TW515191B - Voice code switching method, voice code switching device and voice communication terminal - Google Patents

Abstract

Because the encoding and decoding of the high efficiency voice encoding method are performed by using information in the frame and the bygone encoding and bygone decoding results of the frame, an output voice becomes discontinuous when a voice sign is received at a terminal and switched in the middle of sign series to be decoded, and the continuous interrupts of bygone encoding. In a voice sign switching method which receives a voice sign A in which a voice inputted by a speaker A is encoded and a voice sign B in which at voice inputted by a speaker B is encoded, and selects and outputs either the voice signs A and B for every frame for a prescribed time, the selection control of an output frame is performed on the basis of a bygone selection result.

Description

515191 V. Description of the invention (1) Technical field of the invention The present invention relates to voice communication regarding input sound and input sound code, outputting the output sound code after encoding the input sound and decoding the output sound after decoding the input sound code. Terminal. The present invention also relates to an audio communication system using the audio communication terminal. In particular, it is related to the distribution of the input sound input to each voice communication terminal by using multiple voice communication terminals on a low-speed line, so that the voice of one-to-one call content or conference content can be heard at each voice communication terminal. Communication system and voice communication terminal suitable for the voice communication system. Known technology In business communications, etc., an audio communication system that uses related personnel to monitor business contact calls between two speakers. Such conventional audio communication terminals and audio communication systems have a structure in which two types of audio are superimposed on an analog audio signal, and the superimposed signal is transmitted on one line. For example, when talking between speaker A and speaker B, the voice communication terminal A on the side of speaker A overlaps the received voice signal of speaker B and the voice signal of speaker A, and then speaks to the speaker. Method B for the voice communication terminal B on the side and another terminal to transmit the superimposed voice signal. Fig. 22 shows the overall structure of a conventional audio communication system. In Fig. 22, 1 is the terminal A used by speaker A, 2 is the terminal B used by speaker B, and 3 is the terminal C used by other related personnel when monitoring. Set the terminal A to demodulate the modulation signal B from the terminal B

2103-4159-P.ptd Page 4 515191 V. Description of the invention (2) Analogue demodulation device for outputting sound A 4. The input sound A spoken by speaker A and the output sound A are overlapped to produce overlapping sounds. Sound overlapping device and analog modulation device 6 for modulating overlapping sound. The terminal B is provided with an analog modulation device 7 for modulating the input sound B said by the speaker B and an analog demodulation device 8 for demodulating the modulation signal A from the terminal A. An analog demodulation device 9 for demodulating the modulation signal A from the terminal A is also provided in the terminal C. In addition, 10 is a line A that transmits a modulation signal A from the terminal A, and 11 is a line B that transmits a modulation signal B from the terminal B. The operation of the conventional audio communication system will be described below with reference to the drawings. Input the voice spoken by the speaker A at the terminal A as the input voice A. A modulation signal B is input from a terminal B to be described later via a line B. The analog demodulation device 4 demodulates the modulation signal B, and outputs the obtained sound signal as an output sound A. After the sound superimposing device 5 generates the superimposed sound which overlaps the input sound A and the output sound A, it inputs the analog modulation device 6. The analog modulation device 6 modulates the input superimposed sound, sets the obtained modulation signal to a modulation signal A, and outputs the modulation signal A to the line A. In the terminal B, the voice spoken by the speaker B is input as the input voice B and the analog modulation device 7 is input. The modulation signal A is input from the terminal A to the analog demodulation device 8 via the line A. The analog modulation device 7 modulates the input sound B, and uses the obtained modulation signal as a modulation signal B and outputs it to the line B. The analog demodulation device 8 demodulates the modulated signal A, and outputs the obtained audio signal as an output audio B.

2103-4159-P.ptd Page 5 515191 V. Description of the invention (3) The terminal C inputs the modulation signal A from the terminal A via the line A. The analog demodulation device 9 demodulates the modulated signal A, and outputs the obtained sound signal as an output sound C. With the configuration described above, when the speaker A and the speaker B each use the terminal A and the terminal B to talk, the terminal C can monitor the conversation between the speaker A and the speaker B. Figure 2 3 In the above-mentioned conventional voice communication system, in order to improve the introduction of data communication, the line utilization efficiency, the combination of the use and the error correction code improve the quality, the use of the password and the combination of passwords in the weak areas of radio communication to ensure the secret. Other structural diagrams that try to digitize the system. In FIG. 23, 1 is the terminal A used by the speaker A, 2 is the terminal B used by the speaker B, and 3 is the terminal C used by other related persons when monitoring. The terminal A includes a demodulation device 1 for demodulating the modulation signal B from the terminal B 2. A sound decoding device 1 for decoding the sound code B obtained at the demodulation device 1 2 3. A sound decoding device 1 The output of 3 is subjected to digital analog conversion (D / A conversion). The output is analog signal output sound D / A conversion device 14 of the sound A, the input sound A said by speaker A and the output of the D / A conversion device 14 Sound overlapping device 15 for sound A overlapping, A / D conversion device 16 for performing analog digital conversion (A / D conversion) on the overlapping sound, sound encoding device 17 for encoding the output of the A / D conversion device 16 and encoding the sound After the digital output of device 17 is digitally modulated, the modulation device 18 is output to line A 10. Terminal B includes demodulation device 20 for demodulating modulation signal A from line A, sound decoding device 21 for decoding the output of demodulation device 20, and

2103-4159-P.ptd Page 6 V. Description of the Invention (4)

For the sound decoding device 21, the analog output to the φ, Kuraichi office, s L output is analogous to 卢 f 之 之 ί ΐ ΐ 仃 仃 仃 仃 仃 Digital analog conversion (D / A conversion) ^ Song ratio on the output of the sound B D / A conversion equipment? 9. ~ Terminal B also includes A / D implementation of the digital conversion (A / D conversion) that speaker B said Device 23, 8/1 conversion device 23 will release you: " 9 encoding device 2 4 and modulation device 25 after modulating the input of voice encoding device 2 4 and outputting it to line B1 1. $ Terminal C includes demodulation to demodulate the modulation signal A from the informant ^ stand: digital analog conversion of the sound decoding device 27 which decodes the output of the demodulation device 26 and the output of the second decoding device 27 (D / A conversion) D / a conversion device 28 which outputs the analog sound signal B is output. ^ Level Second 'will describe the operation of this conventional sound communication system. The two-terminal set A inputs the voice spoken by the voice a. In addition, since 1 ^, the terminal B inputs the modulation signal B via the line B. The demodulation device 12 demodulates the second modulation “No. B, and then outputs the obtained sound second code β to the sound decoding device 丨 3. The sound decoding device 13 decodes the sound code B and converts it to D / A. 1 4 outputs the obtained digital output sound a. D / Α conversion device 1 4 After performing digital analog conversion (D / A conversion) on the digital wheel-out sound A, the output sound A is an analog signal output sound. In the terminal A, after the sound overlapping device 15 generates the overlapping sound that overlaps the input sound A and the output sound A, it is input to the A / D conversion device 16. The A / D conversion device 16 performs the obtained overlapped sound. After the analog digital conversion (A / D conversion), the digital input sound a of the coefficient signal is obtained. The audio encoding device 17 encodes the digital input sound A and outputs the obtained audio code A to the modulation device 18. Transformer device 18 digitizes the entered sound code a

Page 7 515191 V. Description of the invention (5) After modulation, the obtained modulation signal is used as the modulation signal A and output to the line A. Input the voice spoken by the speaker B at the terminal B as the input voice B. A modulation signal A is input from the terminal A via a line A. The A / D conversion device 23 performs analog-to-digital conversion (A / D conversion) on the input sound B to obtain the digital input sound B of the coefficient bit signal. The voice encoding device 24 encodes the digital input voice B, and outputs the obtained voice code B to the modulation device 25. The modulation device 25 digitally modulates the input voice code B, and uses the obtained modulation signal as a modulation signal B to output to the line B. The demodulation device 20 in the terminal B demodulates the modulation signal A, and then outputs the obtained audio code A to the audio decoding device 21. The sound decoding device 21 decodes the sound code A and outputs the digital output sound B to the D / A conversion device 22 ° The D / A conversion device 22 performs digital analog conversion (D / A conversion) on the digital output sound B ), The output is the output sound B of the analog signal. In the voice communication terminal C, a modulation signal A is input from the terminal A via the line A. The demodulation device 26 demodulates the modulated signal A, and outputs the obtained audio code to the audio decoding device 27. The audio decoding device 27 decodes the audio code A and outputs the digital output sound C to the D / A conversion device 28. The D / A conversion device 28 performs digital analog conversion (D / A conversion) on the digital output sound C, and outputs the output sound C which is an analog signal. With the structure shown above, the digital communication system is also exactly the same as the conventional analog communication system. When speaker A and speaker B each use terminal A and terminal B to talk, the terminal C can monitor the speaker. A talks with speaker B.

2103-4159-P.ptd Page 8

V. Description of the Invention In JP 5 7-4 1 0 64, S discloses a mainstay mainline that enables the three to communicate with each other simultaneously. The digital main line is for each speaker, after comparing the voice PCM codes of the other two speakers f, select the higher output sound for each sample.

Said the PC code. By constructing in this way, the voices of the other two speakers, pc M, are decoded and added at one time without having to restore the pCM code, which can change the hardware scale / J. The problem to be solved by the invention has been widely used in recent years. The high-efficiency sound encoding method represented by the CELP method, because the output sound obtained by once decoding the input sound is re-encoded or the sound of multiple speakers who speak at the same time has the feature of greatly degrading the encoding quality. When this high-efficiency voice coding method is used to implement the digital communication system shown in FIG. 23, the coding quality of the voice coding device 17 is deteriorated, and the quality of the call sound sent to the terminal b and the terminal C via the line A is deteriorated. . To avoid this problem, please refer to JP 5 7-4 1 0 64, in order to select one of the sound code A and the sound code b to send out on the terminal A, and think of the structure of outputting after switching the sound code. By constructing in this way, it is possible to avoid multiple encodings and audio encoding of multiple speakers, but it has various problems as described below and cannot be easily achieved. In addition, JP-A-Sho 5 7-4 1 0 64 realizes outputting after selecting one of a plurality of sound codes, but it is limited to a sound pCM code that can be switched for each sample, and the following problems have not occurred. In the high-efficiency voice coding method, a frame of about 丨 〇 ~ 50 is used as a frame.

