CN101414463B - A kind of sound mixing coding method, device and system - Google Patents

Abstract

The invention discloses a terminal side coding method, which comprises the steps of setting a sound mixing identifier for sound information according to a sound mixing strategy, and coding the sound information according to the sound mixing identifier to obtain core coding data; if the audio mixing identification information is the audio mixing requirement, calculating dynamic side information, and generating and outputting an audio coding code stream containing the audio mixing identification, the core coding data and the dynamic side information; and if the audio mixing identification information does not need audio mixing, the terminal generates and outputs an audio coding code stream containing the audio mixing identification and the core coding data. The invention also discloses a corresponding network side audio mixing coding method, and a device and a system for audio mixing coding. The scheme of the invention can solve the problems of signal overflow and error introduction during sound mixing, and can not reduce the coding efficiency.

Description

A kind of sound mixing coding method, device and system

technical field

The present invention relates to the technical field of multimedia communication, in particular to a sound mixing encoding method, device and system.

Background technique

At present, there are more and more applications of real-time multimedia communication services, such as multimedia conferencing systems, etc., to meet growing service requirements. Therefore, various technologies related to multimedia conferencing systems are very important.

In multimedia conferencing, audio interaction is the most basic element. In a centralized conference, each terminal establishes a unicast-based connection with a Multi-point Controlling Unit (MCU), and sends and receives audio streams to and from the MCU in real time. Therefore, the input of the MCU is the audio code stream encoded by various coding schemes, and its output is the audio code stream after the mixing process according to the synthesis strategy.

Figure 1 is a schematic diagram of a multimedia conference system, where the dotted box can be regarded as an MCU unit. The input audio streams at terminal position 1, terminal position 2, etc. are respectively decoded, and the decoded audio streams are mixed by the mixing unit, and then the mixed audio streams are respectively encoded, and then output to the corresponding terminals. In the multimedia conference system shown in Figure 1, there are M terminals participating in audio mixing. For a specific time t, each terminal sends audio data to the MCU, and the MCU first decodes the audio data, calculates the mixing parameters for each signal, and finally performs mixing processing on the multiple decoded signals. The commonly used algorithm for audio mixing processing is to add all the decoded data, and then encode the added data through the encoder, and finally transmit it to each terminal.

With the above-mentioned time-domain superposition and mixing scheme, noise is often introduced. This is because the audio signal transmitted by each terminal to the MCU has a certain range [min, max], where min represents the lower limit of the range, and max represents the upper limit of the range. When adding all signals directly, it is likely to exceed the signal value range [min, max]. Due to the upper and lower limits of quantization of digital audio signals, the superposition operation is likely to cause the result to overflow. The usual processing method is to perform overflow detection, and then perform saturation operation, that is, the result exceeding the upper limit is set as the upper limit value, and the value exceeding the lower limit is set as the lower limit value. This operation itself destroys the original time-domain characteristics of the speech signal, thereby introducing noise, which is why there are pops and speech discontinuities in some systems.

As the data of the terminals participating in the mixing increases, the frequency of overflow also continues to rise. Therefore, there is an upper limit for the number of terminals in this time-domain superposition mixing solution, and the upper limit is very low. Experiments have proved that in many cases, if the When 4 terminals participate in the audio mixing, the result is a lot of noise and interruptions, and the speech flow cannot be distinguished.

Contents of the invention

In view of this, an embodiment of the present invention proposes a mixing coding method, which can overcome the noise problem of time-domain mixing coding in the prior art. Described sound mixing coding method comprises the steps:

Set the mixing flag for the sound information according to the mixing strategy, encode the sound information according to the mixing flag information, and use the encoded result as the core coded data;

If the audio mixing flag information needs to be mixed, then calculate the dynamic side information, generate and output the audio code stream that contains the audio mixing flag, core coded data and dynamic side information; if the audio mixing flag information does not need Mixing, then generate and output the audio coded stream that contains the mixing identification bit and core coded data;

The network side receives the audio coded stream from the terminal, and judges whether the audio coded stream needs to be mixed according to the mixing flag information therein, and for the M' audio coded streams that need to be mixed, according to Among them, the dynamic side information selects N channels of audio coded streams, performs mixing processing on the core coded data of the selected N channels of audio coded streams, and outputs the mixed audio coded streams, wherein N is less than or equal to M' .

