JP2007312265A - Voice packet communication system and speech reproducer - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a voice packet communication system in which even when a delay time of a voice packet varies, a high quality voice can be reproduced. <P>SOLUTION: In the voice packet communication system, a voice reproducer has a stored amount detecting means for detecting an amount of the voice packet stored in a voice jitter buffer, an information amount calculating means for calculating an information amount of a voice packet, and an information amount determining means for determining whether or not the information amount calculated by the information amount calculating means is equal to or less than a predetermined value. Further, a reproduction control means changes a reproduction time of the voice packet in which an information amount determinator determines that the information amount is equal to or less than the predetermined value in accordance with an amount of the voice packet detected by the stored amount detecting means. <P>COPYRIGHT: (C)2008,JPO&INPIT

Description

  The present invention relates to a voice packet communication system and a voice playback device.

  Conventionally, as a voice reproduction method in a packet communication type voice reproduction apparatus, the received voice packets are sequentially loaded into the voice packet queue while inspecting the sequence number of the voice packets. The packet is taken out, the voice packet is reverse-transformed while checking the sequence number added to the voice packet, and the voice frame sequence in the correct order is obtained. However, since arrival of voice packets fluctuates from a predetermined time interval due to network delay or the like, there is a problem that voice breaks occur during reproduction.

  In order to deal with such a problem, a method has been proposed in which an alternative voice frame is prepared in advance and the alternative voice frame is reproduced when the arrival of the voice packet is delayed from a predetermined time (see, for example, Patent Document 1). .

  On the other hand, as a voice communication technique using a packet communication method, a voice compression technique that distinguishes between voiced and voiceless in units of packet size in order to reduce the bandwidth of the network and does not transfer a voiceless part, or ITU-T Recommendation G . As shown in 729 Annex B, a compression method for transmitting background noise is known only when a background noise pattern is changed during switching from voiced voice to voiceless voice and in a voiceless zone.

In addition, in order to provide the network part with a function for distinguishing the three states of voiced, non-voiced background noise, and no-voiced state, the packet header has identification information regarding the three states, and the packet has not arrived. A method of reproducing alternative voices such as background noise has been proposed (see, for example, Patent Document 2).
JP 7-38608 A Japanese Patent Laid-Open No. 10-285212

  In the method of Patent Document 1, since an alternative audio frame prepared in advance is reproduced, there is a case where it is unnatural and unnatural when the reproduced sound is heard.

  On the other hand, in recent years, the communication band of networks has become broadband, and there is little necessity for compressing and transmitting audio to silent portions and background noise.

  However, in the method of Patent Document 2, the transmission side discriminates the background noise and the no-voice state in the voiced and no-voice states, and the background noise and the no-voice state are greatly compressed and transmitted. Important information to bring out a sense of reality is lost, and there is a possibility that the voice quality is impaired.

  The present invention has been made in view of the above problems, and an object of the present invention is to provide a voice packet communication system capable of reproducing high-quality voice even when the delay time of voice packets varies.

1.
A voice packet encoder that packetizes voice signals into voice packets;
A reproduction information adding unit for adding reproduction information to the voice packet;
A voice transmitting device comprising: a transmitter that transmits the voice packet to which the reproduction information is attached;
A receiving unit for receiving the transmitted voice packet;
A voice jitter buffer for sequentially storing the voice packets received by the receiver;
An audio packet decoder for decoding and reproducing the audio packet;
Reproduction information detecting means for detecting the reproduction information;
An audio playback device comprising: playback control means for instructing the audio packet decoder to play based on the playback information detected by the playback information detection means;
In a voice packet communication system having
The audio playback device
Accumulated amount detecting means for detecting the amount of voice packets accumulated in the voice jitter buffer;
An information amount calculating means for calculating an information amount of an audio packet to be reproduced by the audio packet decoder;
An information amount determination means for determining the information amount calculated by the information amount calculation means;
The reproduction control means includes
Voice packet communication characterized in that, according to the amount of voice packets detected by the accumulated amount detection means, the playback time of the voice packets determined by the information amount judgment means to be information amount equal to or less than a predetermined value is changed. system.

