用HD38880语音合成芯片构成的汉语合成器

数字语音合成技术作为语音智能的一部分有着广泛的应用前景,它是利用语音的参数模型,从语音信号中提取若干个参数,存放在固定的内存中,再生时用语音合成芯片还原成语音的,我们的工作是将日语专用的语音合成芯片HD38880移植到汉语上来,获得了满意的效果。本文介绍的汉语合成器是用合成芯片 HD38880以及参数EPROM 等少量硬件制成,合成声音清晰自然,且具有体积小結构简单的特点,在仪器仪表语音智能化、计算机自动应答等方面都可以应用。汉语合成器中参数EPROM的制作直接影响合成声音的自然度。文中将对参数EPROM的研制过程,汉语合成器的硬件結构和软件流程作详细描述,同时对其应用实例“报时时钟”作简要介绍。

THE CHINESE SYNTHESIS DEVICE CONSISTS OF VOICE SYNTHESIZER HD38880

A digital speech synthesis technique, which plays an important role in speech intelligence, is being widely applied. The approach to synthesizing speech is to store the words in a digital memory using a few parameters, which are extracted from speech signals, and to retrieve them using a voice synthesis processor. Our work is a successful application of the voice synthesizer HD38880, which is specially designed for Japanese, into Chinese. In this paper, a Chinese synthesis device which consists of voice synthesizer HD38380, parametric EPROM and some other 1C, is introduced. The procedures for making parametric EPROM, which are influences on the quality of synthetic speech, are described. The device is small in size and simple in construction so that it is easy to use in digital instrument and alarm systems as well as in computer terminals, The synthetic Chinese speech produced from the device is highly intelligible and natural. Finally, the paper gives the construction of a talking clock as an example of the application of the device.