JPH0434339B2 - - Google Patents

Abstract

PURPOSE:To attain the transmission and reception of simultaneous conversation of plural persons on one line, by transmitting and receiving voice with a digital signal having a few bit number and providing additionally an address to the signal. CONSTITUTION:Voices of syllables of conversation used daily are sectioned into some 50 sections and converted into syllable coding signals in binary codes and listed as a table, the inputted voice is selected out as the syllable coding signal from the table, and further, the strength/pitch/length of the syllable are detected and represented at a voice accompanying signal in a binary code. Then, an address signal 5 and an order signal 6 are provided additionally to a voice accompanying signal 1 of the syllable code signal and the voice is transmitted between a transmitter 7 and a receiver 8 connected to one line. Then, even if the voice of plural speakers A-D takes place on one line at the same time, the voice is transmitted to listners A'-D' with the signal having a few bit number encoded by the address signal.

Description

DETAILED DESCRIPTION OF THE INVENTION (Field of Industrial Application) The present invention relates to a voice transmission/reception method. Specifically, the present invention relates to a voice communication method in which everyday voice is converted into a coded (digital) signal including the length and shortness of each sound, transmitted, and restored to the original voice on the receiving side.

(Prior Art) In telephone communication, a method is known in which the frequency of voice is analyzed, converted into an analog signal, transmitted, and restored to conversational voice on the receiving side. The voice digitizer used in this transmission method divides the time axis of the audio waveform into extremely fine intervals, captures the amplitude of the audio waveform in each division, extracts it as an analog audio waveform signal, and transmits it as a digital signal. For example, a waveform representing one voice "a" is divided into 2400 bits in the time axis direction, the amplitude signal of each bit is sequentially outputted, and the voice waveform is restored on the input side to obtain a conversational voice. This will be explained in more detail with reference to FIG.

FIG. 1 is a diagram showing the principle of converting a voice waveform into an analog signal. For example, if section A in the diagram represents the sound "a," section B represents the sound "i," and section C represents the sound "woo," then the time axis of section A is
Divide the time into 2400 bits and find the value of the audio amplitude (T ₁ ) for each divided time (t ₁ ).
This is transmitted at once, and the amplitudes of all the bits in section A are totaled to obtain the sound "A" as an analog audio waveform signal.

(Problems to be Solved by the Invention) In the conventional voice transmission/reception system described above, 2400 bits are used to represent a unit of speech, that is, one syllable, so an enormous amount of information is required for the entire conversation. This increases the amount of data required for voice communication, resulting in line congestion. Furthermore, if the number of divisions on the time axis, that is, the amount of bits, is reduced in the conventional method, a clear audio signal cannot be obtained, and in some cases, the audio signal may become indistinguishable from other audio signals.

The present invention encodes audio digital signals,
It is an object of the present invention to provide a voice transmitting/receiving method capable of transmitting a voice signal with an extremely small number of bits compared to the conventional method described above and easily converting the signal to the original voice on the receiving side.

Another object of the present invention is that even if multiple people communicate at the same time on a single line between transmitters and receivers, the voices can be classified and reproduced for each party without causing confusion on the receiver side. The purpose of the present invention is to provide a voice transmission/reception method that can reduce the number of lines between users or allow a large number of people to perform voice communication at the same time with a small number of lines.

(Means for Solving the Problems) The above-mentioned object of the present invention is to sample the input speech at a constant cycle on the transmitting side, segment and encode each syllable, and generate a syllable coded signal. The strength (audio amplitude), pitch (voice frequency), and timbre (voice frequency waveform) of each accompanying sound are detected, each converted into a syllable accompanying encoded signal, and this is attached to the syllable encoded signal. These syllable encoded signals or syllable accompanying encoded signals are sequentially stored in memory, and when a change occurs in these syllable encoded signals or syllable accompanying encoded signals, the time until the change occurs is detected and encoded, and this is used as a time information encoded signal to detect the change. The syllable encoded signal and the syllable accompanying encoded signal immediately before are accompanied by the syllable encoded signal, the syllable associated encoded signal, and the temporal information encoded signal as one speech encoded element signal. This is achieved by sequentially transmitting each signal, converting it to an analog signal on the receiving side and sequentially restoring it to voice according to this transmission order.

