JPS621230B2 - - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION The present invention relates to an electronic timepiece system that announces time by voice.

Conventionally, audio time clocks using magnetic recording media have been known, but they require a recording media, its mechanical drive, a magnetic head, a large-capacity power supply, etc., and although relatively large clocks are possible, However, it could not possibly be made into a small pocket watch or wristwatch.
However, in recent years, audio processing technology has become extremely advanced, converting audio into digital data, permanently storing it in IC memory, and calling it out one after another under the control of a CPU (Central Processing Unit), decoding it into an analog waveform and playing it through a speaker. This has become possible with just two or three IC chips and a small amount of external circuitry, and English word pronunciation learning devices using this technology have also appeared. If more integrated circuits become C-MOS in the future,
Since the problems of speakers, external circuits, and power supplies can be solved, it is expected that watch-sized audio devices will emerge. In view of this situation, the present invention aims to provide a portable voice clock with built-in solidified audio data, which can be used by visually impaired users or general users in dark places or when they do not want to use their eyesight. The purpose is to use it as a clock in various situations, or to be incorporated into other audio equipment such as radios and tape recorders (so that the audio circuit can be shared), and its gist is numbers, time units, day of the week names, month names, etc. voice storage means for storing digital data representing the audio waveform of each word necessary for reporting for a plurality of Japanese languages; a switch means for selecting the type of Japanese language; and a storage means for storing digital information corresponding to the selected type of Japanese language. a Japanese language type storage means, a voice control switch means, a clock device that counts reference signals to generate time information, a decoding circuit for digital data of the voice waveform, and a sound generation means connected to the output of the decoding circuit. , when an operation signal of the voice control switch is generated, digital data of a speech waveform in a predetermined Japanese language is extracted from the voice storage means in accordance with the output of the Japanese language type storage means and the time information produced by the timekeeping device, and is arranged in chronological order. The audio clock is characterized in that it has circuit means for supplying an input signal to the decoding circuit, and announces the time by voice.
Embodiments of the present invention will be described below based on the drawings.

Figure 1 is a block diagram of the clock system of the present invention. 1 is a crystal oscillator that serves as a timekeeping standard, 2 is a frequency divider that generates a second signal for timekeeping, and 3 to 9 are used to count seconds and minutes. , a group of counters connected almost in series to create time information (time data) such as hour, day, day, month, year, etc. year, month counter 8,
The status of 9 is fed back to day counter 7,
Although it is desirable to control the full count and automatically correct the size of the moon, the feedback circuit for this is not shown. Reference numeral 10 denotes an input circuit that generates a signal for controlling the CPU, which will be described later, from a group of manually operated switches S ₁ to _{S 5.} In this example, S ₁ is a push button switch for voicing the time and a push button switch for voicing the calendar. If you press the button once, it will say the time (hours, minutes, seconds), and if you press it twice in quick succession, it will say the calendar (month, sunday, year) in sequence. Discrimination between the two operations is made based on whether or not an operation signal is received again during the operation of a short-time timer that operates every time a switch is operated in the input circuit 10, and each generates a different signal. Since this button is the most important operation button, it is best to make it particularly large and differentiate it from other operation members, and to place it on the front of the watch. _S2 is a switch for selecting and specifying the Japanese language to be generated in advance, and advances a ring counter in the input circuit 10, which has the same number of stages as the number of Japanese languages in the audio data stored in the memory, and changes the state of the counter. One of the national languages is selected by the , and a status signal (so-called bit flag) corresponding to that national language is generated. The embodiment uses two languages, Japanese and English. S ₃
is a switch for interrupting voice production at any time during the program. When this switch is pressed, the program counter in the CPU is forcibly reset even from the state in the middle of the voice program, and the program returns to the standby state it was in before the voice button _S1 was pressed. I'll go back. This switch is preferably used when you already know the required time at the beginning of the utterance and don't want to hear what comes after. Since the time required for one utterance is at most several seconds, this switch may not be necessary. _S4 , _S5
is used when adjusting the time, _S4 is a switch that selects the timekeeping element to be corrected, and _S5 is a switch that adds (or subtracts) a unit count only to the selected timekeeping element each time it is operated. However, the first position is reset only when the selected time measurement element is a second counter). In the time adjustment operation, the CPU 13 controls the input/output gate 11 using signals generated from both switches according to a time adjustment program to be described later, so that the addition signal reaches only the necessary time measurement elements.
12 is digital amplitude data obtained by sampling the audio waveform at high frequency (4.5KHz to 30KHz) and directly A-D converting the amplitude, or by sampling at large time intervals (for example, several tens of milliseconds), and using LPC (Linear Prediction Code). Speech parameter data compressed to the required number of bits using the technology
ROMized information, the same ROM of the voice activation program described later, the ROM of the operation control program of the linear predictive decoding circuit 14 (not needed when the audio data is uncompressed) described later, and information before or during calculation IC consisting of RAM etc. for storage
15 is a DA converter which converts the digital output of the audio waveform memory or the audio waveform parameters into audio waveform data by linear predictive coding technology into an analog waveform. The output is strengthened by an audio amplification circuit 16 as necessary to drive a speaker 17. The circuits 14 and 15 will be collectively referred to as an audio decoding circuit, and the circuit 16 and the speaker 17 will be collectively referred to as a voice generating means.
Linear predictive coding technology inputs the sound source, white noise (in the case of unvoiced sounds) or periodic pulse signals (in the case of voiced sounds), to a multi-stage lattice-type digital filter, and calculates the difference between the data passing through the filter, or between the data and the coefficient parameters. This method obtains a modulated digital output by performing addition and subtraction many times, and an audio waveform is obtained by DA converting the digital audio data of this output. In other words, it can be said that the process in which the input of vocal fold vibration is modulated in the vocal tract is digitally imitated. Audio data is stored in the form of frequencies and filter coefficients, depending on whether the input is white noise or periodic, and each data is sent to a filter (the main part of the decoder) at relatively long intervals, and is stored in the form of filter coefficients. Calculated interpolation operations are also preferably performed so that sudden changes over time are not noticeable. The timing of calculations within the filter is determined by the central processing unit (CPU).