2.
The reproduction control means includes
When the amount of voice packets detected by the accumulated amount detection means is greater than or equal to the first value,
2. The voice packet communication according to 1, wherein voice packets whose information amount is equal to or greater than a predetermined value are sequentially played back without performing playback of the voice packets determined by the information amount determination unit as being equal to or smaller than a predetermined value. system.

3.
The reproduction control means includes
When the amount of voice packets detected by the accumulated amount detection means is less than or equal to the second value,
2. The reproduction of the voice packet determined by the information amount determination means as being an information amount equal to or less than a predetermined value is repeated according to the amount of the voice packet, and then the next voice packet is sequentially reproduced. Voice packet communication system.

4).
The reproduction control means includes
Calculating the amount of change in the amount of voice packets detected by the accumulated amount detection means;
When the fluctuation amount is less than a predetermined value, the first value and the second value are decreased,
When the fluctuation amount is a predetermined value or more, the first value and the second value are increased.
The voice packet communication system according to any one of claims 1 to 3, wherein

5).
The voice packet communication system according to any one of claims 1 to 4, wherein the information amount calculation means calculates the information amount based on an average value of the voice signal.

6).
The information amount calculating means calculates each frequency component of the audio signal, and calculates a difference between the addition average value of each frequency component value of the audio packet calculated up to the previous time and the value of each frequency component calculated this time. 5. The voice packet communication system according to any one of claims 1 to 4, wherein the information amount is calculated based on the information amount.

7).
A receiving unit for receiving a voice packet to which reproduction information is given;
A voice jitter buffer for sequentially storing the voice packets received by the receiver;
An audio packet decoder for decoding and reproducing the audio packet;
Reproduction information detecting means for detecting the reproduction information;
A playback control means for instructing playback to the audio packet decoder based on the playback information detected by the playback information detection means;
The audio playback device
Accumulated amount detecting means for detecting the amount of voice packets accumulated in the voice jitter buffer;
Information amount calculating means for calculating the information amount of the voice packet;
An information amount determination means for determining the information amount calculated by the information amount calculation means;
The reproduction control means includes
An audio reproduction apparatus characterized in that, according to the amount of audio packets detected by the accumulated amount detection means, the reproduction time of the audio packets determined by the information amount determination means to be an information amount equal to or less than a predetermined value is changed. .

8).
The reproduction control means includes
When the amount of voice packets detected by the accumulated amount detection means is greater than or equal to the first value,
8. The audio reproducing device according to claim 7, wherein the audio packet whose information amount is not less than a predetermined value is sequentially reproduced without reproducing the audio packet determined by the information amount determining means to be the information amount not larger than the predetermined value. .

9.
The reproduction control means includes
When the amount of voice packets detected by the accumulated amount detection means is less than or equal to the second value,
8. The reproduction of the voice packet determined by the information amount determination unit to be an information amount equal to or less than a predetermined value is repeated according to the amount of the voice packet, and then the next voice packet is sequentially reproduced. Audio playback device.

10.
The reproduction control means includes
Calculating the amount of change in the amount of voice packets detected by the accumulated amount detection means;
When the fluctuation amount is less than a predetermined value, the first value and the second value are decreased,
When the fluctuation amount is a predetermined value or more, the first value and the second value are increased.
10. The sound reproducing device according to any one of 7 to 9, wherein

11.
11. The audio reproduction device according to any one of 7 to 10, wherein the information amount calculation unit calculates the information amount based on an average value of an audio signal.

12
The information amount calculating means calculates each frequency component of the audio signal, and calculates a difference between the addition average value of each frequency component value of the audio packet calculated up to the previous time and the value of each frequency component calculated this time. The audio reproduction device according to any one of 7 to 10, wherein the information amount is calculated based on the information.