2103-4159-P.ptd Page 9 515191 V. Description of the invention (7) (frame), because the multiple samples constituting the frame are encoded together, they are received at terminal B and terminal C by switching by voice code. Post-decoded code series. Voice code A and voice code B are frequently switched in the speech of speaker A or speaker B. The important part of the content of the speech is omitted due to code switching. The length of the omission is tens of ms. , Hinder the grasp of the content of the call. In the case of only switching codes based on the size of the sound level in the frame, the output sound obtained after decoding the code is switched to the speech of speaker A or speaker B every tens of milliseconds due to frequent switching codes. , There is a problem that neither party's speech can be correctly understood. In addition, only based on the sound level of the frame, the voice code of the speaker with the highest importance to lead the call cannot be correctly selected, and important parts of the speech content are omitted, which hinders the grasp of the call content. In addition, because the code on the unselected side is discarded, there is a problem that the speech content of the speaker on the discarded side is omitted. In addition, since the high-efficiency sound encoding method uses not only the information encoding and decoding in the frame, but also the encoding, decoding result encoding and decoding of the previous frame, the terminal B and the terminal C are switched by the sound code. After receiving the decoded code series, when the sound code A and the sound code B are switched halfway, the output sound obtained after decoding becomes discontinuous because the continuity with the past decoding result is interrupted. In the case where a code series consisting of only a voice code A and a voice code B is output through voice switching, it is impossible to determine when the code sequence occurs when decoding the code series, and it is impossible to determine whether the output sound obtained after decoding becomes Discontinuities have the problem that countermeasures to eliminate discontinuities cannot be taken.

2103-4159-P.ptd Page 10 515191 V. Description of the invention (8) In the voice communication system of business communication, etc., when a business connection between two speakers is in a conversation, one party is communicated to the other person who is listening. The speaker's instructions or contact priority. When the voice code A and the voice code B are switched only according to the size of the voice level, there are cases where the speaker who has not been selected for priority transmission has been selected, and there is a problem of missing instructions such as instructions or contact. In addition, since the code of the input audio B is transmitted a plurality of times, there is a problem that the quality of the decoded audio deteriorates due to accumulation of transmission errors. In order to solve the above-mentioned problem, the present invention aims to provide a voice code switching method, which can utilize the advantages of a high-efficiency voice coding method on the one hand, and realize high-quality calling, monitoring, or conference calling on the other. Means for Solving the Problem The voice code switching method of the present invention is to input a voice code A encoded by an input voice of a speaker A and a voice code B encoded by an input voice of a speaker B, and select a voice at a fixed time frame. One of the code A and the sound code B is output. Among them, the output frame is selected according to the past selection result. In the voice code switching method of the present invention, the voice code A encoded by the input voice of the speaker A and the voice code B encoded by the input sound of the speaker B are input, and the voice code A is selected at a fixed time frame. And one of the sound codes B is output. Among them, after the sound codes A and B of the N frames are stored, according to the stored sound codes A and B of the N frames, before the N frames are selected,

2103-4159-P.ptd Page 11 515191 V. Description of Invention (9) One of sound code A or sound code B is output. In the voice code switching method of the present invention, the voice code A encoded by the input voice of the speaker A and the voice code B encoded by the input sound of the speaker B are input, and the voice code A is selected at a fixed time frame. And one of the sound codes B is output. Among them, when it is determined that the output of both parties is required when one of the sound code A and the sound code B is selected, the non-selected sound codes stored in the front frame in order are not required to output the sound code selected in the front frame. After the output frame is output, the sound code series stored until the time when it becomes unnecessary to output is sequentially output in chronological order. In addition, when the sound code switching method of the present invention sequentially outputs the pre-stored sound code series in chronological order, the sound code of the current frame is arranged at the end of the sound code series, and the stored code is determined. When all or part of the voice code series is a non-voice interval, the configuration at the end of the voice code of the interval is suspended or discarded. In the voice code switching method of the present invention, the voice code A encoded by the input voice of the speaker A and the voice code B encoded by the input sound of the speaker B are input, and the voice code A is selected at a fixed time frame. And one of the sound codes B is output. Among them, when a predetermined input for selecting a voice code of a certain party is selected from the outside, the voice code of one party is preferentially selected according to the predetermined input. In the voice code switching method of the present invention, the voice code A encoded by the input voice of the speaker A and the voice code B encoded by the input sound of the speaker B are input, and the voice code A is selected at a fixed time frame. And sound code B

2103-4159-P.ptd Page 12 515191 V. Description of invention (ίο) Output after one party. Wherein, when the sound code A satisfies a predetermined condition, the sound code A is preferentially selected. Furthermore, the sound code switching method of the present invention inputs the sound code A encoded by the input voice of the speaker A and the sound code input by the speaker B. The next sound code B is selected by outputting one of the sound code A and the sound code B at a fixed time frame. Each of the sound codes is constituted by a data code representing the sound data and an error correction code thereof. The error code B is subjected to error correction decoding and error correction coding, and then selected and output. In the voice code switching method of the present invention, the voice code A encoded by the input voice of the speaker A and the voice code B encoded by the input sound of the speaker B are input, and the voice code A is selected at a fixed time frame. And one of the sound codes B is output. Among them, the sound code A and the sound code B are constructed using only a part of the calibration code that can also decode the sound. The sound code A 'is a part of the sound code A and the sound code B' is a part of the sound code B. The combined code is one of the candidates for selection output. In addition, the voice code switching method of the present invention enables the frame at the time of switching of the selected voice code to output information that can identify the selected voice code before the output of the frame, and sequentially outputs the selected information after the information is output. Voice code. In addition, the voice code switching device of the present invention includes a control device that inputs a voice code A encoded by the input voice of the speaker A and an input of the voice code A by the speaker B.

2103-4159-P.ptd Page 13 515191 V. Description of the invention (11) After entering the sound code B, the sound code A and the sound code B are selected at a fixed time frame and output. The control device selects the output frame according to the past selection result. In the voice communication terminal of the present invention, the voice code A encoded by the input voice of the speaker A and the voice code B encoded by the input voice of the speaker B are constructed by using only a part of the calibration code that can decode the voice. ; Including: Demodulation device, input the code composed of the sound code A 'which is a part of the sound code A and the sound code B' which is a part of the sound code B, and demodulate it; a separation device, which demodulates the sound after demodulation; The code A 'and the sound code B' are separated; the sound decoding device decodes the separated sound code A 'and the sound code B', respectively, and outputs the output sound A and the output sound B; the addition device separates the decoded output sound A and output sound B are added; and a D / A conversion device performs D / A conversion on the output of the addition device. Embodiments of the Invention Embodiments of the present invention will be described below with reference to the drawings. Embodiment 1 FIG. 1 shows a configuration of a terminal A using a voice code switching method according to Embodiment 1 of the present invention. In addition, the use of this terminal A as the terminal A of FIG. 23 becomes a structure of a voice communication system to which the voice code switching method of Embodiment 1 of the present invention is applied. In Figure 1, 16 series A / D conversion devices, 17 series voice encoding devices, 30 series switching devices, 18 series modulation devices, 12 series demodulation devices, 13 series sound decoding devices, and 1 series 4 D / A conversion device.

2103-4159-P.ptd Page 14 515191 V. Description of the invention (12) 31 is a control device constituting the switching device 30, and 3 2 is the same selection device constituting the switching device 30. The operation will be described below with reference to the drawings. Input the voice spoken by the speaker A at the terminal A as the input voice A. A modulation signal B is input from a terminal B to be described later via a line B. The demodulation device 12 demodulates the modulation signal B at a fixed time frame, and outputs the obtained audio code to the sound decoding device 13 and the control device 3 1 and the selection device 3 2 in the switching device 30. B. The audio decoding device 13 decodes the audio code B and outputs the digital output audio A to the D / A conversion device 14. The D / A conversion device 14 performs digital analog conversion (D / A conversion) on the digital output sound A, and outputs the output sound A which is an analog signal. The A / D conversion device 16 performs analog-to-digital conversion (A / D conversion) on the input sound A to obtain the digital input sound A of the coefficient-bit signal. The voice coding device 17 encodes the digital input voice A at a frame of a fixed time, and outputs the obtained voice code A to the control device 31 and the selection device 32 in the switching device 30. The control device 3 1 in the switching device 30 selects the sound code A and the sound code B from the input result of the selected sound code A and sound code B and the sound code determined by the control device 3 1 in accordance with a predetermined selection criterion. After one of them is selected, information on which one is selected is output to the selection device 32. In the selection criteria example, there are the selection m _ 1 according to the amplitude A and amplitude B of the two decoded sounds obtained by actually decoding the sound code A and the sound code B, and the selection result in the past (time of m-1). Select the method according to the flow shown in Figure 2.

