including segmental duration, pitch, and spectral balance, i.e., the amplitude pattern across different frequency ranges of the power spectrum. To synthesize speech that sounds natural, these effects need to be accurately modeled. In this study we describe and evaluate a synthesis method that mimics the effects of prosodic factors on spectral balance. We measure spectral balance by using the energy in four broad frequency bands that correspond to formant frequency ranges. An additive model is used to capture the effects of prosodic factors on spectral balance. A new sinusoidal synthesis module is implemented under Festival to predict the target spectral balance value for each band from analysis results and apply it to the amplitude parameters of the sinusoidal model during synthesis. In this study we evaluate an important strength of this system, which is its ability to reduce spectral discontinuities in unit concatenation.
AB - In natural speech, prosodic factors such as accent, stress, phrasal position and speaking style play important roles in controlling several acoustic features, including segmental duration, pitch, and spectral balance, i.e., the amplitude pattern across different frequency ranges of the power spectrum. To synthesize speech that sounds natural, these effects need to be accurately modeled. In this study we describe and evaluate a synthesis method that mimics the effects of prosodic factors on spectral balance. We measure spectral balance by using the energy in four broad frequency bands that correspond to formant frequency ranges. An additive model is used to capture the effects of prosodic factors on spectral balance. A new sinusoidal synthesis module is implemented under Festival to predict the target spectral balance value for each band from analysis results and apply it to the amplitude parameters of the sinusoidal model during synthesis. In this study we evaluate an important strength of this system, which is its ability to reduce spectral discontinuities in unit concatenation.
UR - http://www.scopus.com/inward/record.url?scp=77954366438&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=77954366438&partnerID=8YFLogxK
M3 - Conference contribution
AN - SCOPUS:77954366438
T3 - Proceedings of the International Conference on Speech Prosody
BT - 3rd International Conference on Speech Prosody 2006
A2 - Hoffmann, R.
A2 - Mixdorff, H.
PB - International Speech Communications Association
Y2 - 2 May 2006 through 5 May 2006
ER -