  ADVANTAGE OF THE INVENTION According to this invention, the audio | voice packet communication system which can reproduce | regenerate a high quality audio | voice can be provided even when the delay time of an audio | voice packet fluctuates.

  Hereinafter, the present invention will be described in detail with reference to embodiments, but the present invention is not limited thereto.

  FIG. 1 is a block diagram showing an example of a voice packet communication system according to the present invention.

  FIG. 1A is a block diagram showing the overall configuration, and FIG. 1B is a detailed block diagram of the main microcomputer unit 713.

  The audio transmission device 701 and the audio reproduction device 702 in FIG. 1A are connected to a transmission line 717. The voice transmitting device 701 encodes the input voice into a voice packet and transmits it to the transmission line 717, and the voice reproducing device 702 that receives the voice packet via the transmission line 717 plays back the voice.

  The audio transmission device 701 includes an audio input unit 703, an audio packet encoder 704, a reproduction information addition unit 707, and a transmission unit 708.

  The voice signal input from the voice input unit 703 is packetized by the voice packet encoder 704 to become a voice packet. The audio packet is then input to the reproduction information adding unit 707, and the reproduction information adding unit 707 records the information of the reproduction order as reproduction information in the audio packet. The transmission unit 708 transmits the voice packet provided with the reproduction information to the transmission line 717.

  The transmission line 717 is a wired network or a wireless network, and may be either the Internet or an intranet.

  The audio reproduction device 702 includes an audio output unit 709, an audio packet decoder unit 710, a main microcomputer unit 713, a reception unit 714, and an audio jitter buffer 716.

  Voice packets received by the receiving unit 714 via the transmission line 717 are sequentially stored in the voice jitter buffer 716. The voice jitter buffer 716 is a buffer memory provided for correcting variations in arrival times of voice packets received via the transmission line 717, and the received voice packets are temporarily stored in the voice jitter buffer 716.

  The audio packet decoder unit 710 decodes the packetized audio packet, reproduces the decoded audio signal, and outputs it from the audio output unit 709.

  Next, the main microcomputer unit 713 will be described with reference to FIG.

  The main microcomputer unit 713 includes a microcomputer. That is, the main microcomputer unit 713 includes a CPU (Central Processing Unit) 20 that performs various arithmetic processing, a RAM 21 that is a work area for performing arithmetic, and a ROM 22 that stores a control program and the like, and an audio playback device 702. Centrally controls the operation of each processing unit. As the ROM 22 which is a non-volatile memory, for example, an EEPROM capable of electrically rewriting data is adopted.

  The CPU 20 of this embodiment includes a reproduction information detection unit 10, a reproduction control unit 11, an accumulation amount detection unit 12, an information amount calculation unit 13, and an information amount determination unit 14. The reproduction information detection unit 10, the reproduction control unit 11, the accumulation amount detection unit 12, the information amount calculation unit 13, and the information amount determination unit 14 are the reproduction information detection unit, reproduction control unit, accumulation amount detection unit, and information amount calculation according to the present invention. Means for determining the amount of information.

  The reproduction information detection unit 10 detects information related to the reproduction order from the reproduction information of the audio packet.

  The playback control unit 11 instructs the audio packet decoder 710 to perform decoding and playback based on the playback information detected by the playback information detection unit 10.

  The accumulation amount detection unit 12 detects the amount of voice packets accumulated in the voice jitter buffer 716. Since voice packets are sequentially stored in the voice jitter buffer 716, the storage amount detection unit 12 starts from the address of the head voice packet stored in the voice jitter buffer 716 and the amount of voice packets stored in the voice jitter buffer 716. Can be detected.

  The information amount calculation unit 13 calculates the information amount of the voice packet, and the information amount determination unit 14 determines whether or not the calculated information amount is a predetermined value or less. The amount of information is, for example, the level of an audio signal, and the predetermined value is a level at which it can be determined that there is no sound. Thus, when the level of the audio signal is low, it is considered that the amount of information is low and important information is not included.