2103-4159-P.ptd Page 15 515191 V. Description of the invention (13) Figure 2 shows the case where the amplitude difference between two decoded sounds is larger than the threshold th, and the amplitude is larger than the threshold th. The choice of the situation and the past choice result in — 1 is the same as the 'then' decision, which is the choice m of the current (in time) choice result. By constituting in this way, the terminal C, which is monitoring the conversation between the speaker A and the speaker B, can also be obtained on the output sound C because the speaker A and the speaker B are talking at the same time. The voice of the louder speaker will not hinder the grasp of the call content. In addition, because the difference between the sound levels of speaker A and speaker B is not sufficiently large, the continuity of the selection result is maintained, which can avoid frequent switching of speech in a short time that terminal C hinders the grasp of the content of the call Speaker A and Speaker B speak. In addition, in FIG. 2, the threshold th of the amplitude difference may be a fixed value, or may be an adaptively variable value according to the amplitude of the voice code or decoded sound, the past selection result, or the threshold value of the amplitude. can. The selection device 3 2 switches the switch according to the information input from the control device 31, sets the selected sound code to the sound code A, and outputs it to the modulation device 18. Then, the modulation device 18 modulates the voice code A 'input from the switching device 30 digitally, and uses the obtained modulation signal as the modulation signal A to output to the line A10. In addition, in the above embodiment, the method shown in FIG. 2 is used on the basis of the sound code selection of the control device 31 in the switching device 30, but other options are used, for example, according to the flow shown in FIG. 3 or FIG. 4 The construction of the benchmark is of course possible. In addition, these can be switched adaptively or according to external instructions

2103-4159-P.ptd Page 16 515191 V. Description of the invention (14) Of course, the structure used after selecting the standard is also possible. Fig. 3 shows that the amplitude of the decoded sound of the selected speaker m-1 is smaller than the threshold th in the selection result of the past (m-1 moment), and it is smaller than the other one which is not selected according to the selection m-1. When the amplitude of the speaker is small, choose the speaker that has not been selected according to the selection m-1, otherwise choose the same result as the previous selection m-1, and then decide the result of the current (time m) selection. . In addition, in FIG. 4, the amplitude of the decoded sound of a non-selected speaker according to the selection result of the past (m-1 moment) is greater than the threshold th, and according to the selected m-1 non-selected speaker In other cases, the selection is the same as the past selection result m — 1, and then, the decision is the current (time m) selection result. Also, in the above-mentioned embodiment, when the sound code of the current frame is selected, only the previous frame is used in the past selection result, but the structure of the selection result of the previous frame is of course also used. may. In addition, when the structure of the system is different, it is of course possible to use the structure of the switching device 30 while changing the structure of other devices in the terminal. According to the embodiment 1 ', after inputting the voice code A encoded by the input voice of the speaker A and the voice code B encoded by the input voice of the speaker B, the voice code A and the voice are selected at a fixed time frame. In the method of switching the sound code output after one of the codes B, the current frame selection is controlled according to the past selection result, which has the effect of maintaining the continuity of the selection result. The voice code switching method of the first embodiment is applied to share two lines between multiple terminals, and can also monitor between two terminals at other terminals.

2103-4159-P.ptd Page 17 515191 V. Description of the invention (15) Digitization of the call I. In the case of a call or monitoring-like frequency communication system, it has code switching between # and dual-talking voice In the following, it is possible to avoid frustration. If there is an effect of quality deterioration caused by rain horses, it is particularly encoded once. Embodiment 2 FIG. 5 shows a structure using the terminal A of the present invention. In addition, receiving & = β example Z heart sound tone code switching party, 1 becomes the structure of the application of the present invention / two-terminal: A stand for the end of: Figure 2 3: machine ~ terminal system. The sound code switching method of the second embodiment, sound machine tone 1 is shown in Figures 5 and 30 as the switching device, 3 and 3 are installed late. 31 is the control device for the 31 series and 32 is the storage device for selection. Set 30 and the embodiment shown in FIG. 1 ^. The second embodiment is the same, and the same description is given. The difference from the first embodiment is omitted, and the other operations are described below with reference to the drawings. The terminal A is input with a turn-in sound A, and a voice from a terminal β to be described later via a line 6 is used. Set the frame at 12 every fixed time to change ^ = transformed signal B. The demodulation device 13 stores the data stored in the decoding device 13 and the switching device 30. After the demodulation, the obtained sound code B is output to the sound. After the sound decoupling device 3 5 and the delaying device 3 6, it outputs the obtained set 1 to the D / A conversion device 14 and the 3 decodes the sound code B and converts the device 14 into the digital output sound A]. . After D / A conversion), the output is analog signal output. 曰 Digital analog conversion (D / A instead, A / D conversion device 16 after the second person (A / D conversion), the coefficient bit is obtained. , Said A to perform analog digital conversion. Now the digital input sound A, to the sound 515191 V. Description of the invention (16) The encoding device 17 outputs the digital input sound A. The sound encoding device 17 sends this digital frame at a fixed time frame. After digitally inputting the sound A code, the obtained sound code A is output to the storage device 33 and the delay device 34 in the switching device 30. The storage device 33 in the switching device 30 stores the input sound code A of the past N frames. And output the sound codes of the N frames to the control device 31. In addition, the storage device 35 stores the input sound codes B of the N frames in the past, and then outputs the sound codes of the N frames to the control device 31. 。Control device 3 1 According to the inputted sound code A of the N frames and the sound code B of the N frames, the sound code A before the N frames and the sound before the N frames are selected according to the predetermined selection criterion. One of the codes B then outputs to the selection device 32 information on which party is selected In the selection benchmark example, after actually decoding the sound code A and the sound code B, after analyzing the decoded sounds obtained, it is determined whether each frame is sound or non-sound. According to the determination result, as shown in FIG. 6 The method of process selection. In FIG. 6, first, the sound / non-sound judgment result of the decoded sound of the sound code A before N frames is set to state A, and the sound / non sound of the sound code B before N frames is set. The sound determination result is set to state B. Also, the number of consecutive frames in the sound section or non-sound section of the decoded sound of the sound code A in front of the N frames is set to the continuous length A, and the number of frames in the N frames is used. The number of consecutive frames in the sound interval or non-sound interval of the decoded sound of the previous sound code B is set to the continuous length B. However, in the case where one of the states A and B is a sound and the other is a non-sound, select the system The voice code of one of the voices; and in the case where both the state A and the state B are voices, after comparing the continuous length A and the continuous length B, the continuous length of the sound interval is selected.

2103-4159-P.ptd Page 19 515191 V. Description of the invention (17) One-degree-long sound code; In addition, in the case where both sides are non-sound, after comparing the continuous length A and continuous length b, select the non-sound interval The short one-way voice code. By constituting in this way, the terminal C, which is monitoring the conversation between the speaker A and the speaker B, can also be obtained on the output sound C because the speaker A and the speaker B are talking at the same time. The voice of the speaker who speaks for a long time will not hinder the grasp of the content of the call. Also, because the voice of the ancient speaker who chose to speak for a long time between speaker A and speaker B keeps the continuity of the selection result, which can avoid frequent calls in a short time that terminal C hinders the grasp of the call content. Switch, speaker B speaks. & ^ #A and, the delay device 34 causes the delay code input from the voice encoding device 17 to output to the selection device 32 the two codes before the N frames, A. The delay means 34 gives the demodulation means 12 =

After each frame, the device 32 is selected to delay the delay and the code B is delayed by the device 32 according to the voice code B before the frame = Ν: frame. After the selected sound is set to J: 3, the second f information switch is input, and the modulation device 18 is automatically switched; ′ 〇 is output to the decay device 18. Then, the obtained modulation 彳: j f is entered into the audio code A, and the digital modulation is performed. In addition, in the above, the modulation signal A is output to the line A. 3 1. After analyzing the decoded sound, the parameters included in the control device code in θ ^ switching device 30 directly determine the sound ^ or non-sound, but from the sound, in the above embodiment, I " said the structure of course also possible. Non-sound judgment results, but select ^? Select the reference using only sound / 疋 select the reference using the amplitude of the decoded sound,

515191 V. Description of the invention (18) Of course, the structure of other information such as the result of the past selection is also possible. In addition, the selection device 3 2 and the delay devices 3 4 and 36 are deleted, and the control device 31 selects the sound code of the first one of the N frames and outputs the sound code, omitting the structure of the memory required for the delay. may. If according to the second embodiment, the voice code A encoded by the input voice of the speaker A and the voice code B encoded by the input voice of the speaker B are input, and the voice code A and the voice are selected at a fixed time frame. In the method of switching the sound codes output after one of the codes B, because the sound codes A and B of the N frames are stored in advance, the N codes are selected according to the stored sound codes A and B of the N frames. One of the sound code A and the sound code B in front of the frame is output after it, which has the effect of correctly selecting the speaker who leads the call and maintaining the continuity of the selection result. In the case where the voice code switching method of the second embodiment is applied to a case where a digital voice communication system is used in which two lines are shared between multiple terminals and another terminal can also monitor the calls between the two terminals, It has the effect of not missing the speech content of the speaker who leads the call, and can avoid the quality deterioration caused by the secondary encoding and the encoding of the two-speaking voice under the frequent occurrence of the suppression code switching. Embodiment 3 Fig. 7 shows a configuration of a terminal A using a voice code switching method according to Embodiment 3 of the present invention. In addition, the use of the terminal A as the terminal A of FIG. 23 becomes a structure of a voice communication system to which the voice code switching method of the present invention is applied.

