JPS6098495A - Voice synthesizer - Google Patents

Abstract

(57) [Summary] This bulletin contains application data before electronic filing, so abstract data is not recorded.

Description

DETAILED DESCRIPTION OF THE INVENTION The present invention relates to a speech synthesis device that receives external signals and performs speech synthesis.

Conventionally, speech synthesizers extract waveform information from A/D converted data, store the waveform information, and synthesize speech waveforms from the waveform information as needed. . At this time, there are roughly two types of methods for extracting waveform information.

One, as typified by PCMK, is a method of converting the amplitude value of the audio signal into a code of 1-1, which is called a waveform encoding method.

The other method is called the parameter editing method, in which parameters that characterize the waveform are extracted at regular time intervals, and the parameters are stored in a temporal sequence. One example is the Percall method. Regardless of whether this parameter editing method or waveform encoding method is used, the K to be stored as information necessary for speech synthesis is
Requires a large amount of storage capacity. Therefore, when a speech synthesizer synthesizes various voices, there is always a storage capacity problem. At the same time, it is very important whether this waveform information extraction can be easily performed in real time. If it is difficult to extract waveform information in real time, it is necessary to extract waveform information in advance for speech synthesis. Furthermore, if the voice is long, a large amount of storage capacity is required in proportion to the length of the voice, which has the disadvantage that it is difficult to use as a voice synthesizer and is difficult to use. Therefore, the next idea was to memorize speech waveform information in units of phonemes and phonemes in advance, and to sequentially arrange the phonemes and phoneme tubes according to the digits of the speech. This is an idea that people are clamoring to realize. According to this method, if all the phonemes and phonemes are memorized in advance, then by specifying the arrangement of these (phonemes), it is possible to respond to any vocal message. Especially in the case of a language like Japanese, which is easy to treat in phonological (phoneme) units,
This is a very powerful method, and the applications of speech synthesis devices will expand dramatically. By the way, even with this method, the point that phonemes (phonemes) must be memorized in advance is the same as the conventional method, and it has the disadvantage that messages other than those specified in advance cannot be produced.

The present invention has been made in order to completely eliminate the above-mentioned drawbacks, and is intended to realize a complete speech synthesis device that can generate any message without being concerned about storage capacity. Sara K
provides a new type of wireless receiver that receives external signals and synthesizes speech.

Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings.

FIG. 1 shows an embodiment of the present invention, in which 1 is a receiving circuit for receiving signals from the outside, a 2-digit demodulation circuit, a 3-digit storage circuit, and 4
5 is a control circuit, 5 is a speech synthesis circuit, and 6 is a speaker.

To explain this operation, first, the receiving circuit 1 receives a signal from the outside. At this time, the external signal is electrical,
Possible causes include magnetism, light, and sound. This signal is a signal specifying a phoneme (or phoneme), and the signal received by the receiving circuit 1 is demodulated by the demodulating circuit 2. The memory circuit 3 stores speech parameters necessary for synthesizing speech in units of all phonemes (or phonemes). Here, various cases can be considered as the voice parameter, such as a percall related i%PcM code. The control circuit 4 sequentially extracts phonemes (
Alternatively, the speech parameters necessary for speech synthesis in units of phonemes are selected and sent to the speech synthesis circuit 5. The speech synthesis circuit b synthesizes a speech synthesis signal sequentially in units of phonemes (or phonemes) based on the speech parameters sent from the control circuit 4, and outputs the synthesized signal to the speaker 6. At this time, it goes without saying that the voice parameters stored in the storage circuit 3 must match the voice synthesis method of the voice synthesis circuit 5. In the above process, the memory circuit 3 stores all the phonemes (or phonemes) necessary to synthesize speech, but the rest can be done without being aware of the memory capacity required for speech synthesis. It is possible to realize a speech synthesis device that can synthesize messages using various voices. Furthermore, based on this embodiment, it is possible to realize an unprecedented wireless receiver that receives signals from the outside and performs speech synthesis.

Next, an embodiment of a watch to which the present invention is fully applied will be described.