  In this embodiment, as an example, the voice transmission device 701 packetizes 20 msec of voice into voice packets. Further, the following processing will be described on the assumption that the reproduction information adding unit 707 assigns numbers to audio packets as reproduction information in the order of transmission and transmits the audio packets from the transmission unit 708 to the audio reproduction device 702. The reproduction information is not limited to the numbers described in the present embodiment as long as the reproduction information is information that can specify the reproduction order on the reproduction side, such as time information.

  A specific example of the processing of the CPU 20 of this embodiment will be described.

  The playback control unit 11 does not start playback until the received voice packet is stored in the voice jitter buffer 716 by a predetermined amount. For example, the playback control unit 11 starts playback control when the amount of voice packets stored in the voice jitter buffer 716 detected by the storage amount detection unit 12 reaches 50% of the storage capacity of the voice jitter buffer 716. To do.

  The reproduction control unit 11 performs control so that reproduction is performed in the order of the reproduction information described in the reproduction information of the audio packet detected by the reproduction information detection unit 10. The reproduction control unit 11 retrieves the audio packet to be reproduced next from the audio jitter buffer 716 from the reproduction information of the audio packet to be reproduced, and sends it to the audio packet decoder unit 710 to reproduce the next audio.

  The audio jitter buffer 716 may underflow or overflow due to the influence of the propagation delay of the network or the data processing time of the audio playback device 702. This phenomenon will be described with reference to FIG.

  FIG. 2 is an explanatory diagram for schematically explaining underflow and overflow. In FIG. 2, reference numeral 716 is a diagram schematically showing the state of the voice packet stored in the voice jitter buffer 716. The outer shape 716 in FIG. 2 represents the maximum memory capacity of the audio jitter buffer 716. The horizontal axis represents the memory address, and the voice packets are stored in FIFO (First In First Out) in order from the memory address at the left end of FIG. The hatched portion represents a portion where the voice packet is accumulated in the voice jitter buffer 716. Therefore, the voice packet stored in the arrow direction on the left side of FIG. 2 is a “new” packet, and the voice packet stored in the arrow direction on the right side is an “old” packet.

  When audio packets are depleted (underflow) or overflow (overflow) from the audio jitter buffer 716, the audio playback is interrupted. It must be stored in the jitter buffer 716. If the risk of underflow is less than or equal to the threshold value U (X), and there is a risk of underflow when the value is greater than or equal to the threshold value O (X), the appropriate range is greater than U (X) and less than O (X). is there. The threshold value O (X) is the first value of the present invention, and the threshold value U (X) is the second value of the present invention.

  Such underflow and overflow are phenomena that cannot be avoided because data is transmitted and received via the IP network. It also occurs due to individual differences in the processing time of the encoder and decoder.

  FIG. 2A is an explanatory diagram for explaining a case where the communication state is good and voice packets are being reproduced smoothly.

  An arrow A1 in FIG. 2A indicates the head audio packet to be reproduced next. In this way, when audio packets are being played back smoothly, it is not necessary to store many audio packets in the audio jitter buffer 716, so that the delay from the real time of the audio to be played is reduced by reducing the storage amount. is doing. The threshold value O (X) is O (1) indicated by an arrow, the threshold value U (X) is U (1) indicated by an arrow, and the accumulated amount of voice packets is within a range from O (1) to U (1). good.

  FIG. 2B is an explanatory diagram for explaining a case where the communication state is bad and the voice packet is not reproduced smoothly.

  An arrow A2 in FIG. 2A indicates the head audio packet to be reproduced next. As described above, when the voice packet is not reproduced smoothly, more voice packets than those in FIG. 2A are accumulated in the voice jitter buffer 716. The threshold value O (X) is O (2) indicated by an arrow, the threshold value U (X) is U (2) indicated by an arrow, and the accumulated amount of voice packets may be in the range of O (2) to U (2). It ’s fine.

  Next, a process of setting the voice packet accumulation amount to an appropriate range in the present embodiment will be described.