2103-4159-P.ptd Page 21 515191 V. Description of the invention (19) In Figure 7, 30 series switching device, 31 series control device, 3 series selection device, 40 series storage device, 41 series storage device Two selection devices. The third embodiment is the same except that the switching device 30 is different from the first embodiment shown in FIG. 1, and the same reference numerals are assigned, and descriptions thereof are omitted. The operation will be described below with reference to the drawings. Input the voice spoken by the speaker A at the terminal A as the input voice A. A modulation signal B is input from a terminal B to be described later via a line B. The demodulation device 12 demodulates the modulation signal B at a fixed time frame, and outputs the modulation signal B to the sound decoding device 13, the control device 31 in the switching device 30, the second selection device 41, and the selection device 32. The voice code B is obtained. The sound decoding device 13 decodes the sound code B and outputs the digital output sound A to the D / A conversion device 14. The cD / A conversion device 14 performs digital analog conversion (D / A conversion) on the digital output sound A. ), The output is analog output signal A. In addition, the A / D conversion device 16 performs analog digital conversion (A / D conversion) on the input sound A of the speaker A to obtain the digital input sound A of the coefficient signal, and outputs the digital input sound to the sound encoding device 17 A. The voice coding device 17 encodes the digital input voice A at a frame of a fixed time, and outputs the obtained voice code A to the control device 31, the selection device 32, and the second selection device 41 in the switching device 30. The control device 31 in the switching device 30 is stored in the storage device 40 that can be grasped from, for example, the inputted sound code A, sound code B, and the selection result for the second selection device 41 determined in the past according to a predetermined selection criterion. Status, output selection information about the code stored in the storage device 40 to the second selection device 41 and output selection information about the code output to the modulation device 18 to the selection device 32

2103-4159-P.ptd Page 22 515191 V. Description of the invention (20) Selection information. In the selection criterion example, after actually decoding the sound code A and the sound code B, after analyzing the decoded sounds obtained, it is determined whether each frame is sound or non-sound. According to the determination result, the sound in the storage device 40 is determined. The storage status of the code and the past selection results are selected according to the table in Figure 8. In FIG. 8, first, the voice / non-sound judgment result of the decoded voice of the voice code A is set to state A, and the voice / non-sound judgment result of the voice code B is set to state B. The storage state of the audio code in the storage device 40 is set to the condition M. The selection result of the selection device 3 2 in the past (time of m — 1) is set to selection 3 2m — 1, the selection result of the selection device 32 at now (time of m) is set to selection 32 m, and will be selected in the second selection device 41 The selection result is set to be 4 1 m 0, and if there is a sound or code in any of the state A, state B, and state M, the sound code is selected and output. In addition, in the case where there are two or more sounds or codes, the sound codes currently being output are preferentially selected and output, and one of the sound codes A and B which is not output is stored in the storage device 40. By constituting in this way, since the speaker A and the speaker B speak at the same time, the voice of the speaker of the other party can be stored before the speaker of the other party finishes speaking, and the voice of the other party can be output beforehand. The stored voice of the speaker of the other party is not lost on the terminal C that monitors the conversation between the speaker A and the speaker B, and does not hinder the grasp of the call content. The second selection device 41 switches in accordance with the information input from the control device 31

2103-4159-P.ptd Page 23 515191 V. Description of the invention (21) After the switch is turned on, the selected audio code is output to the storage device 40. The storage device 40 is, for example, a F I F 0 type memory buffer. After storing the sound code series inputted by the second selection device 41, the sound device outputs the sound code series to the selection device 3 2 in order. In addition, the selection device 32 switches the switch according to the information input from the control device 31, and sets the selected sound code to the sound code A ', and outputs it to the modulation device 18. Further, the modulation device 18 digitally modulates the voice code A 'input from the switching device 30, sets the obtained modulation signal as the modulation signal A, and outputs it to the line A. In addition, in the above embodiment, the control device 31 in the switching device 30 analyzes the decoded sound and determines whether it is sound or non-sound. However, the parameters of the sound code directly determine the structure of the sound / non-sound. may. Also, in the above-mentioned embodiment, the audio / non-sound determination result is used as the selection criterion for code switching, but it is of course possible to use other information structures such as the amplitude of the decoded speech as the selection criterion. In the above-mentioned embodiment, the method shown in the table of FIG. 8 is used for the selection criterion of the control device 31 in the switching device 30, but a structure using another selection criterion is of course possible. In addition, it is a matter of course that a configuration that switches a plurality of selection bases adaptively or in accordance with an external instruction is also possible. According to the third embodiment, the voice code A and the voice code B encoded by the input voice of the speaker A and the voice code B after the voice input of the speaker B are input, and the voice code A and the voice are selected at a fixed time frame. In the method of switching the sound code output after one of the codes B, when it is determined that the selection output of both the sound code A and the sound code B is required, it is stored in order in the one that is not selected in the previous message box.

2103-4159-P.ptd Page 24 515191 V. Description of the invention (22) The sound code is output after the selection of the sound code selected in the previous frame becomes the unnecessary frame, and then is output in that order in that order. The sound code series stored so far has the effect of not missing the required sound code. In the case where the voice code switching method of the third embodiment is applied to a case where a digital voice communication system is used in which two lines are shared between multiple terminals, and the other terminals can also monitor calls between the two terminals, It has the effect of not missing the content of the call, and can avoid the quality deterioration caused by the secondary encoding and the encoding of the two-talk voice. In addition, when the pre-stored sound code series is sequentially output in chronological order, the sound code of the current frame is stored at the end of the sound code series. When it is determined that the stored sound code series is a non-sound interval, This makes it possible to suspend the storage of the audio code in the section, and it is not necessary to memorize the non-sound section that is not related to the content of the call, because the memory capacity required for the storage of the code can be reduced, which has the effect of reducing the device scale. Embodiment 4 Fig. 9 shows a configuration of a terminal A using a voice code switching method according to Embodiment 4 of the present invention. In addition, the use of the terminal A as the terminal A of FIG. 23 becomes the structure of a voice communication system to which the voice code switching method of the present invention is applied. In FIG. 9, 30 is a switching device, and 31 is a control device. 3 2 series selection device, 40 series storage device, 4 2 series second storage device. The fourth embodiment is the same except that the switching device 30 is different from the first embodiment shown in FIG.

2103-4159-P.ptd Page 25 515191 V. Description of the Invention (23) The following describes the operation according to the figure. Input the voice spoken by the speaker A at the terminal A as the input voice A. A modulation signal B is input from a terminal B to be described later via a line B. The demodulation device 12 demodulates the modulation signal B at a fixed time frame, and outputs the obtained sound to the sound decoding device 13, the control device 31 in the switching device 30, the storage device 42, and the selection device 32. Code B. The sound decoding device 13 decodes the sound code B, and outputs the digital output sound A to the D / A conversion device 14. The D / A conversion device 14 performs digital analog conversion (D / A conversion) on the digital output sound A. ), The output is analog output signal A. The A / D conversion device 16 performs analog digital conversion (A / D conversion) on the input sound A of the speaker A to obtain the digital input sound A of the coefficient signal, and outputs the digital input sound A to the voice encoding device 17. The voice encoding device 17 encodes the digital input sound A at a fixed time frame, and outputs the obtained voice code A to the control device 31, the storage device 40, and the selection device 32 in the switching device 30. The control device 31 in the switching device 30 is stored in accordance with a predetermined selection criterion from the storage status of the inputted audio code A, audio code B, and audio codes determined in the past of the storage devices 40 and 42. The storage device 40 outputs information about whether or not the voice code A is stored, outputs information about whether the voice code B is stored to the storage device 42, and outputs selection information about the code output to the modulation device 18 to the selection device 32. In the selection criterion example, after actually decoding the sound code A and the sound code B, after analyzing the decoded sounds obtained, it is determined whether each frame is sound or non-sound. According to the determination result, it is stored in the storage device 40 and stored Device 4 2 Voices

2103-4159-P.ptd Page 26 515191 V. Description of the Invention (24) Tone = storage f status and past choices-part of the table selection method. In FIG. 10, firstly: j "0 table, the sound voice / non-sound judgment is 100 dollars, the decoded tone of code code A " ... the determination result is set to two, and the sound state is set For the condition MA, the storage MB of the storage device 40 will be stored in the storage device 42. The selection device 32 will also be selected in the past two 41 moments) = = = The condition is set to the condition selection m-1, and the selection device 32 will be selected at the moment) k The selection result is set to selection m. You V 丨 丨 丨 %% J's selection result is set to selected =: In state A, state B, state MA, and state J JC state 'select this sound code and output. And :: ：： There is a case of Feng or OR, it is preferred to output one that is not stored in J storage and sound code b. It is stored in storage device 40 or storage device 42. 洚 Saiter A and speaker 3 Simultaneous speaking = person ... a perfect composition, because the other one is stored before speaking: ΐΐί 情 = You can also output a pre-voice after the speaker of one side finishes speaking, 'Wait until the speaker of the side finishes speaking, no η = the voice of the speaker of the other party. Save, by "ί; Er Chu Λ installs the switch voice code" ° Voice code B and then store the pre-stored saying ㈣ The device, in the case of outputting the words, is also due to Ϊ Ϊ ΒΐII, and the speaker A and the speaker B simultaneously listen to the speaker A and the Go: son and § have recalled the voice codes of the two speakers, The terminal C 'of the call that does not speak the right ^ will not miss the two ΜΛ, which hinders the grasp of the content of the call. The storage 4 and 4 0 are, for example, F 丨 F 〇 stored in the control device 31, ^, The hidden punch is stored from the sound according to whether the sound code a entered in 31 is stored or not.

515191 V. Description of the invention (25) After the sound code series input by the encoding device 17 is output, the sound code series is sequentially output to the selection device 32 in chronological order. The storage device 42 is, for example, a FIFO-type memory buffer. According to whether or not the information of the sound code B input from the control device 31 is stored, and the sound code series input from the demodulation device 12 is stored, it is sequentially sent to the chronological order. The selection device 3 2 outputs the sound code series. The selection device 3 2 switches the switch according to the information input from the self-control device 31, and sets the selected sound code A or B as the sound code A ', and outputs it to the modulation device 18. Further, the modulation device 18 digitally modulates the voice code A 'input from the switching device 30, and sets the obtained modulation signal as the modulation signal A and outputs it to the line A. In addition, in the above-mentioned embodiment, the control device 3 1 in the switching device 30 analyzes and decodes the sound and determines whether it is sound or non-sound. However, the parameters of the sound code directly determine the structure of the sound / non-sound. may. Also, in the above-mentioned embodiment, the audio / non-sound determination result is used as the selection criterion for code switching, but it is of course possible to use other information structures such as the amplitude of the decoded speech as the selection criterion. Further, in the above-mentioned embodiment, the method of selecting the code of the control device 31 in the switching device 30 according to the table showing a part of it in FIG. 10 is used, but a structure using another selection standard is of course possible. . In addition, it is of course possible to adapt a configuration that switches multiple selection criteria according to external instructions. If according to the fourth embodiment, the voice code A encoded by the input voice of the speaker A and the voice code B encoded by the input voice of the speaker B are input, and the voice code A and the voice are selected at a fixed time frame. Output after one of the codes B

2103-4159-P.ptd Page 28 515191 5. In the voice code switching method of invention description (26), when it is determined that the choice of both voice code A and voice code B is required, it is stored in the front frame in order. If the voice code of one party is not selected, after the selected output of the voice code selected in the previous frame becomes the unnecessary frame, the sound code series stored up to that time will be output in order in chronological order. The effect of voice code. In addition, since the voice code A and voice code B devices are separately included, the voice code switching method of the fourth embodiment is applied to share two lines between multiple terminals, and can also be used in other terminals. Monitor the situation of digital audio communication system for calls between 2 terminals, with speaker A and speaker B outputting the pre-stored voice code series, and will not miss the call content, and avoid secondary coding And the quality deterioration caused by the encoding of the two-speaking voice. In addition, when the pre-stored sound code series is sequentially output in chronological order, the sound code of the current frame is stored at the end of the sound code series. When it is determined that the stored sound code series is a non-sound interval, This makes it possible to suspend the storage of the audio code in the section, and it is not necessary to memorize the non-sound section that is not related to the content of the call, because the memory capacity required for the storage of the code can be reduced, which has the effect of reducing the scale of the device. Embodiment 5 Fig. 11 shows the structure of a terminal A using a voice code switching method according to Embodiment 5 of the present invention. In addition, the use of the terminal A as the terminal A of Fig. 23 has a structure of a voice communication system to which the voice code switching method of the present invention is applied.