The embodiment of the present invention also proposes a terminal-side encoding method, including the following steps:

Setting a mixing identifier for the sound information according to the mixing strategy, and encoding the sound information according to the mixing identifier information to obtain core coded data;

If the audio mixing identification information is that audio mixing is required, then calculate the dynamic side information, generate and output an audio coded stream containing the audio mixing identification, core coded data and dynamic side information; if the audio mixing identification information is not If audio mixing is required, the terminal generates and outputs an audio coded stream including the audio mixing identifier and core coded data.

The embodiment of the present invention also proposes a network-side audio mixing encoding method, including the following steps:

Receive M-channel audio coded streams, judge whether the audio coded streams need to be mixed according to the mixing identification information, and for the M' audio coded streams that need to be mixed, according to the dynamic side information Select N channels of audio coded streams, perform mixing processing on the core coded data of the selected N channels of audio coded streams, and output the mixed audio coded streams, where M, M' and N are all positive integers , N is less than or equal to M', and M' is less than or equal to M.

An embodiment of the present invention proposes a multimedia conference system, including M terminals and a multipoint control unit;

Including M terminals and multi-point control units, characterized in that,

The terminal is used to set the mixing identification bit for the collected sound information according to the local mixing strategy, encode the sound information according to the mixing identification bit information, and use the encoded result as core coded data; and The policy sets the mixing flag, generates and outputs the audio code stream containing the core coded data, and the mixing flag needs to be mixed and dynamic side information, or generates and outputs the core coded data and the mixing flag Encode streams for audio that does not require mixing;

The multi-point control unit is used to receive the audio coded stream from the terminal, judge whether the audio coded stream needs to be mixed according to the value of the audio mixing flag, and the M' that needs to be mixed audio code streams, N audio code streams are selected according to the dynamic side information therein, the core coded data of the selected N audio code streams are mixed, and the audio coded streams after mixing are output, wherein M, M' and N are all positive integers, N is less than or equal to M', and M' is less than or equal to M.

An embodiment of the present invention proposes a multimedia conference terminal, including:

Sound collection module, used for collecting sound information;

A sound mixing strategy module, configured to set a sound mixing identification bit for the sound information collected by the sound collection module according to a preset sound mixing strategy;

A core encoding module, configured to encode the sound information and output core encoded data;

A framing module, configured to calculate dynamic side information according to the audio mixing flag set by the audio mixing strategy module, and generate a frame containing the core coded data, the audio mixing flag, and the dynamic side information according to the value of the audio mixing flag. The audio coded data frame of the side information, or generate the audio coded data frame containing the core coded data and the mixing identification bit;

The output module is configured to output the audio coded data frame generated by the framing module as an audio coded stream.

An embodiment of the present invention proposes a multi-point control unit, including:

The selection unit is used for receiving audio coded streams from M terminals, judging whether the audio coded stream needs to be mixed according to the value of the audio coded bit stream of the audio coded stream, and performing mixing M' road audio coded streams for audio processing, select N road audio coded streams according to the dynamic side information therein;

A sound mixing unit, for performing mixing processing on the core coded data in the N-way audio coded streams selected by the selection unit, to obtain M' road-mixed audio coded streams;

A sending unit, configured to send the audio coded stream from the mixing unit to a corresponding destination terminal.

From the above technical solutions, it can be seen that on the terminal side, the audio mixing flag is calibrated in the encoded code stream and the corresponding dynamic side information is added; on the network side, the required audio mixing is selected according to the audio mixing flag and dynamic side information The audio coded bit stream is mixed, which can solve the noise problem when mixing and encoding.

Description of drawings

Fig. 1 is a schematic diagram of a multimedia conference system in the prior art;

FIG. 2 is a schematic diagram of a multimedia conference system according to an embodiment of the present invention;

3 is a structural diagram of an encoded data frame in an audio encoded code stream output by a terminal encoder unit according to an embodiment of the present invention;

Fig. 4 is the coding flowchart of the terminal side of the embodiment of the present invention;

FIG. 5 is a flow chart of audio mixing encoding on the MCU side of an embodiment of the present invention;

FIG. 6 is a block diagram of a multimedia conference terminal proposed by an embodiment of the invention;

FIG. 7 is a block diagram of a multi-point control unit proposed by an embodiment of the present invention.

Detailed ways

The embodiment of the present invention proposes a mixing encoding method based on the mixing identification bit. In addition to the core coded stream carrying the voice, the data stream output by the terminal also includes the mixing identification bit and dynamic side information, wherein the dynamic side information carries the mixing audio The information required for encoding, if the audio mixing flag is set to require audio mixing, then set the dynamic side information; if the audio mixing flag is set to not require audio mixing, then the dynamic side information will not be set. The MCU selects the core coded streams that need to be subjected to the audio mixing process according to the audio mixing flag to perform the audio mixing process.