According to another aspect of the present invention, the input voice is sampled at a constant cycle on the transmitting side, and each individual syllable is segmented and encoded to produce a syllable encoded signal, so that each syllable associated with each syllable of the voice is The strength (amplitude of the voice), pitch (voice frequency), and timbre (waveform of the voice frequency) are detected, each converted into a syllable accompanying encoded signal, which is attached to the syllable encoded signal and sequentially stored in the memory. , when there is a change in these syllable encoded signals or syllable accompanying encoded signals, the time up to the change is detected and encoded, and this is used as a time information encoded signal to detect the syllable immediately before the change. The coded signal and the syllable accompanying coded signal are associated with each other, and the syllable coded signal, the syllable accompanied coded signal and the time information coded signal are made into one unit of speech coded element signal, and furthermore, it corresponds to a plurality of speakers. The syllable encoded signal or the syllable accompanying encoded signal is obtained by adding to the voice encoded element signal a continuous code signal indicating the signal order in the voice encoded element signal of each speaker identified by the address signal and the address signal. Each time there is a change, the audio encoding element signal is sequentially transmitted along with the audio encoding element signal, and the receiving side classifies the audio encoding element signal according to the address signal and the continuous code signal, converts it into an analog signal, and transmits the audio encoding element signal for multiple conversations. Provided is a voice transmission/reception method characterized by sequentially restoring the voices of each person.

(Example) Next, the present invention will be described in more detail with reference to the drawings.

FIG. 2 shows the principle of digitally encoding and representing human speech in units of speech according to the present invention. First, each syllable, that is, the sound of a syllable in everyday conversational speech, is divided into about 50 categories and represented by binary code, and this is made into a table in advance. This only requires that the syllables be distinguished from each other, and one syllable can be displayed using at most about 8 bits for each syllable. For example, to digitally code and represent the sound "A", the sound (syllable) code signal that specifies the sound of the first syllable "A" is determined as (00000001), and the code 1 in FIG.
Input this syllabic code signal into the item (block) indicated by . Next, the length of the first sound A is detected based on the pronunciation of R, and a signal representing the degree of prolongation of A is expressed in several bits and input into the item indicated by 2 in the figure. From now on, do the same thing to adjust the strength, pitch, and pitch of the “A” sound.
Detect and tone color, convert it into a binary code, and input it to items 3 and 4. Here these signals 2, 3, 4
will be referred to as a signal accompanying the syllable code signal of "a", that is, a syllable accompanying coded signal.

Here, an example will be described in which the length of a sound is detected and a coded signal is generated. First, the input audio signal is sampled at a fixed period and each sound is encoded, but as long as the encoded signal of each sound does not change, the encoded signal is held (stored) as it is within the transmitting device. No transmission takes place. If a change occurs in the next input audio signal (strength, weakness, pitch, etc.), the next encoding is performed, and the time interval from the previous encoding to the current encoding is detected and encoded.
An encoded signal of time information indicating this time interval is added to the immediately preceding encoded audio signal and transmitted. In this way, each time a change appears in the input audio signal, the previous coded signal is replaced with the accompanying time information coded signal (this time information coded signal is used to detect the change in the next signal, for example, the input of a different sound). This can be confirmed after the audio signal is input). Although this method makes real-time mutual conversation difficult, the amount of data transmitted is only a very small number of bits, and the amount of memory required for storage is very small.
The time information encoded signal indicating the time interval is restored to actual long and short speech on the receiving side, resulting in more natural restored speech.

The syllable accompanying encoded signal includes the length and shortness of the syllable,
In addition to strength and weakness, high and low, consonants such as “kotsupu”,
This includes signals representing semi-voiced sounds such as ``pu'' and voiced sounds such as ``go''. Next, for the syllable of "ru", select the syllable code signal that specifies "ru" from the table and input it into the item 1', and check its length and shortness.
Strength, weakness, pitch, etc. are detected, and a syllable accompanying coded signal obtained by binary coding is input to each item 2', 3', and 4'. Here, one syllable code signal and the corresponding syllable accompanying coded signal will be collectively referred to as a speech coded element signal. By combining the audio encoded element signals shown in the figure, the pronunciation language for "R" can be identified. In order to transmit the voice R, it is sufficient to successively transmit the two voice encoded element signals described above. On the receiving side, each digital code signal sent from the transmitting side is analyzed by a code recognition circuit, converted to analog, and a synthesized speech generator outputs the conversational voice "R". FIG. 3 is a diagram showing the transmitting and receiving system in this case. The transmitting side 7 is provided with a microphone 10, a voice analyzer 11, an analog-to-digital converter 12, and a digital code signal transmitter 16, and the receiving side 8 is provided with a code recognition circuit 13, a digital-to-analog converter 14, and a synthesized sound generator. A receiver (telephone receiver, speaker) 15 is provided.