2103-4159-P.ptd Page 29 515191 Five descriptions of the invention (27) In Figure 1, 1, 3 0 series switching device, 31 1 control device, 3 2 selection device, 40 storage device, 4 2 series Two storage devices, 4 3 series change code generation settings. The fifth embodiment is the same except that the switching device 30 and the first embodiment shown in FIG. 1 are the same, and the same reference numerals are assigned, and the description is omitted. The operation will be described below with reference to the drawings. On the terminal A, the voice of the speaker a is input as the voice A. Also, a modulation signal b is input from a terminal B to be described later via a line B. The demodulation device 12 demodulates the modulation signal b at a fixed time frame, and outputs the obtained audio code to the sound decoding device 13, the control device 31 in the switching device 30, the storage device 42, and the selection device 32. B. After the sound decoding device 13 has decoded it, it outputs to the D / A conversion device 14 the digitally turned sound A. After the D / A conversion device 14 converts the digital output sound bit-to-ratio (D / A conversion), the output is a wheel of an analog signal ^ The digital A / D conversion device 16 performs vertical conversion on the input sound of the speaker a ( After A / D conversion), the coefficient-bit signal is obtained, and the analog digital is output to the audio encoding device 17 to output the digital input sound a. The sound A is input in 1 round, and the digital input signal is set at a fixed time frame. The sound control code 31 is installed in the control device 31. After storing the code, the known sound code A is output to the receiver. And the selection device 3 2 the control device 3 in the switching device 3 〇 Press the key to indicate the other: the input of the sound code A, the sound code b, and the sound code determined in the past by the selection base 42 The device 40 comes out with information about whether to store the sound code A ^, stores information on whether to store sound in the storage department and:

515191 V. Description of the invention (28) The selection information of the code for the eighteen rounds of the changing device. f In the selected reference example, there are the sound code A and the sound code β, which are divided into two parts: after the decoded sound is obtained, for each frame: 2 is determined: the result is stored in the storage device 40 and the health storage device. The state of horse storage and the method of table selection as part of the voice of the past. "", "Sheyuan 12" shows it-in Figure 12, first the sound / iit of the decoded sound of sound code A is set to the state ^ the sound / non-sound of sound code β is determined ^ the fruit sound column will be in The storage status of the storage device 40 is set to the state of jealousy 7¾ and the storage status of the device 42 is set to the status Mβ. The selection result is also selected, 2 time) is set to select m -2, i is set to 32 32 In the past (m-1 time), the selection result is set to set the selection result to 32J. Now (m time,), the selection result is set to selection, and the selection is installed, and the code of the stored device 40 changed as the sound code A 'turns out. It is a sound code β or storage, "Voice code A may be stored in storage, or vice versa, from the sound code 6 ^ code of storage device 42 ί ΐ ί Π" or stored in storage device 40 ^: ί 袭 置 42 的In the case of a horse that is rotated out by the code change code generating device 43, the self-variation is chosen to be constituted in this way, and it is rotated out after exhaustion. ίΐ: ν Because the identifiable code switching time is' ° in advance 1, it is possible that the J of the call has a possibility to produce an output sound. See when the code is switched and the special place S, it can improve the quality discontinuity of the product. Later, a slow storage device 40 such as a FIF0 type can be stored in the voice code punch inputted by the control device 31. # Photo is stored as <Beer, stored from the sound 2103-4159-P.ptd page 31 515191 V. Description of the invention (29) After the sound code series input by the encoding device 17 is output, the sound code series is sequentially output to the selection device 3 2 in order of time. The storage device 4 2 is, for example, a memory buffer of type FIF 0, according to whether the information of the sound code B input from the control device 31 is stored, and the sound code series input from the demodulation device 12 is stored in chronological order. The sound code series is sequentially output to the selection device 3 2. In addition, the change code generating device 43 generates a special code C indicating the switching of the audio code, and outputs it to the selecting device 32. The selecting device 3 2 switches the switch according to the information inputted from the self-controlling device 31, and sets the selected voice code to the voice code A 'to output to the modulation device 18. In addition, the modulation device 18 digitally modulates the voice code A 'input from the switching device 30, and sets the obtained modulation signal as the modulation signal A, and outputs it to the line A. In addition, in the above embodiment, the control device 31 in the switching device 30 analyzes the decoded sound and determines whether it is sound or non-sound. However, the parameters of the sound code directly determine the structure of the sound / non-sound. may. Also, in the above-mentioned embodiment, the audio / non-sound determination result is used as the selection criterion for code switching, but it is of course possible to use other information structures such as the amplitude of the decoded speech as the selection criterion. In the above-mentioned embodiment, the method of selecting the code of the control device 31 in the switching device 30 according to the table shown in FIG. 12 is used, but a structure using another selection standard is of course possible. . In addition, it is of course possible to adapt a configuration that switches multiple selection criteria according to external instructions. Moreover, in the above-mentioned embodiment, the code output from the change code generating device 43

2103-4159-P.ptd Page 32 515191 V. Description of the invention (30) It may be a single code, which means that the code representing the change from voice code A to voice code B and the code representing the change from voice code B to voice code A Of course, it is also possible to use a structure with multiple codes. According to the fifth embodiment, the voice code A and the voice code B encoded by the input voice of the speaker A and the voice code B of the input voice encoded by the speaker B are input, and the voice code A and the voice are selected at a fixed time frame. In the method for switching the sound code output after one of the codes B, the selection of the sound code A and the sound code B is changed, so that the sound code of the frame is used to output the information indicating that the selection has changed. After the changed information is selected, the pre-stored voice code series is output in sequence, which has the effect of correctly identifying the switching time of the voice code A and the voice code B from the output voice code series. In the case where the voice code switching method of the fifth embodiment is applied to a case where a digital voice communication system is used in which two lines are shared between multiple terminals, and another terminal can also monitor calls between the two terminals, Since the code switching time that has the possibility of discontinuity of the output sound can be identified, countermeasures such as special processing to eliminate the discontinuity can be adopted, which has the effect of improving call quality. Embodiment 6 Fig. 13 shows a structure of a terminal A using the voice code switching method of the present invention. In addition, the use of the terminal A as the terminal A of Fig. 23 becomes a structure of a voice communication system to which the voice code switching method of the present invention is applied. In Figure 1, 3, 30 series switching devices, 31 series control devices, and 32 series selection devices. In the sixth embodiment, the structure of the control device 31 in addition to the switching device 30

2103-4159-P.ptd Page 33 515191 、 Instruction of the invention (31) Manufactured, that is, the input of the control device 3 1 is about the Li Pi ♦ i ^ J 1 i &quot; Bu /, he is all the following, according to the figure to explain the action. , Terminal A inputs speaker A with input voice A = [from terminal B, which will be described later, enters a modulation letter via line B. ♦ As mentioned above, the input is about the sound code: _bu, enter E. The demodulation device 12 selects / a at a fixed time, and then sends it to the sound decoding device 13 and the switch f ^ to decode the variable signal B and output the obtained sound to the selection device 32: Β; 30. The internal control device 31 decodes the sound code β, and then outputs the code to the D / A conversion device. 4 The code device 1 outputs the sound ... After the conversion device converts to the ratio conversion (D / A conversion), The output is the output sound of the analog signal A. The A / D conversion device 16 performs an analog digit on the input sound a: Qiao) to obtain the digital input sound A of the coefficient signal. After the frame encodes the digital input sound A, the control device 31 and the selection device 32 in the clothing switching device 30 output the resulting ° code A. The control device 3 1 selects the input sound according to the external input E and the predetermined selection. One of the code A and the voice code B, and outputs information about which one is selected to the dream ^ 32. On the selection reference example, there is a case where the voice code A and the voice code B are implemented after the code. Get the amplitude A and amplitude b of the two decoded sounds and the choice of the time (m — 1 time) If the selection of m - 1 ^ according to the method of FIG Shu select lines of FIG. 14 in the case where there is input from the outside indicating the E 4, selected from; select * ^ 1.