FIG. 2 is a flowchart showing an overview of a program stored in the ROM 12 that operates the CPU 13. The program starts running along the a column when either S ₁ or S ₄ (assumed to be a voice switch) is pressed. If it is the correction button _S4 that is pressed, the process branches to column d and below. Further, if a calendar voice signal is output, the process branches to column c. In the subsequent operation, the CPU 13 first controls the select gate group 11 to send the current time data to the memory 1 in a predetermined order.
Read the 1st and 10th digits into the RAM section of 2 or the data stack in the CPU 13 (there are blocks such as the day of the month that do not require this), and use the read numeric values as addresses to read the corresponding audio data. ROM the group
12 and input to audio decoding circuits 14 and 15. At this time, the program is branched depending on the selected state of the Japanese language, and a different ROM section is of course addressed. The corresponding words in each language are
They are arranged in different parts of the ROM in approximately the same order for each national language, and are addressed by the time data read in each language. Furthermore, when entering the correction state (columns d, e,), the count value is also changed along with the utterance.
As shown in the lower half of column a, if you keep pressing the utterance button, the seconds part will be read repeatedly and utterances will be made every second. The day of the week is automatically corrected from the year and month data. When you press the utterance button, the CPU captures all the time data all at once, fixes it, and then converts it into a series of sounds, but in this program, each time data is input immediately before the utterance of each stage of hours, minutes, seconds, etc. The time data of each stage is read, and each stage has its own special characteristics. That is, in the former method, there is no contradiction in the time signal voice, but the time of utterance is based on data several seconds in the past. (You can also set the clock slightly ahead so that the end of the utterance is almost on the hour.)
In the latter case, if you perform an operation that causes the vocalization period to span the transition between minutes, hours, and days, discrepancies will occur when reporting.
The program to output seconds intermittently is simplified. At the top stage of the d column, the counter in the first place of seconds is reset to minus 1 second (the frequency divider 2 in the previous stage is reset to zero) when the second is uttered.
This is to accurately synchronize the timing with the end of the utterance, anticipating the second. A in FIG. 3 is a subroutine of the number utterance stage in the flowchart in FIG. 2, where a is for Japanese and b is for English radix. The latter takes into account the decimal system. In addition, English numerals are used as ordinal numbers (used for dates, etc., 1st, 2nd, etc.)
Although the subroutine for uttering is not shown,
It's not much different.

Next, a modification example of this embodiment will be described. First, we used a timekeeping counter similar to conventional electronic watches, but instead of using a timekeeping program and CPU to calculate the time,
The calculation results of each block may be stored in a RAM latch or stack memory (updated all the time) and used as time information. An optical display device may be used in the watch in parallel with the sound device. In addition to the time data in this example, as an application of audio display, function instructions and usage of operation buttons on multi-function watches with audio (for example, in the case of a chronograph, pronouncing "button A to start") or operation Speak a confirmation (such as “start” when pressing a button),
In the world clock, if you switch the display to another standard time with a time difference, the time will be displayed along with the place name every time you say something, such as "What hour and minute is it in Japan time?" or "What hour and minute is it in New York time?" Good too. This can be achieved by adding a place name utterance stage that uses the time difference as address information to accompany the time difference calculation program. Furthermore, you can enter a small memo such as a phone number obtained from another person and create a program to play it when necessary, or add the purpose of the alarm as an additional device for an acoustic alarm (sounding an electronic buzzer, etc.). When you recycle a watch, it becomes even more closely connected to your daily life. It is also possible to switch between male and female voices depending on the user's preference. It is also extremely preferable to confirm the number aloud each time a number key is operated in a computer watch in order to prevent erroneous operations. Watches that are organized around generation functions like this eliminate or do not place emphasis on the optical display, so even if they have multiple functions, the optical display does not become unnecessarily complicated and difficult to see, and the front or It is possible to create a clock with a simple and sleek design centered on the speaker provided on the side.

As described above, the present invention makes it possible to realize a portable audio timepiece with advantages such as a multilingual display and the ability to adjust the time while listening to vocalizations, and its range of applications is extremely wide.

[Brief explanation of the drawing]

FIG. 1 is an electronic system diagram of an embodiment of the present invention.
Figures 2a, b, c, d, and e are flowcharts of the operating program, and Figures 3a and b are flowcharts of the numeric utterance subroutine. 10... Switch input circuit, 11... Time information input/output gate, 12... Audio data, program storage device, 13... Central processing unit, 14, 15... Audio decoding circuit, 16, 17... Voicing means.

Claims (1)

[Claims] 1. Audio storage means for storing digital data representing the audio waveform of each word necessary for reporting, such as numbers, time units, day of the week names, month names, etc., for multiple national languages, a switch means for selecting the type of Japanese language, and a selection device. Japanese language type storage means for storing digital information corresponding to the type of Japanese language that has been given, a voice control switch means, a clock device that counts reference signals to generate time information, and a decoding circuit for digital data of the voice waveform; A sounding means connected to the output of the decoding circuit, when an operation signal of the sounding control switch is generated, outputs a predetermined Japanese language from the voice storage means in accordance with the output of the Japanese language type storage means and the time information produced by the timekeeping device. 1. An audio clock comprising circuit means for extracting digital data of an audio waveform in chronological order and supplying the digital data to the decoding circuit as an input signal in chronological order, and reporting the time by audio.