JPS5950076B2 - audio output equipment - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION The present invention relates to an audio output device (for example, an audio alarm clock, an audio output type calculator, an audio alarm tester, etc.) that outputs the content to be outputted by audio, and particularly relates to an audio output device that outputs the content to be outputted by audio, and particularly relates to an audio output device (for example, an audio alarm clock, an audio output type calculator, an audio alarm tester, etc.) that has a simple configuration This is intended to improve the quality of voice notification of numerical information.

Generally, when numerical information is output as voice, the numerical value as voice can be constructed by combining several phonemes.

For example, 2534 "Nisengohyakusanjiyuyon" is "two", "sen", "go", "hyaku", "san", "
It can be constructed by combining seven phonemes: ``Jiyuu'' and ``Yon''. By storing a necessary number of basic phonemes in the memory in this way and extracting them in a desired order and performing speech synthesis, desired numerical information can be output as speech. However, as a result of our research, it has become clear that such simple combinations of basic phonemes can cause problems in some cases.

As an example, when outputting numerical information of 123,445,000 as voice, it was found that a certain period of silence is required immediately after 100,000,000 ``OK''.

If this silent section is not provided, in the case of the above example,
The sound between ``ok'' and 2 ``two'' sounds crowded,
It was found that not only was it unnatural and difficult to hear, but there was also a risk of mishearing. In such an example, a similar phenomenon occurs not only immediately after 100 million but also immediately after 10,000. Also, the hundreds place is 1, that is, "
When pronouncing ``hiyaku'', the tens place is 1, i.e. ``jiyuu''.
It has become clear that when pronouncing ``hiyaku'' and ``jiyu'', it is also necessary just before these words.

For example, between 6,120, 1,000 "sen" and 100, "hyaku",
A certain period of silence is also required between the 4,312 + ``jiyuu'' and the 100 ``hiyaku''. Also, the pronunciation of the decimal point “
A certain period of silence is required just before "chi", for example, 123.45 "Hyaku Niji Yu Santen Yongo"
A certain period of silence is required between the ``chi'' and ``san''.

The case explained above is an example of pronouncing words with place values such as 100 million, 10,000, 1000, 100, etc., but in the case of so-called stick reading in which place words are not uttered, the pronunciation of the digit separator and the decimal point are pronounced. A certain period of silence is also required immediately before the call. for example,
2,345,678 “Nikon Masan Yongo Kon Maroku Nanahachi” between “two” and “comma”, “go” and “comma”, and between “san” and “Ichini Santen Yongo” in 123.45 A certain period of silence is required between each "ten". The present invention significantly improves the quality of audio notification of numerical information with a simple configuration, and one embodiment of the present invention will be described below with reference to the drawings.

FIG. 1 is a main part block diagram showing this embodiment.

X is a numerical data storage register, X is a decimal point position storage register, and both are RAM (random access memory).
configured within.

0C is an output control circuit that extracts the contents of the X register in the order of audio output and outputs them to buffer B for one digit.

J, is... This is a digit determination circuit that determines the digit of the value output to buffer B based on the decimal point position stored in If it is in the 10,000's place, it is a signal S, and if it is in the first decimal place, it is a signal S. , second decimal place. If it is below, signal S is output, otherwise signal S is output. Similarly to J, J, determines the digit of the value output to buffer B based on the decimal point stored in X and the signal Sa indicating which digit of the x register 0C outputs to buffer B. It is something that... If it is in the hundreds place or + place, signal S.
Output. Further, J3 determines whether the content of buffer B is 1 or not, and if it is 1, the signal S is sent. Output. AG
is an AND gate, and the signal S. , S, are output together, the signal S is generated. Output to 0CG. CGd and CGp are code generating units, respectively, where CGd generates a code for words such as 100,000,000,000, 1,000, etc. corresponding to each digit, and CGp generates a code representing a silent section and outputs it to 0CG.

0CG is CGd, CGp according to the signals S,,S,,S,
, B to the CC in a predetermined order.

VCC is a voice synthesis section, which outputs voice corresponding to the code output from 0CG. CC is 0 in code conversion section
The start address of the audio corresponding to the CG output code is output to the address counter AC. VR is a memory that stores data that is the basis of audio, and AD is its address decoder. D/A is a digital-to-analog converter, and JE is for detecting that the END code stored in VR is output, and if detected, outputs a signal Se. SP is an audio output section. For example, if the contents of register X are 254325678 and the contents of register

When outputting the 100 million digit of this numerical value to buffer B, J outputs a signal S. 0CG first outputs the contents of buffer B to VCC by signal S3, and makes it utter "two".

When the generation of "2" is completed, the signal Se is received, and then CGd
Outputs the output code to VCC. At this time, CGd was 100 million yen.
Since the code for "Oku" is generated, VCC utters "Oku" using this code. After the utterance ends, the signal Se
Then, the output of CGp, that is, the code of the silent section is outputted to VCC. VCC receives this code and remains silent for a certain period of time. By this operation, a silent section can be provided immediately after the 100 million "ok". Next, 0C outputs the tens of millions to buffer B.

In this case, J, outputs a signal S, and 0CG first outputs the contents of B to VCC using this signal, causing it to utter "Go", and then CGd outputs "Sen". Output the code to VCC and make ``Sen'' utter. In a similar operation, 0C outputs the millionth place to buffer B, and "Yon" and "Hiyaku" are uttered. Thereafter, audio is output in the same way, and when 0C outputs the contents of the tens of thousands place of register X to buffer B, J,
is signal S.

Output. The output control unit 0CG is the same as the billions digit, the contents of 1 buffer B, the output code of 2CGd, and the output code of 3CG.
Output to VCC in the order of the output code of p. As a result, a certain period of silence is provided immediately after the word ``man''. If the content of register x is 3245 and the content of X is 2, then 32.45 is stored by X and X.