515191 V. The sound code indicated in the description of the invention (32). In the case where there is no instruction from external input E, when the amplitude difference between the two decoded sounds is greater than the threshold th, the larger one is selected. In the case where the threshold th is small, the selection is the same as the previous selection result m — 1, and then, it is determined that the selection m is the current (time m) selection result. By constituting in this way, the terminal C that monitors the conversation between the speaker A and the speaker B can send the content to be notified in priority when the speaker A and the speaker B are talking at the same time, and can give instructions or Contact, etc. The selection device 3 2 switches the switch according to the information input from the control device 31, sets the selected sound code to the sound code A, and outputs it to the modulation device 18. Then, the modulation device 18 modulates the voice code A 'input from the switching device 30 digitally, and uses the obtained modulation signal as the modulation signal A to output to the line A. In addition, in the above-mentioned embodiment, the method shown in FIG. 14 is used as the selection standard of the voice code of the control device 31 in the switching device 30, but a structure using another selection standard is of course possible. In addition, it is of course possible to use a configuration that is adapted after switching between these multiple selection criteria based on external instructions. Moreover, in the above-mentioned embodiment, after a predetermined input is accepted from the outside according to the first embodiment, a voice code of one party is preferentially selected according to the predetermined input, but a structure according to another embodiment is of course possible. If according to this embodiment 6 ', the voice code A coded by the input voice of the speaker A and the voice code B coded by the input voice of the speaker B are selected, and the voice code A and the voice are selected at a fixed time frame. In the method of switching the voice code output after one of the codes B, because the predetermined input from the outside is accepted

2103-4159-P.ptd Page 35 515191 5. In the description of the invention (33), the voice code of one party is preferentially selected according to the predetermined input, which has the effect of reliably sending the desired voice code. In the case where the voice code switching method of the sixth embodiment is applied to a case where a digital voice communication system is used in which two lines are shared between multiple terminals, and the other terminals can also monitor calls between the two terminals, It has the effects of giving priority to the content you want to send, and giving reliable instructions and contact. Embodiment 7 The structure of the terminal A using the voice code switching method of the embodiment 7 is the same as that of the embodiment 1 shown in FIG. FIG. 15 shows a selection criterion of a code used by the control device 31 in the switching device 30 of the terminal A in the voice code switching method of the seventh embodiment of the present invention. In FIG. 15, according to the amplitude of the decoded sound obtained by decoding the sound code A and the sound code B and the past selection results, the sound is preferentially selected when the amplitude of the decoded sound of the sound code A is greater than a certain threshold th Code A. Otherwise, the selection is made on the same basis as in the first embodiment. With this structure, speaker B who talks to speaker A and a person monitoring at another terminal can give priority to the contents of speaker A's speech, and can reliably give instructions or contact information. In addition, in the above embodiment, a structure using a selection criterion different from that shown in Fig. 15 is of course possible. In addition, it is of course possible to use a structure that is adapted or switched after switching among these multiple selection bases according to external instructions. Moreover, in the above-mentioned embodiment, according to Embodiment 1, regardless of the voice code B, an exception condition is given to preferentially select when the voice code A meets a predetermined condition.

2103-4159-P.ptd Page 36 515191 V. Description of the Invention (34), but of course the structure according to other embodiments is also possible. If the voice code A encoded by the input voice of the speaker A and the voice code B encoded by the input voice of the speaker B are input according to the present embodiment, the voice code A and the voice are selected at a fixed time frame. In the method of switching the sound code output after one of the codes B, because the sound code B is ignored, the exception condition is preferred when the sound code A meets the predetermined conditions, and the effect of sending the sound code A can be guaranteed. In the case where the voice code switching method of the seventh embodiment is applied to a case where a digital voice communication system is used in which two lines are shared between multiple terminals, and another terminal can also monitor calls between the two terminals, It has the effect that the call content will not be missed among the callers, and the instructions or contacts can be reliably performed. Embodiment 8 Fig. 16 shows the structure of a terminal A using a voice code switching method according to Embodiment 8 of the present invention. In the eighth embodiment, an error correction process is added to the first embodiment described above. In addition, the use of the terminal A as the terminal A of Fig. 23 becomes a structure of a voice communication system to which the voice code switching method of the present invention is applied. In Figs. 16, 4, 5 are error correction coding devices that add error correction code A to data code A from sound encoding device 17, and output them as sound code A. 4 6 are sound codes from demodulation device 12. Error correction code for B. Error correction decoding device for B decoding. 30 is a switching device, 31 is a control device, and 32 is a selection device. 4 7 is the error correction decoding device in the switching device 30, and 48 is the same error correction encoding device in the switching device 30.

2103-4159-P.ptd Page 37 515191 V. Description of the invention (35) H is omitted, because the same symbols as in Fig. 1 indicate the same or equivalent parts. Hereinafter, the operation will be explained in accordance with the drawings. Also, = "1 to input the voice, the voice spoken by A. Also input the modulation from the terminal B described below via line B. Set 12, the fixed time frame to the modulation" d, mouth :: ϊγ 音 6: Switching device 3 code B constructs this point of sound: Β is constructed using data code 8 representing the sound data and its error correction dance. The error correction decoding device 46 outputs 〇 ear = code B to the audio decoding device 13 to enter the data code β obtained after the error correction. "lose

Κ After VI digital analog conversion (D / A conversion), the output is "the digital signal conversion device 16 performs analog digital conversion on the input sound A (A / D error correction code J ί ΐ) after encoding the digital input sound A , To the code device 45, the data code A obtained by the ^ material code / attached force. Wrong correction coded code A. w device 31 and selection device 32 round out the error in the sound switching device 30 ^ device 48 outputs a right The input ^ f of the decoding device 47 is obtained by performing error correction on the error correction code Zhuo code B.

2103-4159-P.ptd Page 38 515191 V. Description of the invention (36) Data code B. The error correction coding device 48 adds the error correction code B 'to the data code B as a sound code B', and outputs the sound code B 'to the control device 31 and the selection device 32. By structuring in this way, the sound code B 'eliminates the influence of transmission errors on line B. When outputting to line A, it is only affected by transmission errors on line A, which can avoid the two lines of line B and line A. The effects of transmission errors accumulate. The control device 31 selects one of the inputted voice code A and voice code B 'according to a predetermined selection criterion of the reference shown in the first embodiment, and outputs information to the selection device 32 about which one is selected. The selection device 32 sets the selected sound code to the sound code A 'according to the information switching switch input from the control device 31, and outputs the selected sound code to the modulation device 18. The modulation device 18 digitally modulates the voice code A 'input from the switching device 30, and then uses the obtained modulation signal as a modulation signal A and outputs it to the line A. In addition, in the above-mentioned embodiment, the error correction decoding and the error correction encoding are performed on the sound code B according to Embodiment 1, but the structure according to other embodiments is of course possible. In addition, since the operations of the error correction coding device 45 and the error correction coding device 48 are the same in the above embodiment, the position of the error correction coding device 45 is moved to the switching device 30 and the modulation device 18 as shown in FIG. 17. Of course, the structure of the error correction encoding device 48 after the deletion is also possible. In addition, since the operations of the error correction decoding device 46 and the error correction decoding device 47 are the same in the above embodiment, the position of the error correction decoding device 46 is moved to the switching device 30 and the demodulation device 12 as shown in FIG. 18. It is of course possible to construct the error correction decoding device 47 after a while.

2103-4159-P.ptd Page 39 515191 V. Description of the Invention (37) According to the first embodiment, after inputting the voice code A encoded by the input voice of the speaker A and the voice input A encoded by the speaker B, The sound code B is a sound code switching method in which one of the sound code A and the sound code B is selected at a fixed time frame. Each sound code is composed of a data code representing the sound data and its error correction code. The error correction decoding and error correction encoding of the sound code B can output the sound code without the influence of transmission errors. In the case where the voice code switching method of the first embodiment is applied to a case where a digital voice communication system is used in which two lines are shared between multiple terminals and another terminal can also monitor calls between the two terminals, It has the effect of avoiding the accumulation of transmission errors of each line and achieving a high-quality call for the sound code B passing through multiple lines. Embodiment 9 Fig. 19 shows a structure of a terminal A using a voice code switching method according to Embodiment 9 of the present invention. Fig. 20 shows the structure of a voice communication system using this terminal A. In Figure 19, 30 series switching devices, 31 series control devices, 50, 51 series code extraction devices, 52 series code combination devices, and 32 series selection devices. The terminal A of the ninth embodiment is the same except that the switching device 30 is different from the first embodiment shown in FIG. 1, and the same reference numerals are assigned, and the description is omitted. Also in Fig. 20, 5, 3, and 5 are 4 separate devices, 5 5 are audio decoding devices, and 5 6 are addition devices. In addition, the same reference numerals as those in FIG. 23 indicate the same or corresponding parts, and the description is omitted. The operation will be described below with reference to the drawings.

2103-4159-P.ptd Page 40

V. Description of the invention (38) Grasp the input sound A on the terminal A, and from the terminal B described later, set the modulated signal B through the voice of the speaker Lübei 8 and then demodulate the modulated signal B. , Slightly β input modulation signal B. The control device 31 in the demodulation device 30, the code cylinder &apos; to the sound decoding device 13, and the sound code β obtained by the switching device. Here, Α is placed in 2 and the selection device 3 2 outputs the code B 1 as the second part of the two parts. ^ Ma B uses a sound code B 0 and a sound B 0. Although the quality is slightly worse, 3 fixed codes are also obtained. Constructed using only sound codes. After decoding the intelligible sound, the coded sound decoding device 13 decodes the digital output sound obtained by outputting the sound &lt; 1 4 and then outputs the sound A to the D / Α conversion device for digital analog conversion 榼 f . The 0 / eight conversion means 14 outputs sound A to the digital number. , CD / A conversion) after the ’output is analog signal A / D conversion device 16

After conversion), the coefficient-bit signal ^ tone A intake analog digital conversion (after A / D encoding) is output to the control device 31, code extraction device 50, and selection device 32 in the switching device 30, and the sound H Ϊ The sound code A uses a two-stage calibration code structure composed of a sound code A0 and a sound code M. The sound code A is only a poor quality, but also has a code structure that can decode the understandable sound. The code extraction device 50 in the switching device 30 extracts the audio code A0 from the inputted audio code A and outputs it to the code combination device 52. The code extraction device 51 in the switching device 30 extracts the inputted audio code b. After the sound code β0 is extracted, it is output to the code group ^ device 52. The code combination device 52 generates a code composed of the inputted sound code A0 and sound code B0, and outputs it to the selection device μ. Selection