In order to make the purpose, technical solution and advantages of the present invention clearer, the present invention will be further elaborated below in conjunction with the accompanying drawings.

Fig. 2 shows a schematic diagram of a multimedia conference system according to an embodiment of the present invention. The multimedia conference system includes M terminals, that is, terminal 1, terminal 2...terminal M; and an MCU.

Taking terminal 1 as an example, the terminal includes an encoder unit 201, which encodes the sound collected by the sound collection device of terminal 1, such as a microphone, to generate a core coded stream carrying the sound information. The encoder unit 201 also sets the audio mixing flag according to the locally set audio mixing policy. The sound mixing strategy is used to determine whether the sound coding output by the terminal needs to be mixed. Different sound mixing strategies can be set according to actual needs. For example, different priorities can be set for different terminals. The audio stream of a high-end terminal is prioritized for mixing; the sound energy threshold can also be set, and when the sound energy collected by the terminal exceeds the energy threshold, the audio stream of the terminal will be mixed and so on. And multiple mixing strategies can be used at the same time.

If the set mixing flag indicates that mixing is required, the encoder unit 201 will also generate dynamic side information and write it into the audio code stream; Only the core encoding and mixing flags are included in the code stream.

Fig. 3 shows a structural diagram of encoded data frames in an audio encoded code stream output by a terminal encoder unit according to an embodiment of the present invention. Let the total length of a data frame be n bits, when the audio mixing flag indicates that audio mixing is required, the coded data frame is shown in the upper figure in Figure 3, including the audio mixing flag of t bits, and the dynamic edge of m bits information, and a core encoding of n-m-t bits. Wherein, the audio mixing identification bit is set in the frame header, which is convenient for the MCU to identify. When the audio mixing flag indicates that audio mixing is not required, the coded data frame is shown in the lower figure of FIG. 3 , including t-bit audio-mixing flags and n-t-bit core codes.

For G.711 narrowband enhancement layer (Low Band Enhance, LBE) coding, each part in Fig. 3 can take the following values: t=1, n=80, m=9.

The side information includes: frame energy (Frame Energy) and voice score (Voicing score). If the code length of the side information is 9 bits, 6 bits are the quantized frame energy, and 3 bits are the quantized voice score.

Among them, the calculation of frame energy is expressed by formula (1):

$\mathrm{<span\; class="notranslate">\; frame</span>}<span\; class="notranslate">\; \_</span>\mathrm{<span\; class="notranslate">\; 能源</span>}<span\; class="notranslate">\; =</span>\frac{\underset{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 0</span>}}{\overset{\mathrm{<span\; class="notranslate">\; frame</span>}<span\; class="notranslate">\; \_</span>\mathrm{<span\; class="notranslate">\; Length</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}}{\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}}}{\mathrm{<span\; class="notranslate">\; S</span>}}^{\mathrm{<span\; class="notranslate">\; 2</span>}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; )</span>}{\mathrm{<span\; class="notranslate">\; frame</span>}<span\; class="notranslate">\; \_</span>\mathrm{<span\; class="notranslate">\; Length</span>}}<span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; )</span>$

Frame_Length is the frame length, S(i) is the low frequency band signal passed through the quadrature mirror filter (Quadrature MirrorFilter, QMF), and i is the sample value sequence number in the frame.

The voice score is calculated using formula (2):

$\mathrm{<span\; class="notranslate">\; Voicing</span>}<span\; class="notranslate">\; \_</span>\mathrm{<span\; class="notranslate">\; score</span>}<span\; class="notranslate">\; =</span>\frac{\mathrm{<span\; class="notranslate">\; Zero</span>}<span\; class="notranslate">\; \_</span>\mathrm{<span\; class="notranslate">\; Cros</span>}\mathrm{<span\; class="notranslate">\; sin</span>}\mathrm{<span\; class="notranslate">\; g</span>}<span\; class="notranslate">\; \_</span>\mathrm{<span\; class="notranslate">\; Rate</span>}}{\mathrm{<span\; class="notranslate">\; Scale</span>}<span\; class="notranslate">\; \_</span>\mathrm{<span\; class="notranslate">\; factor</span>}}$

Among them, the zero-crossing rate (Zero_Crossing_Rate) indicates the number of zero-crossing times of the time-domain waveform within 10ms. The reduction factor (Scale_Factor) is a preset reduction constant, and its value is [0, 1].