In this case, the J beam signal S, is output for the tens digit, so
Since 0CG is output to VCC in the order of the output code of 2CGd, the contents of 1 buffer B, "san" and "jiyuu" are output as voices, and when the next 1's digit is input to B, J1 outputs signal S1. 0CG is 1 CG content of B
The output codes are output to the CC in the order of the output codes of d. But CGd
Because the 1's digit does not generate a code for digits such as 100 million, 10,000, 1000, etc., ``2'' is simply pronounced. Next, when the first decimal place is input to buffer B, J1 outputs signal S2. 0
CG output code of 1CGp, 2CGd by this signal.
output code, and the contents of 3 buffer B are output to CC in this order.

This creates a silent section before the "ten" in "ten" and "yon." Next, the contents of the second decimal place are input to buffer B, and signal S7 is output from J1. In response to this signal, 0CG outputs only the contents of output buffer B to VCC, and "Go" is uttered.

With the above operations, the voice output is "Sanjiyuunitenyongo". If the contents of register X are 6125 and the contents of register x are 0, then 6125 is stored by x and
, J2 outputs signal S4, and J2 outputs signal S4.
3 determines that the content of B is 1 and outputs S5.

Therefore, signal S6 is output to 0CG. This results in 0
The CG outputs codes to the CC in the order of the output code of 1CGp and the output code of 2CGd. For this reason, a certain period of silence is provided immediately before the "hiyaku". Also, 3210
Even if a value is stored where the + digit is 1, as in
Similarly, since the signals S4 and S5 are forced, "JIUU"
A certain period of silence is provided immediately before. As explained above, when outputting numerical information including scale, a certain period of silence is provided immediately after 100 million ``ok'' and immediately after 10,000 ``man'', and ``hiyaku'' is created in which the hundreds place is 1. Immediately before , the tens place becomes 1. The above-mentioned problem is solved by providing a certain period of silence immediately before the ``jiyu'' and immediately before the decimal point ``ten''.

FIG. 2 is a block diagram showing an embodiment in which so-called stick reading is performed in which the scale value of numerical information is not uttered, and blocks common to those in FIG. 1 are indicated by the same symbols.

In the figure, J4 determines whether the value input to buffer B is at the comma or decimal point position of the digit separator, and if it is after the comma or after the decimal point, it outputs signal S8, and other outputs S7. 0CG is set to 1 by signal S8
Each code is output to VCC in the order of the output code of CGp, the output code of 2CGc, and the contents of 3 buffer B, and the signal S7
As a result, only the contents of buffer B are output to VCC.

Note that CGc is a digit separator code generator that generates codes corresponding to "comma" and "ten". Therefore, for example, 12345679 in register x, 2 in register
is stored, and 1,234,567.
89 is stored, a silent period of a certain time is provided between "ichi" and "comma" and between "yon" and "comma". Additionally, a certain period of silence is provided between "Nana" and "Ten". Although all the embodiments provide silent sections of the same length of time, the present invention is not limited to this, and the length of the silent sections may be set depending on the situation.

In addition, in the silent section, a sound added to each phoneme from the beginning can be treated as a phoneme. For example, if ``Oku'' and ``Man'' have ``Oku + silent section'' and ``Man + silent section'' as one phoneme, there is no need for a circuit to insert a silent section immediately after 100 million and 10,000 million. becomes. Similarly, in the case of hyaku "hiyaku" or + "jiyuu", "silent interval + hiyaku" or "silent interval + jiyuu" can be treated as one phoneme, but "hiyaku" or "jiyuu" without a silent interval Needless to say, it is also necessary. As described above, the audio output device of the present invention has a
A means for determining units of thousands, a means for detecting decimal points and digit separators, a means for determining numerical values in hundreds and tens digits, and in connection with these means, upper memory, units of thousands, decimal points, digits separators, immediately before audio notification of numerical values with 1 in the digits of 100s and +. Alternatively, since it consists of a means for forming a silent section of a certain period of time immediately after, the quality of audio notification of numerical information can be significantly improved with a very simple configuration.

[Brief explanation of the drawing]

FIG. 1 is a block diagram showing one embodiment of the present invention, and FIG. 2 is a block diagram of essential parts showing an embodiment of values. X...Numeric data storage register, x...
・Decimal point storage register, 0C... Output control circuit, JlJ2... Output digit detection section, B...
Buffer, J3... Numerical detection unit, AG...
...AND gate, CGd...Scale code generation section, CGp...Silent section code generation section, 0C
GL: Output control unit, VCC: Voice synthesis unit, CC: Code conversion unit, AC:
・Address counter, AD...address decoder, R...audio data storage section, D/A...
...D/A conversion section, SP...Audio output section, J
E...END code detection section, J4...
Digit separator: Gc...Digit separator sound code generation part.

Claims (1)

[Claims] 1. By retrieving appropriate phonemes from phoneme memory,
In a device that notifies numeric information as well as information such as scale, digit separator (comma), decimal point, etc. by sound, a means for determining whether the numeric value to be notified is in 100 million or 10,000 units, and a numeric value. Detection means for detecting the decimal point and digit separator included in the information, and the hundreds and tens digits of the numerical information
In relation to the discrimination output of the numerical discrimination means for discriminating whether or not the unit is 100,000, and the discrimination output of the pre-memory unit discrimination means, a silent interval of a certain time is formed immediately after the audio notification of the unit of 100,000,000, and the detection means and the numerical discrimination 1. An audio output device comprising means for forming a silent period of a certain time immediately before audio notification of a decimal point, digit separator, hundreds and tens digits in connection with the detection and discrimination output of the means.