摘要
LF微分声门波模型是分析嗓音源的一个重要的参考模型,它与实际的声门波形通过时域参数相对应,其分析所得的模型参数可用来合成声门激励信号。嗓音源合成是语音转换和语音合成的一个重要基础。在LF-4微分声门波模型的基础上,提出了一种LF-4模型的改进算法,进行嗓音源的合成。实验结果和分析表明,该算法可实现对嗓音源参数的灵活控制,使嗓音源的合成具有更高的融合度。
LF derivative glottal flow model is an important reference model in analyzing voice source, which is related to actual glottal waveforms via the timing parameters. Analysed parameters of the model can be used to synthesize glottal excitation signal The synthesis of voice source is an important foundation in speech conversion and speech synthesis. On the base of LF-4 derivative glottal flow, an improved algorithm of LF-4 derivative model is proposed to synthesize voice source. The experimental results shows that this improvement can be used to control the voice source parameters flexibly. It makes the synthesis of voice source take on higher degree of fusion.
出处
《电声技术》
2006年第5期53-57,共5页
Audio Engineering
关键词
微分声门波
变长度的平均幅度差函数
嗓音源合成
derivative glottal flow
LVAMDF（Length Varied Average Magnitude Difference Function）
synthesis of voice source