According to actual conditions, the dynamic side information can also be set to other quantities that can be used as a basis for judging the sound mixing process, for example, it can be set to a silent activity detection (VAD).

After the audio code stream output by the terminal is sent to the MCU, it is first input into the selection unit 202 . The selection unit 202 first recognizes the audio mixing flag from the received audio coded stream, and determines whether the audio coded stream needs to be mixed according to the value of the audio coded stream. , the selection unit 202 outputs the coded audio stream to the corresponding destination terminal. For all M' (M' is less than or equal to M) channels of audio coded streams that need to be mixed, the selection unit 202 selects N (N is less than or equal to M') channels of audio coded streams according to the dynamic side information therein, and these The audio coded streams are sent to the corresponding decoders respectively, and after being decoded, they are then sent to the mixing unit 203 for mixing processing to obtain M' mixed audio streams, and then the M' audio streams are respectively After encoding with an encoder, send it to the corresponding terminal.

The encoding process on the terminal side of the embodiment of the present invention is shown in Figure 4, including the following steps:

Step 401: Set the mixing identification bit for the collected sound information according to the local mixing strategy, and then encode the sound information, and the encoded result is used as the core encoded data;

Step 402: If the audio mixing flag is set to require audio mixing, then calculate the dynamic side information. The frame energy and sound score can be calculated according to the aforementioned formula (1) and formula (2) as the dynamic side information.

Step 403: Generate and output an audio coded stream. The generating the audio coded stream specifically includes: if the set audio mixing flag is valid, generating an audio coded data frame including the audio mixing flag, core coded data and dynamic side information; If the identification bit is invalid, an audio encoding data frame including the mixing identification bit and core encoding data is generated. The audio mixing identification bit is set at the beginning of the data frame, preferably, the length is 1 bit.

The audio mixing encoding process on the MCU side of the embodiment of the present invention is shown in Figure 5, including the following steps:

Step 501: The MCU receives the audio coded stream from the terminal, and judges whether the audio coded stream needs to be mixed according to the value of the audio mixing flag. If so, execute step 502; otherwise, execute step 503 .

Step 502: Send the coded audio stream directly to the corresponding destination terminal, and end the processing of the coded audio stream.

Step 503: For the audio coded streams received from M' terminals at the same time, and the audio mixing flags in these audio coded streams all need to be mixed, the MCU according to the dynamic side information in these coded streams , select N channels of audio coded streams, and discard the remaining M'-N channels of audio coded streams. Where N is less than or equal to M'.

According to the value of the energy in the side information, if it is greater than a certain threshold T, the sound will be mixed, and if it is smaller than it, the sound will not be mixed.

504: Decode the core coded data of the selected N channels of audio coded streams respectively, perform audio mixing processing on the decoded core coded data, and obtain M' audio coded streams after mixing.

Step 505: Encode the M' mixed audio streams respectively, and send the encoded M' encoded and mixed audio encoded streams to M' destination terminals respectively.

Fig. 6 is a kind of multimedia conferencing terminal proposed by the embodiment of the invention, including:

Sound collection module 601, for collecting sound information;

A sound mixing strategy module 602, configured to set a sound mixing identification bit for the sound information collected by the sound collection module 601 according to a preset sound mixing strategy;

The core encoding module 603 is used to encode the sound information and output the core encoded data; if the mixing strategy module 602 sets the mixing identification bit as not requiring mixing, then the core encoding module 603 does not need to consider the dynamic Bit allocation of side information; if the audio mixing flag is set to require audio mixing, the core encoding module 603 needs to consider the bit allocation of dynamic side information when performing encoding. For example, if the total number of bits of the coded data frame is n bits, the audio mixing flag is t bits, and the dynamic side information is m bits, then for the case where the bit allocation of the dynamic side information does not need to be considered, the encoding obtained by the core encoding module 603 The length of the core coded data is n-t bits; for the case where the bit allocation of dynamic side information needs to be considered, the length of the core coded data obtained by encoding by the core coding module 603 is n-m-t bits.

The framing module 604 is configured to calculate dynamic side information according to the audio mixing flag set by the audio mixing strategy module 603, and generate a frame containing the core coded data and the audio mixing flag according to the value of the audio mixing flag. and an audio data frame of dynamic side information, or generate an audio data frame comprising the core coded data and a mixing identification bit;

The output module 605 is configured to output the audio data frame generated by the framing module 604 as an audio coded stream.