Figure 4 is a system diagram showing a communication network in which multiple subscriber lines are connected to the sending side and/or the receiving side and multiple people are talking at the same time. FIG. 3 is a diagram showing an example of an audio encoded element signal in the case of FIG. in this case,
Following the aforementioned syllable code signal 1 and syllable accompanying coded signals 2, 3, and 4, an address signal 5 and a continuous code signal 6 indicating the signal order in the speech coded element signal of each speaker specified by the address signal are provided. It is encoded (binary encoded) and added.
Even if multiple speakers talk simultaneously using their respective transmitters, a voice encoding element signal is created for each address based on the address signal, and each voice encoding element signal is sent in a certain order determined by the transmitting side. On the receiving side, the received speech coded element signals are classified according to the address signal 5 into received speech encoded element signals corresponding to each speaker on the sending side. At the same time, on the receiving side, the speech coded element signals of each speaker are arranged according to the continuous code signal 6 carried on the element signal, and the transmitting side speaker A, speaker B,
The synthesized restored speech for... is connected to desired speakers on the receiving side, for example, speaker A', speaker B', and so on. In this way, the voice encoded element signals of each speaker can be sent over the same line between the transmitting side and the receiving side, and the number of lines can be significantly reduced compared to normal telephone communication. If a binary continuous code signal is used as the continuous code signal, for example, if some of the voice coded element signals of a certain speaker are received missing due to some accident during transmission, It is possible to immediately notify the transmitting side of this fact and instruct it to retransmit only the audio encoded element signal of the missing portion.

In the above-mentioned embodiment, the speech encoding element signal is in the form of a syllable code signal followed by a syllable accompanying coded signal, but the present invention is not limited to this form, and the present invention is not limited to this form. After the syllable accompanying encoded signal is stored, a corresponding syllable code signal may be followed to form a speech encoded element signal. Similarly, it is also possible to bring the address signal and the continuous code signal before the audio encoded element signal. In addition,
In the above description, syllables refer to utterances of syllable parts of conversational speech, that is, syllables.

(Effects of the Invention) As explained above, according to the present invention, since the length information of the voice is also transmitted, more natural voice can be reproduced. Although it is a little inconvenient for real-time conversation communication with the other party, it is possible to compress the amount of data transmitted and minimize the number of bits.
This is extremely effective in reducing the amount of memory required for code signal storage. Furthermore, even if a large number of people communicate by voice at the same time with a small number of lines, the voices can be classified and reproduced by each caller without confusion on the receiving side. Note that the present invention is not only applicable to remote telephone communication, but can also be applied to voice mail systems and other voice transmission and reception systems, and can also be applied by being incorporated into a document creation device using a voice input method, for example. be.

[Brief explanation of drawings]

FIG. 1 is a diagram showing a voice waveform as an analog signal, FIG. 2 is a diagram showing an example of a voice encoding element signal, and FIG. 3 is a diagram showing an example of a transmitting/receiving system when implementing the present invention. FIG. 4 is a diagram showing a transmission/reception system in the case of a simultaneous call with multiple people, and FIG. 5 is a diagram showing an example of a voice encoding element signal applied in the case of a simultaneous call with multiple people. 1... Syllable code signal, 2, 3, 4... Syllable accompanying coded signal, 5... Address signal, 6... Continuous code signal, 7... Transmitting side, 8... Receiving side, 11...
Speech analyzer, 12... Analog-digital converter, 13... Code recognition circuit, 14... Digital-analog converter, 15... Synthetic sound generator, 16
...digital code signal transmitter, ...speech coded element signal.

Claims (1)

[Claims] 1. On the transmitting side, the input voice is sampled at a constant cycle and each syllable is segmented and encoded to produce a syllable coded signal, and the intensity, pitch, and pitch of each sound accompanying each syllable of the voice are The timbres are detected and converted into syllable-associated encoded signals, which are associated with the syllable-encoded signals and stored sequentially in memory, and when there is a change in these syllable-encoded signals or syllable-associated encoded signals. The time until the change is detected and encoded, and this is made to accompany the syllable encoded signal immediately before the change and the syllable accompanying encoded signal as a time information encoded signal, and these syllable encoded signals are generated. , the syllable accompanying coded signal and the time information coded signal are sequentially transmitted as one audio coded element signal every time there is a change, and according to this transmission order, the receiving side converts it into an analog signal and sequentially restores it to audio. A voice transmission/reception method characterized by: 2 On the transmitting side, input speech is sampled at regular intervals and each syllable is segmented and coded to produce a syllable coded signal, and the intensity, pitch, and timbre of each sound accompanying each syllable of the speech are detected and each syllable is coded. Converting to an accompanying encoded signal and sequentially storing it in memory along with the syllable encoded signal, and when there is a change in these syllable encoded signals or syllable accompanying encoded signals, until the change occurs. Detect and encode the time between, and make it accompany the syllable encoded signal and the syllable accompanying encoded signal immediately before the change as a time information encoded signal, and generate these syllable encoded signals and syllable accompanying encoded signals. and a time information encoded signal as one unit of voice encoded element signal, and furthermore, an address signal corresponding to a plurality of speakers and a sequence indicating the signal order in the voice encoded element signal of each speaker specified by the address signal. A code signal is added to the voice encoded element signal and sequentially transmitted together with the voice encoded element signal every time there is a change in the syllable encoded signal or the syllable accompanying encoded signal, and the receiving side receives the address signal and the continuous A voice transmission and reception method, characterized in that the voice encoded element signals are classified according to code signals, converted into analog signals, and sequentially restored to voices for each of a plurality of speakers.