  FIG. 3 is a flowchart process showing the flow of reproduction control by the main microcomputer unit 713.

  In the flowchart of FIG. 3, the initial threshold values are, for example, U (1) = 10% and O (1) = 50%, and the number of received voice packets is, for example, 30% of the storage capacity of the voice jitter buffer 716. A process after reaching and starting reproduction will be described.

  S101: This is a step of decoding a voice packet having the oldest reproduction information.

  The reproduction control unit 11 instructs the reproduction information detection unit 10 to search for an audio packet having the oldest reproduction information stored in the audio jitter buffer 716, and sends the searched audio packet to the audio packet decoder unit 710. The packet decoder unit 710 decodes the audio packet into audio data.

  S500: This is a step of calculating the information amount of the decoded voice packet.

  The information amount calculation unit 13 calls a subroutine for calculating the information amount of the decoded voice packet, and obtains the value of the information amount of the voice packet. For example, the amount of information in a voice packet is large when speech is actively performed during a conference, and the amount of information is small when speech is interrupted and the sound is almost silent, or only background sounds such as air conditioning sounds are heard. An algorithm for calculating the amount of information from the audio data will be described in detail later.

  S102: This is a step of starting audio reproduction.

  The reproduction control unit 11 instructs the audio packet decoder unit 710 to reproduce the decoded audio data and output it from the audio output unit 709.

  S800: This is a step of calling a subroutine for setting threshold values O (X) and U (X).

  A threshold setting subroutine for setting the threshold values O (X) and U (X) according to the fluctuation amount of the voice packet amount accumulated in the voice jitter buffer 716 is called. The processing of the threshold setting subroutine will be described in detail later.

  Appropriate threshold values O (X) and U (X) are returned by the threshold setting subroutine. For example, when voice packets are being played back smoothly, threshold values O (X) = O (1) and U (X) = U (1) are set, and when voice packets are not played back smoothly, threshold values are set. O (X) = O (2), U (X) = U (2), and O (2)> O (1), U (2)> U (1).

  S104: This is a step of determining whether or not the accumulated amount of voice packets is O (X) or more.

  The accumulation amount detection unit 12 checks the start address of the voice packet accumulated in the voice jitter buffer 716 and detects the accumulation amount. The reproduction control unit 11 determines whether or not it is greater than or equal to O (X).

  If it is less than O (X) (No at Step S104), the process proceeds to Step S105.

  When it is O (X) or more (step S104; Yes), the process proceeds to step S110.

  S110: A step of determining whether or not the amount of audio information is equal to or less than a predetermined value.

  The information amount determination unit 14 determines whether or not the value of the audio information amount calculated in step S500 is equal to or less than a predetermined value. The predetermined value is a threshold value for determining, for example, a silent state or a state in which only background sound is heard, and is set in advance according to the use environment.

  If it is greater than or equal to the predetermined value (step S110; No), the process proceeds to step S112.

  The audio reproduction is continued, and the process proceeds to step S112.

  If the value is equal to or smaller than the predetermined value (step S110; Yes), the process proceeds to step S111.

  S111: This is a step of stopping audio reproduction.

  The reproduction control unit 11 instructs the audio packet decoder unit 710 to stop audio reproduction, and proceeds to the next step S112. In this case, since the processing of the voice packet is delayed, the delay from the real time of the voice is large, for example, a sense of incongruity when talking is large. For this reason, the reproduction of a voice packet with a small amount of information, for example, a silent state or a background sound that is only heard, is stopped to reduce the voice delay. In this way, there is little sense of incongruity for the person who is listening, and the voice delay time can be shortened without missing important information.

  S105: A step of determining whether or not the amount of voice packets stored is U (X) or less.

  The accumulation amount detector 12 checks the head address of the voice packet stored in the voice jitter buffer 716 and detects the amount of voice packet stored in the voice jitter buffer 716. The playback control unit 11 determines whether or not it is equal to or less than U (X).

  When it exceeds U (X) (step S105; No), it progresses to step S112.