Fig. 7 is a kind of multi-point control unit proposed by the embodiment of the present invention, including:

The selection unit 701 is used for receiving audio coded streams from M terminals, judging whether the audio coded stream needs to be mixed according to the value of the audio coded code stream of the audio coded stream, and performing M' channel audio coded streams processed by sound mixing, and N-channel audio coded streams are selected according to the dynamic side information therein;

The audio mixing unit 702 is used to perform mixing processing on the core coded data in the N-way audio coded streams selected by the selection unit, to obtain M' road-mixed audio streams;

The sending unit 703 is configured to send the audio code stream from the mixing unit to a corresponding destination terminal.

The selection unit 701 sends the audio coded stream that does not need to be mixed to the sending unit 703; then the sending unit 703 sends the audio coded stream from the selection unit to the corresponding destination terminal.

The multipoint control unit further includes: a decoder 704, configured to decode the core coded data in the audio coded stream selected by the selection unit 701, and send the decoded core coded data to the audio mixing Unit 702;

The encoder 705 is configured to encode the mixed audio stream from the mixing unit 702 , and send the encoded audio encoded stream to the sending unit 703 .

The solution of the embodiment of the present invention performs the marking of the audio mixing identification bit in the encoded code stream and adds the corresponding dynamic side information, and allocates the side information bits according to the audio mixing identification bit and dynamic allocation. The MCU selects the audio coded stream that needs to be mixed according to the mixing flag and dynamic side information for mixing processing, which can solve the problem of signal overflow and the introduction of errors when mixing large signals, and reduce the computational complexity of the MCU ; When not performing audio mixing, it can make full use of the code stream bit allocation to improve the core coding quality. The solution of the invention can be used in a sound mixing system, and can also be applied to a codec of a commonly used codec system, which is beneficial to realize the intelligent control of the code stream and enhance the interactivity of the MCU unit.

The above descriptions are only preferred embodiments of the present invention, and are not intended to limit the present invention. Any modifications, equivalent replacements and improvements made within the spirit and principles of the present invention should be included in the protection of the present invention. within range.

Claims (13)

1. an audio mixing coding method is characterized in that, comprises the steps:

Acoustic information is provided with the audio mixing flag according to the audio mixing strategy, according to audio mixing flag information described acoustic information is encoded, the result of coding is as the core encoder data;

If audio mixing flag information is the needs audio mixing, then calculate dynamic side information, generate and export the stream of audio codes that comprises described audio mixing flag, core encoder data and dynamic side information; If audio mixing flag information for not needing audio mixing, then generates and exports the stream of audio codes that comprises described audio mixing flag and core encoder data;

Network side receives the stream of audio codes of self terminal, judge whether that according to wherein audio mixing flag information needs carry out audio mixing to this stream of audio codes and handle, needs are carried out M ' the road stream of audio codes that audio mixing is handled, select N road stream of audio codes according to dynamic side information wherein, the core encoder data of selected N road stream of audio codes are carried out audio mixing to be handled, and the stream of audio codes behind the output audio mixing, wherein N is smaller or equal to M '.

2. method according to claim 1 is characterized in that, described dynamic side information comprises frame energy, sound score value and/or quiet motion detection.

3. method according to claim 2 is characterized in that, the dynamic side information of described calculating comprises: according to formula

Calculate the frame energy, wherein, Frame_Energy represents the frame energy, and S (i) is the low band signal through Quadrature Mirror Filter QMF, and i is the sampled value sequence number in the frame.

4. method according to claim 2 is characterized in that, the dynamic side information of described calculating comprises: according to formula

Calculate the sound score value, wherein Voicing_score represents the sound score value; Zero_Crossing_Rate represented in the schedule time, the time domain waveform zero passage number of times of described acoustic information; Scale_Factor is the reduction constant that sets in advance, and value is [0,1].

5. method according to claim 1, it is characterized in that, the information of described basis audio mixing flag wherein judges whether that needs carry out audio mixing to this stream of audio codes and handle, its judged result is handled for not needing that this stream of audio codes is carried out audio mixing, then further comprises: export described stream of audio codes to the purpose terminal.

6. according to each described method of claim 1 to 5, it is characterized in that, described core encoder data to selected N road stream of audio codes are carried out audio mixing and are handled, and the audio code stream behind the output audio mixing comprises: the core encoder data in the audio code stream of selected N road are decoded respectively, decoded N road core encoder data are carried out audio mixing to be handled, obtain the audio code stream behind the audio mixing of M ' road, audio code stream behind the audio mixing of described M ' road is encoded respectively, the stream of audio codes behind coding of the M ' road behind the coding and the audio mixing is sent to the individual purpose terminal of M ' respectively.