FIG. 2 shows an embodiment in which the present invention is incorporated into a watch, where 21 is a receiving circuit, 22 is a demodulating circuit, 5 is a phoneme (phoneme) voice parameter storage circuit, Sae is a control circuit, 5 is a voice synthesis circuit, 26 speakers, 27 a clock, 3 a voice message storage circuit, α a voice message reception command signal,
b is an alarm time notification signal.

To explain the operation of the circuit in this configuration, in the "voice message reception mode" provided in this watch,
When the voice message reception command signal is output from clock n, the control circuit array becomes ready for voice message reception. At this time, the receiving circuit 2 receives signals from the outside one after another and sends them to the demodulating circuits. This external signal is a signal specifying a phoneme (or phoneme). When the control circuit array receives a signal from the demodulation circuit 22 so that it can accept voice messages, it sequentially stores them in the voice message storage circuit. The signal from the demodulation circuit 22 is a phoneme (or phoneme)
Voice message storage circuit 2, which is specified by 'fr
8, codes for specifying phonemes (or phonemes) are stored in chronological order. In this way, voice messages can be stored in the "voice message receiving mode" of the watch.

The clock will then display the preset alarm time (
When the notification time is reached, an alarm time notification signal is output. When the control circuit Jun receives the alarm time notification signal, it outputs a phoneme (
Based on the specified time series of phonemes (or phonemes), the voice synthesis circuit 2 sequentially acquires the voice parameters necessary for voice synthesizing each phoneme from the phoneme (phoneme) voice parameter storage circuit B.
Send it out on 5th. The speech synthesis circuit 25 synthesizes a speech synthesis signal from the received speech parameters and sends it to the speaker 26. This allows the user of the p1 watch to listen to voice messages received from the outside at the preset alarm time.

Moreover, since this watch includes the configuration as a speech synthesizer shown in Figure 1, it can also be used as a wireless receiver by performing real-time speech synthesis from external signals. be. Furthermore, by making this clock take the form of a wristwatch, it is possible to realize an audio alarm, an audio schedule table, a wrist-sized portable radio receiver, etc.

As described above, according to the configuration of the present invention, it is possible to realize a speech synthesis device that can generate any voice message without being concerned about storage capacity when receiving signals from the outside.

Furthermore, it becomes possible to realize a new type of wireless receiver that receives signals from the outside and synthesizes speech.

[Brief explanation of the drawing]

FIG. 1 is a simplified embodiment of the invention. FIG. 2 shows a simple example in which the present invention is incorporated into a watch. l: Receiving circuit 2: Demodulating circuit 3: Memory circuit 4: Control circuit 5: Speech synthesis circuit 6: Speaker 2'': Receiving circuit 22: Demodulating circuit Z3: Phonological (phoneme) voice parameter 24 2 Control circuit 25 = Speech synthesis circuit G: Speaker 27: Clock 28: Voice message storage circuit α: Voice message reception command signal b: Alarm time notification signal and above Applicant Seiko Electronic Industries Co., Ltd. Agent Patent Attorney Mogami Tsumu

Claims (1)

[Claims] A receiving circuit that receives a signal from the outside, a demodulation circuit that demodulates the output of the receiving circuit, and a device that stores speech parameters necessary for synthesizing speech on a phoneme or phoneme basis. a first storage circuit that selects a predetermined audio parameter from the first storage circuit in accordance with the output signal of the demodulation circuit;
1. A speech synthesis device comprising: a speech synthesis circuit that synthesizes a speech signal using speech parameters of a storage circuit; and a speaker that outputs speech based on a signal from the speech synthesis circuit. 12. The speech synthesis device according to claim 1 has a timekeeping function with a time reporting function, and further includes a second storage circuit for sequentially storing the signals of the demodulation circuit. , 1° RUNushibζ MUMO
(3) A speech synthesis device characterized by having a speech synthesis function for synthesizing a speech signal according to the information stored in the second storage circuit. [3] A speech synthesis device characterized in that the speech synthesis device according to claim 2 is in the form of a wristwatch. 41. A speech synthesis device, characterized in that the speech synthesis device according to claim 2 is in the form of a portable electronic on-vehicle computer.