  In the case of U (X) or less (step S105; Yes), the process proceeds to step S122.

  S122: This is a step of determining whether or not the amount of audio information is a predetermined value or less.

  The reproduction control unit 11 determines whether or not the information amount value of the audio data calculated in step S500 is equal to or less than a predetermined value. The predetermined value is a threshold value for determining, for example, a silent state or a state in which only background sound is heard, and is set in advance according to the use environment.

  If it is equal to or greater than the predetermined value (step S122; No), the process proceeds to step S112.

  The audio reproduction is continued, and the process proceeds to step S112.

  If it is equal to or smaller than the predetermined value (step S122; Yes), the process proceeds to step S123.

  S123: This is a step of repeatedly reproducing the sound.

  The reproduction control unit 11 repeatedly stores the audio data determined as the audio data having the information amount equal to or less than the predetermined amount in step S122 until the audio packet is accumulated in the audio jitter buffer 715 by U (X) or more. Command to play. In this case, since the arrival of the audio packet is delayed, an audio packet with a small amount of information, for example, an audio packet in which only a silent state or background sound is heard is repeatedly reproduced, and the video packet is transferred to the video jitter buffer 715 by U (X) or more. Wait until the amount is accumulated. When the accumulation amount detection unit 12 detects that a predetermined amount of video packets have been accumulated in the video jitter buffer 715, the main microcomputer unit 713 proceeds to step 112.

  S112: A step of determining whether or not there is a next voice packet.

  The reproduction control unit 11 searches the audio jitter buffer 716 and searches for the next audio packet to be reproduced.

  If there is a voice packet (step S112; Yes), the process returns to step S101.

  Returning to step S101, the next voice packet is reproduced.

  When there is no voice packet (step S112; No), the process ends.

  As described above, in the present invention, a small amount of information transmitted from the audio transmission device 701, for example, an audio packet in which only a silent state or background sound is heard can be stored in the video jitter buffer 715 when the communication state is good. If the packet is within a predetermined range, it is played back. On the other hand, when the video jitter buffer 715 may overflow or underflow, the audio jitter buffer 715 adjusts the delay time from the real time of the audio by stopping or repeatedly reproducing the audio packet with a small amount of information. The amount of voice packets stored in 716 is set to a fixed amount to secure voice packets. By doing so, it is possible to provide a high-quality voice packet communication system in which important voice information is not impaired and the caller feels uncomfortable with the reproduced voice.

  Next, the routine for calculating the amount of audio information described in step S500 will be described with reference to FIGS.

  FIG. 4 shows a first embodiment of the audio information amount calculation routine of the present invention.

  S501: The average signal level of the audio signal is calculated.

  The information amount calculation unit 13 calculates the average signal level of the audio data decoded by the audio packet decoder 710, and uses this value as the audio information amount.

  For example, it is assumed that the audio data is 12 bits and the complete silence state is a digital value of 0. If the average value calculated in step S500 is a digital value 80, the digital value 80 is the value of the amount of audio information. For example, if the predetermined value determined in step S110 is 100, if the value of the audio information amount is 80, it is determined that there is no sound.

  FIG. 5 shows a second embodiment of the sound information amount calculation routine of the present invention.

  S601: Frequency conversion of an audio signal is performed.

  The information amount calculation unit 13 performs frequency conversion on the audio data decoded by the audio packet decoder 710, and obtains a signal level value for each frequency.

  S602: Addition averaging with the previous value.

  The main microcomputer unit 713 adds the current value to the signal level value for each frequency stored in the built-in memory until the previous time, and obtains an average value for each frequency.

  When the addition average of the frequency distribution of each time is calculated in this way, the average value of the frequency distribution of the background sound such as the air-conditioning sound that always generates the same sound is calculated.

  S603: The difference between the addition average value and the current value is calculated.

  The information amount calculation unit 13 calculates the difference between the addition average value for each frequency obtained in step S602 and the current value.

  S604: The sum of the differences for each frequency is calculated.