7. an end side coding method is characterized in that, comprises the steps:

According to the audio mixing strategy acoustic information is provided with the audio mixing sign, according to described audio mixing identification information described acoustic information being encoded obtains the core encoder data;

If described audio mixing identification information is the needs audio mixing, then calculate dynamic side information, generate and export the stream of audio codes that comprises described audio mixing sign, core encoder data and dynamic side information; If described audio mixing identification information is not for needing audio mixing, then terminal generates and exports the stream of audio codes that comprises described audio mixing sign and core encoder data.

8. the audio mixing coding method of a network side is characterized in that, comprises the steps:

Receive M road stream of audio codes, whether needs carry out audio mixing to this stream of audio codes handles according to wherein audio mixing identification information judgment, needs are carried out M ' the road stream of audio codes that audio mixing is handled, select N road stream of audio codes according to dynamic side information wherein, the core encoder data of selected N road stream of audio codes are carried out audio mixing handle, and the stream of audio codes behind the output audio mixing, wherein M, M ' and N are positive integer, N is smaller or equal to M ', and M ' is smaller or equal to M.

9. a multimedia conference system comprises M terminal and multipoint control unit, it is characterized in that,

Described terminal is used for the acoustic information collected is provided with the audio mixing flag according to the audio mixing strategy of this locality, according to audio mixing flag information described acoustic information is encoded, and the result of coding is as the core encoder data; And the audio mixing flag is set according to the audio mixing strategy of this locality, generate and output to comprise described core encoder data, audio mixing flag be to need the audio mixing and the dynamic stream of audio codes of side information, perhaps generate and export to comprise described core encoder data and audio mixing flag for not needing the stream of audio codes of audio mixing;

Described multipoint control unit is used to receive the stream of audio codes of self terminal, value according to wherein audio mixing flag judges whether that needs carry out audio mixing to this stream of audio codes and handle, needs are carried out M ' the road audio code stream that audio mixing is handled, according to selecting N road audio code stream in the dynamic side information wherein, the core encoder data of selected N road audio code stream are carried out audio mixing to be handled, and the stream of audio codes behind the output audio mixing, wherein M, M ' and N are positive integer, N is smaller or equal to M ', and M ' is smaller or equal to M.

10. a multimedia conferencing terminal is characterized in that, comprising:

The sound collecting module is used to collect acoustic information;

The audio mixing policy module is used for according to the audio mixing strategy that sets in advance the collected acoustic information of described sound collecting module being provided with the audio mixing flag;

The core encoder module is used for described acoustic information is encoded, output core encoder data;

Become frame module, be used for calculating dynamic side information according to the audio mixing flag of described audio mixing policy module setting, and according to the value of described audio mixing flag, generation comprises the coded audio data frame of described core encoder data, audio mixing flag and dynamic side information, perhaps generates the coded audio data frame that comprises described core encoder data and audio mixing flag;

Output module, the coded audio data frame that is used for the described one-tenth frame module generation of externally output is as stream of audio codes.

11. a multipoint control unit is characterized in that, comprising:

Selected cell, be used for receiving stream of audio codes from M terminal, value according to the audio mixing flag of described stream of audio codes judges whether that needs carry out audio mixing to this stream of audio codes and handle, needs are carried out M ' the road stream of audio codes that audio mixing is handled, select N road stream of audio codes according to dynamic side information wherein;

The audio mixing unit is used for that the core encoder data of the selected N of described selected cell road stream of audio codes are carried out audio mixing and handles, and obtains the stream of audio codes behind the audio mixing of M ' road;

Transmitting element is used for the stream of audio codes from described audio mixing unit is sent to the corresponding target terminal.

12. multipoint control unit according to claim 11 is characterized in that, the stream of audio codes that described selected cell will not need audio mixing to handle sends to described transmitting element; Then described transmitting element will send to the corresponding target terminal from the stream of audio codes of described selected cell.

13. according to claim 11 or 12 described multipoint control units, it is characterized in that, described multipoint control unit further comprises: demoder, be used for the core encoder data of the selected stream of audio codes of described selected cell are decoded, and decoded core encoder data are sent to described audio mixing unit;

Scrambler be used for encoding from the audio code stream behind the audio mixing of described audio mixing unit, and the stream of audio codes after will encoding sends to described transmitting element.

Images