2103-4159-P.ptd Page 41 515191 V. Description of the invention (39) The device 32 selects one of the inputted sound code A and sound code B, and the sound code A0 and sound code B0 according to a predetermined selection criterion. Then, information on which one is selected is output to the selection device 32. The selection device 32 switches the switch according to the information input from the control device 31, sets the selected sound code to the sound code A ', and outputs it to the modulation device 18. Further, the modulation device 18 digitally modulates the voice code A 'input from the switching device 30, and uses the obtained modulation signal as the modulation signal A to output to the line A. Next, the operation of the voice communication system using this terminal A will be described with reference to FIG. Input the voice spoken by the speaker B at the terminal B as the input voice B. A modulation signal A is input from the terminal A via a line A. The A / D conversion device 23 performs analog-to-digital conversion (A / D conversion) on the input sound B to obtain the digital input sound B of the coefficient bit signal. The voice encoding device 24 encodes the digital input voice B, and outputs the obtained voice code B to the modulation device 25. The modulation device 25 digitally modulates the input voice code B, and uses the obtained modulation signal as a modulation signal B to output to the line B. The demodulation device 20 in the terminal B demodulates the modulation signal A from the terminal A, and outputs the obtained audio code A 'to the separation device 53. The separating means 53 separates only the sound code for the input sound A from the sound code A 'and outputs it to the sound decoding device 21. That is, in the case where the sound code A ′ is the sound code A, the sound code A becomes an input to the sound decoding device 21, and in the case where the sound code A ′ is a combination of the sound code A0 and the sound code B0, the sound code A0 becomes Input to the audio decoding device 21.

2103-4159-P.ptd Page 42 515191 Five Description of the invention (40)

After the 2 A code of the sound decoding device, the D / A conversion device converts the sound of the turn from the device 53 to the D / A conversion device (22,412: 1111: ^^ conversion), and then outputs the analog sound output B . Cheek ratio conversion (D / a in the voice communication terminal C, from the green terminal a through the green di a ~.) The demodulation device 26 tunes the modulation signal into the modulation, and applies the sound code of the input sound A and the input The sound code of the sound B is divided into a sound sound decoding device 27, and the sound code for the input sound A is output to the code device 55, and the sound code for the input sound b is output to the code device 55. The interpretation is that "the sound code A" is the sound code A In the case, the sound code a becomes the input of the Shiyin decoding device 27. In the case where the sound code A is a combination of the sound code and ^ ^^ sound B0, the sound code 40 becomes the &amp; exhaustion of the sound decoding device 27. The input sound code B0 becomes the input to the sound decoding device 55. In the case where the sound is exhausted, it is the sound code B. The sound code B becomes the input to the sound decoding device 55. The sound decoding device 27 will be input from the separation device 54. The input voice code for the input voice A is decoded, and the resulting digital output voice c A is output. In addition, the voice decoding device 55 decodes the voice code for the input voice B input from the separation device 54 and outputs the obtained code. Digital output sound c B. D / A The switching device 2 8 performs digital analog conversion (D / A conversion) on the digital output sound c obtained by adding the digital output sound CA and the digital output sound CB with the addition device 5 6 and outputs the output sound C of the analog signal. In addition, in the above embodiment, when the switching device is judged according to the first embodiment

2103-4159-P.ptd Page 43 515191 V. Description of the invention (41) When the selection of both the calibration code sound code A and the sound code B is required, the code that combines one part of the two sound codes is output, but the basis The construction of other embodiments is of course possible. Also, in the above-mentioned embodiment, the sound code is constructed using a two-stage calibration code, but a structure using a two-stage calibration code structure is of course possible. In addition, the code part extracted from these multi-stage calibration codes, or a combination method of which is set in advance, and a structure in which these methods are switched and output are also possible. In the above embodiment, D / A conversion is performed after the two types of digital output sounds are added by the terminal C, but the two types of ratio signals obtained by performing D / A conversion on the two types of digital output sounds are added and set as The structure of the output sound is of course possible. In addition, instead of adding two types of ratio signals into one type of signal, a structure in which each signal is output individually is of course possible. According to the ninth embodiment, the voice code A and the voice code B encoded by the input voice of the speaker A and the voice code B of the input voice encoded by the speaker B are input, and the voice code A and the voice are selected at a fixed time frame. In the method of switching the sound code output after one of the codes B, each sound code is constructed using only a part of the calibration code that can also be decoded, because the code formed by combining the parts of each sound code is selected as a candidate for output. First, it has the effect of outputting a voice code that can decode the voices of two speakers even when speaker A and speaker B speak at the same time. In the case where the voice code switching method of the ninth embodiment is applied to the case where a digital voice communication system is used in which two lines are shared between multiple terminals and the other terminals can also monitor calls between the two terminals, Because speaker A and

2103-4159-P.ptd Page 44 515191 V. Description of the invention (42) The situation where speaker B speaks at the same time does not increase the amount of data. The audio code that can decode the voice of the two speakers is not missing during the call. It can avoid the effect of great quality deterioration caused by secondary encoding and dual-sound encoding. Embodiment 10 Fig. 21 shows a structure of a terminal A using the voice code switching method of the present invention. In addition, the use of the terminal A as the terminal A of Fig. 23 becomes a structure of a voice communication system to which the voice code switching method of the present invention is applied. The difference between Embodiment 10 and Embodiment 1 shown in FIG. 1 is that the output of the control device 31 is output to not only the selection device 32 but also the modulation device 18. Since it is the same as that of FIG. 1, description is omitted. The operation will be described below with reference to the drawings. Input the voice spoken by the speaker A at the terminal A as the input voice A. A modulation signal B is input from a terminal B to be described later via a line B. The demodulation device 12 demodulates the modulation signal B at a fixed time frame, and outputs the obtained audio code to the sound decoding device 13 and the control device 3 1 and the selection device 3 2 in the switching device 30. B. The audio decoding device 13 decodes the audio code B and outputs the digital output audio A to the D / A conversion device 14. The D / A conversion device 14 performs digital analog conversion (D / A conversion) on the digital output sound A, and outputs the output sound A which is an analog signal. The A / D conversion device 16 performs analog-to-digital conversion (A / D conversion) on the input sound A to obtain the digital input sound A of the coefficient-bit signal. Voice coding equipment

2103-4159-P.ptd Page 45 515191 V. Description of the invention (43) Set 1 7 Encode the digital input sound A at a fixed time frame, and send it to the control device 31 and selection device 32 in the switching device 30 The resulting voice code A is output. The control device 31 selects one of the input voice code A and voice code B according to a predetermined selection criterion, and outputs information about which one is selected to the selection device 32 and the modulation device 18. The selection criterion of the control device 31 is the same as that of the first embodiment. The selection device 3 2 switches the switch according to the information input from the control device 31, sets the selected sound code to the sound code A, and outputs it to the modulation device 18. Then, the modulation device 18 modulates the signal of the sound code A 'input from the switching device 30 after the identification information of the code selected by the control device 31, and then uses the obtained modulation signal as the modulation. Change signal A and output to line A. By constituting in this way, the terminal C that monitors the conversation between the speaker A and the speaker B can identify which speaker the received voice code belongs to, and perform processing suitable for each speaker, or the timing of the identification code switching. After that, special treatments such as pre-detection and relaxation of the discontinuity of the output sound C can be performed, which can improve the quality of the bureau. In addition, in the above embodiment, it is possible to output information that can identify which code is selected according to the first embodiment, but a structure according to other embodiments is of course possible. If according to this embodiment 10, after inputting the voice code A encoded by the input voice of the speaker A and the voice code B encoded by the input voice of the speaker B, the voice code A and After one of the sound codes B

2103-4159-P.ptd Page 46 515191 V. Description of the Invention (44) The output voice code switching method, because the output can identify which code is selected, has the correct output voice code series. It is the effect of sound code A or sound code B. The method of switching the voice code of this embodiment 10 is applied to the case of a digital voice communication system in which two lines are shared between multiple terminals and other terminals can also monitor calls between the two terminals. , Because it can identify which speaker the received sound code belongs to, perform processing suitable for each speaker, or perform special processing such as pre-detection and mitigation of the discontinuity of the output sound C after the identification code switching time, which has The effect of improving quality. Effects of the Invention According to the present invention, after inputting the voice code A encoded by the input voice of the speaker A and the voice code B encoded by the input sound of the speaker B, the voice code A and The method or device for switching the sound code output after one of the sound codes B, because the output frame is selected according to the past selection result, the effect of continuity of the selection result can be maintained. In addition, according to the present invention, after inputting the voice code A encoded by the input voice of the speaker A and the voice code B encoded by the input voice of the speaker B, the voice code A and the voice are selected at a frame every fixed time The method of switching the sound codes output after one of the codes B, because after storing the sound codes A and B of the N frames, the N codes are selected according to the stored sound codes A and B of the N frames. The voice code A or voice code B in front of the frame is output after one of them, which has the effect that the code of the speaker who leads the call can be correctly selected and the continuity of the selection result can be maintained.

2103-4159-P.ptd Page 47 515191 V. Description of the invention (45) In addition, according to the present invention, a sound code switching method is output after selecting one of sound code A and sound code B at a fixed time frame. When it is determined that the output of both parties is required when one of the sound code A and the sound code B is selected, the non-selected sound codes stored in the front frame in order are not required to output the sound code selected in the front frame. After the output frame is output, the sound code series stored until the time when it becomes unnecessary to output in order of chronological order has the effect that the required sound code will not be missed. In addition, according to the present invention, after the sound code A encoded by the input sound of the speaker A and the sound code B encoded by the input sound of the speaker B are input, the sound code A and the sound are selected at a fixed time frame. The method for switching the sound code output after one of the codes B, because when a predetermined input for selecting and instructing the sound code of a certain party is accepted from the outside, the sound code of one party is preferentially selected according to the predetermined input, and the sound code can be sent reliably. The effect. In addition, according to the present invention, after inputting the voice code A encoded by the input voice of the speaker A and the voice code B encoded by the input voice of the speaker B, the voice code A and the voice are selected at a frame every fixed time. The method for switching the voice code output after one of the codes B, because regardless of the voice code B, when the voice code A satisfies a predetermined condition, it has an exceptional condition of preferential selection, and has the effect of reliably sending the voice code A. In addition, according to the present invention, after inputting the voice code A encoded by the input voice of the speaker A and the voice code B encoded by the input voice of the speaker B, the voice code A and the voice are selected at a frame every fixed time. A method for switching the sound code output after one of the codes B is used, because each sound code is constituted by a data code representing the sound data and its error correction code, so that the sound code B is wrong.