  The information amount calculation unit 13 calculates the sum of differences for each frequency obtained in step S603.

  The sum total obtained in step S604 is used as the information amount, and the process returns to the original routine.

  For example, when there is a silent state with a low amount of information, the value calculated in this way is 0. On the other hand, when multiple people speak in a meeting or the like, the frequency distribution is often different each time. The value obtained is large.

  Next, a subroutine for setting the threshold values O (X) and U (X) described in step S800 will be described with reference to FIG.

  FIG. 6 is a flowchart for explaining the processing of the threshold setting routine of the present invention.

  S801: A voice packet accumulation amount is detected.

  The accumulation amount detection unit 12 detects the accumulation amount of the voice packet accumulated in the voice jitter buffer 716. The obtained accumulation amount is stored in the RAM 21.

  S802: The standard deviation is calculated from the value of the accumulated amount up to the previous time and the value of the accumulated amount this time.

  The reproduction control unit 11 calculates a standard deviation from the previous accumulated amount value and the current accumulated amount value stored in the RAM 21.

  S803: It is determined whether or not the standard deviation is greater than or equal to a predetermined value.

The reproduction control unit 11 determines whether or not the calculated standard deviation is greater than or equal to a predetermined value. If the standard deviation is greater than or equal to the predetermined value (step S803; Yes), the process proceeds to step S805.

  S805: O (X) = O (2), U (X) = U (2).

  When the standard deviation is greater than or equal to a predetermined value, the variation in the amount of voice packets accumulated is large, so the playback control unit 11 sets the threshold values O (X) and U (X) to large values O (2) and U (2), Return to the original step.

  When the standard deviation is less than the predetermined value (step S803; No), the process proceeds to step S804.

  S805: O (X) = O (1) and U (X) = U (1).

  When the standard deviation is less than the predetermined value, the amount of accumulated voice packets is large, so the playback control unit 11 sets the threshold values O (X) and U (X) to small values O (2) and U (2), Return to the original step.

  This is the end of the description of the threshold setting routine. Note that the values of the thresholds O and U are not limited to two, and a plurality of values may be set according to the variation of the voice packet.

  As described above, according to the present invention, it is possible to provide a voice packet communication system capable of reproducing high-quality voice even when the delay time of voice packets varies.

It is a block diagram which shows one Embodiment of the structure of the voice packet communication system which concerns on this invention. It is explanatory drawing which illustrates underflow and overflow typically. It is a flowchart process which shows the flow of reproduction | regeneration control by the main microcomputer part 713. FIG. It is 1st Embodiment of the audio | voice information amount calculation routine of this invention. It is 2nd Embodiment of the audio | voice information amount calculation routine of this invention. It is a flowchart explaining the process of the threshold value setting routine of this invention.

Explanation of symbols

DESCRIPTION OF SYMBOLS 10 Reproduction | regeneration information detection part 11 Reproduction | regeneration control part 12 Accumulation amount detection part 13 Information amount calculation part 14 Information amount determination part 20 CPU
21 RAM
22 ROM
701 Audio transmission device 713 Main microcomputer unit 702 Audio reproduction device 717 Transmission line

Claims (12)