2103-4159-P.ptd Page 48 515191 V. Description of the invention (46) Error correction decoding and error correction coding, with sound code that can output without the influence of transmission errors. In addition, according to the present invention, after inputting the voice code A encoded by the input voice of the speaker A and the voice code B encoded by the input voice of the speaker B, the voice code A and the voice are selected at a frame every fixed time. The method for switching the sound code output after one of the codes B, because each sound code is a calibration code structure that can decode the sound using only a part of it, so that a code formed by combining parts of each sound code is selected as a candidate for output One is the effect of outputting a voice code that can decode the voices of two speakers even when speaker A and speaker B speak at the same time. In addition, according to the present invention, after inputting the voice code A encoded by the input voice of the speaker A and the voice code B encoded by the input voice of the speaker B, the voice code A and the voice are selected at a frame every fixed time. The method of switching the sound code output after one of the codes B, because the output can identify which code is selected, has the effect of correctly identifying the output sound code series being the sound code A or the sound code B. Brief Description of the Drawings Fig. 1 is a block diagram showing the structure of a terminal in a voice communication system using the voice code switching method of the first embodiment of the present invention. Fig. 2 is an explanatory diagram showing a code selection step in a voice code switching method according to the first embodiment of the present invention. Fig. 3 is an explanatory diagram showing another code selection procedure in the voice code switching method according to the first embodiment of the present invention.

2103-4159-P.ptd Page 49 515191 V. Description of the invention (47) Fig. 4 is an explanatory diagram showing another code selection procedure in the voice code switching method according to the first embodiment of the present invention. Fig. 5 is a block diagram showing the structure of a terminal of a voice communication system using the voice code switching method according to the second embodiment of the present invention. Fig. 6 is an explanatory diagram showing a code selection procedure in a voice code switching method according to the second embodiment of the present invention. Fig. 7 is a block diagram showing the structure of a terminal of a voice communication system using the voice code switching method according to the third embodiment of the present invention. Fig. 8 is an explanatory diagram showing a code selection criterion in a voice code switching method according to the third embodiment of the present invention. Fig. 9 is a block diagram showing the structure of a terminal of a voice communication system using the voice code switching method according to the fourth embodiment of the present invention. Fig. 10 is an explanatory diagram showing a code selection criterion in the voice code switching method according to the fourth embodiment of the present invention. Fig. 11 is a block diagram showing a structure of a terminal in a voice communication system using a voice code switching method according to Embodiment 5 of the present invention. Fig. 12 is an explanatory diagram showing a code selection criterion in the voice code switching method according to the fifth embodiment of the present invention. Fig. 13 is a block diagram showing a structure of a terminal in a voice communication system using a voice code switching method according to Embodiment 6 of the present invention. Fig. 14 is an explanatory diagram showing a code selection procedure in a voice code switching method according to a sixth embodiment of the present invention. Fig. 15 is an explanatory diagram showing a code selection procedure in a voice code switching method according to a seventh embodiment of the present invention.

2103-4159-P.ptd Page 50 515191 V. Description of the invention (48) Fig. 16 is a block diagram showing the structure of a terminal in a voice communication system using the voice code switching method of the eighth embodiment of the present invention. Fig. 17 is a block diagram showing the structure of a terminal in a voice communication system using the voice code switching method according to the eighth embodiment of the present invention. Fig. 18 is a block diagram showing the structure of a terminal in a voice communication system using the voice code switching method according to the eighth embodiment of the present invention. Fig. 19 is a block diagram showing a configuration of a terminal in a voice communication system using a voice code switching method according to Embodiment 9 of the present invention. Fig. 20 is a block diagram showing the overall structure of a voice communication system using Embodiment 9 of the present invention. Fig. 21 is a block diagram showing the structure of a terminal in a voice communication system using the voice code switching method of Embodiment 10 of the present invention. Fig. 22 is a block diagram showing the overall structure of a conventional audio communication system. Fig. 23 is a block diagram showing the overall structure of a conventional voice communication system which has been digitized. Explanation of symbols 2 Terminal B, 1 2 Demodulation device, 1 4 D / A conversion device, 1 6 A / D conversion device, 1 8 Modulation device, 2 1 Audio decoding device, 1 Terminal A, 3 Terminal C , 1 3 sound decoding device, 15 sound overlapping device, 1 7 sound encoding device, 2 0 demodulation device,

2103-4159-P.ptd Page 51 515191 V. Description of the invention (49) 2 2 D / A conversion device, 2 4 voice encoding device, 2 6 demodulation device, 2 8 D / A conversion device, 3 1 control device , 3 3, 3 5 storage device, 4 0 storage device, 4 2 second storage device, 45 error correction coding device 4 7 error correction decoding device 5 0, 5 1 code extraction device, 5 3, 5 4 separation device, 5 6 Adding device 23 A / D conversion device, 2 5 modulation device, 2 7 sound decoding device, 3 0 switching device, 3 2 selection device, 34, 36 delay device, 4 1 second selection device, 4 3 change code Generating device, 4 6 error correction decoding device 4 8 error correction encoding device 5 2 code combination device, 5 5 sound decoding device,

2103-4159-P.ptd Page 52

Claims (1)

515191 6. Application patent scope 1. A voice code switching method, inputting the voice code A encoded by the input voice of the speaker A and the voice code B encoded by the input voice of the speaker B, at a fixed time frame One of the sound code A and the sound code B is selected and output, which is characterized in that: the output frame is selected according to the past selection result. 2. — A voice code switching method, input the voice code A encoded by the input voice of speaker A and the voice code B encoded by the input voice of speaker B, and select voice code A and The sound code B is output after one of the characters, and is characterized in that after storing the sound codes A and B of the N frames, the N frames are selected in accordance with the stored sound codes A and B of the N frames. One of the previous sound code A or sound code B is output after. 3. — A voice code switching method, input the voice code A encoded by the input voice of speaker A and the voice code B encoded by the input voice of speaker B, and select voice code A and The sound code B is output after one of the characters, and is characterized in that when one of the sound code A and the sound code B is selected and it is determined that the output of both parties is required, the non-selected sound code stored in the previous message frame in order becomes unnecessary. After the output frame of the audio code selected in the previous frame is output, the audio code series stored until the time when it becomes unnecessary to output is output in order in chronological order. 2103-4159-P.ptd Page 53 515191 6. Application for Patent Scope 4 · For the voice code switching method of the third scope of the patent application, when the pre-stored voice code series is output in chronological order, the The sound code of the current frame is arranged at the end of the sound code series, and when it is determined that all or a part of the stored sound code series is a non-voice section, the configuration of the end of the sound code of the section is suspended or discarded. 5. —Voice code switching method, input the voice code A encoded by the input voice of speaker A and the voice code B encoded by the input voice of speaker B, and select voice code A and The voice code B is output after one of the parties, and is characterized in that, when a predetermined input for selecting and instructing a certain party's voice code is accepted from outside, the voice code of one party is preferentially selected according to the predetermined input. 6. —A method of switching voice code, input voice code A encoded by the input voice of speaker A and voice code B encoded by the input voice of speaker B, and select voice code A and The sound code B is output after one of the characters, and is characterized in that when the sound code A meets a predetermined condition, the sound code A is preferentially selected. 7. — A voice code switching method, input the voice code A encoded by the input voice of speaker A and the voice code B encoded by the input voice of speaker B, and select voice code A and The sound code B is output after one of the characters, and is characterized in that each sound code uses a data code representing sound data and its error correction. 2103-4159-P.ptd Page 54 515191 6. Scope of patent application Code composition, error correction decoding and error correction encoding for sound code B, and then select judgment and output. 8. — A voice code switching method, input the voice code A encoded by the input voice of speaker A and the voice code B encoded by the input voice of speaker B, and select voice code A and The sound code B is output after one of the characters, and is characterized in that: the sound code A and the sound code B are constructed by using only a part of the calibration code that can also decode the sound. The sound code A 'and the sound code A are part of the sound code A. A combination of the sound codes B ', which are part of the sound codes B, is used as one of the candidates for selection and output. 9. For example, the method for switching the sound code of item 1, 2, 3, 4, 5, 6, 7, or 8 of the scope of patent application, wherein the frame at the time of selecting the output sound code is switched in the frame. The information that previously recognized the selected sound code is output, and the selected sound code is sequentially output after the information is output. 1 0. A voice code switching device, including a control device, inputting a voice code A encoded by the input voice of the speaker A and a voice code B encoded by the input voice of the speaker B, and a frame at a fixed time interval One of the sound code A and the sound code B is selected and output, which is characterized in that the control device selects the output frame according to the past selection result. 1 1. A voice communication terminal, the voice code A encoded by the input voice of the speaker A and the voice code B encoded by the input voice of the speaker B are calibration codes that can decode the voice using only a part thereof Construction, including: 2103-4159-P.ptd Page 55 515191 VI. Patent application demodulation device, input code composed of sound code A 'which is part of sound code A and sound code B' which is part of sound code B Post demodulation; a separation device that separates the demodulated sound code A 'and a sound code B'; a sound decoding device that decodes the separated sound code A 'and the sound code B', respectively, and outputs output sounds A and An output sound B; an addition device that adds the decoded output sound A and the output sound B; and a D / A conversion device that performs D / A conversion on the output of the addition device. 2103-4159-P.ptd Page 56