A voice packet encoder that packetizes voice signals into voice packets;
A reproduction information adding unit for adding reproduction information to the voice packet;
A voice transmitting device comprising: a transmitter that transmits the voice packet to which the reproduction information is attached;
A receiving unit for receiving the transmitted voice packet;
A voice jitter buffer for sequentially storing the voice packets received by the receiver;
An audio packet decoder for decoding and reproducing the audio packet;
Reproduction information detecting means for detecting the reproduction information;
An audio playback device comprising: playback control means for instructing the audio packet decoder to play based on the playback information detected by the playback information detection means;
In a voice packet communication system having
The audio playback device
Accumulated amount detecting means for detecting the amount of voice packets accumulated in the voice jitter buffer;
An information amount calculating means for calculating an information amount of an audio packet to be reproduced by the audio packet decoder;
An information amount determination means for determining the information amount calculated by the information amount calculation means;
The reproduction control means includes
Voice packet communication characterized in that, according to the amount of voice packets detected by the accumulated amount detection means, the playback time of the voice packets determined by the information amount judgment means to be an information amount equal to or less than a predetermined value is changed. system. The reproduction control means includes
When the amount of voice packets detected by the accumulated amount detection means is greater than or equal to the first value,
2. The voice according to claim 1, wherein the voice packet whose information amount is equal to or greater than a predetermined value is sequentially reproduced without reproducing the voice packet determined as the information amount being equal to or smaller than the predetermined value. Packet communication system. The reproduction control means includes
When the amount of voice packets detected by the accumulated amount detection means is less than or equal to the second value,
2. The audio packet determined by the information amount determination means to be an information amount equal to or less than a predetermined value is repeated according to the amount of the audio packet, and then the next audio packet is sequentially reproduced. The voice packet communication system described in 1. The reproduction control means includes
Calculating the amount of change in the amount of voice packets detected by the accumulated amount detection means;
When the fluctuation amount is less than a predetermined value, the first value and the second value are decreased,
When the fluctuation amount is a predetermined value or more, the first value and the second value are increased.
The voice packet communication system according to any one of claims 1 to 3, wherein 5. The voice packet communication system according to claim 1, wherein the information amount calculation unit calculates the information amount based on an average value of the voice signal. 6. The information amount calculating means calculates each frequency component of the audio signal, and calculates a difference between the addition average value of each frequency component value of the audio packet calculated up to the previous time and the value of each frequency component calculated this time. 5. The voice packet communication system according to claim 1, wherein the information amount is calculated based on the information amount. A receiving unit for receiving a voice packet to which reproduction information is given;
A voice jitter buffer for sequentially storing the voice packets received by the receiver;
An audio packet decoder for decoding and reproducing the audio packet;
Reproduction information detecting means for detecting the reproduction information;
A reproduction control means for instructing the audio packet decoder to reproduce based on the reproduction information detected by the reproduction information detection means;
The audio playback device
Accumulated amount detecting means for detecting the amount of voice packets accumulated in the voice jitter buffer;
Information amount calculating means for calculating the information amount of the voice packet;
An information amount determination means for determining the information amount calculated by the information amount calculation means;
The reproduction control means includes
An audio reproduction apparatus characterized in that, according to the amount of audio packets detected by the accumulated amount detection means, the reproduction time of the audio packets determined by the information amount determination means to be an information amount equal to or less than a predetermined value is changed. . The reproduction control means includes
When the amount of voice packets detected by the accumulated amount detection means is greater than or equal to the first value,
8. The voice according to claim 7, wherein the voice packet whose information amount is equal to or greater than a predetermined value is sequentially reproduced without reproducing the voice packet determined as the information amount being equal to or smaller than the predetermined value. Playback device. The reproduction control means includes
When the amount of voice packets detected by the accumulated amount detection means is less than or equal to the second value,
8. The reproduction of voice packets determined by the information amount determination means to be an information amount equal to or less than a predetermined value is repeated according to the amount of the voice packets, and then the next voice packets are sequentially reproduced. The audio reproducing device according to 1. The reproduction control means includes
Calculating the amount of change in the amount of voice packets detected by the accumulated amount detection means;
When the fluctuation amount is less than a predetermined value, the first value and the second value are decreased,
When the fluctuation amount is a predetermined value or more, the first value and the second value are increased.
The sound reproducing device according to claim 7, wherein the sound reproducing device is a sound reproducing device. 11. The audio reproduction device according to claim 7, wherein the information amount calculation unit calculates the information amount based on an average value of an audio signal. The information amount calculating means calculates each frequency component of the audio signal, and calculates a difference between the addition average value of each frequency component value of the audio packet calculated up to the previous time and the value of each frequency component calculated this time. The audio reproduction device according to claim 7, wherein the information amount is calculated based on the information amount.

Images