Apparatus, method and transmitting medium for transmitting and receiving fragmented coded data

1. Field of the Invention

The present invention relates to a transmitting apparatus and method, a receiving apparatus and method, and a providing medium. More particularly, the present invention relates to a transmitting apparatus and method, a receiving apparatus and method, and a providing medium which are suitably used, for example, in the case of encoding and transmitting digital signals such as digital audio signals of multiple channels, and receiving and decoding the digital signals to reproduce them in real time on the receiving side.

2. Description of the Related Art

For example, when digital signals such as digital audio signals are transmitted from the transmitting side to the receiving side via a network, e.g., Internet, it is customary that the audio signals are highly efficiently encoded into about 1/10 to 1/100 of original data on the transmitting side, and resulting coded data is transmitted to the receiving side where the coded data is decoded to reproduce the original data.

Generally, audio signals provided via a network are monaural or stereo, and are in a state where sounds of a plurality of musical instruments and a singing voice (words) are mixed with one another. On the receiving side, therefore, a volume level of the entire sounds can be adjusted to increase or decrease, but it is difficult to adjust a volume level of the sound of certain one musical instrument or the singing voice.

However, there is a demand for providing service with high user interactivity (i.e, providing highly versatile audio signals) that enables users on the receiving side to adjust a volume level of the sound of certain one musical instrument or the singing voice, listen to only a vocal apart or an accompaniment part, or listen to the sounds of several desired musical instruments in a mixed state.

To satisfy such a demand, audio signals representing the sounds of a plurality of musical instruments, an audio signal representing the singing voice, etc. require to be transmitted separately without being mixed together. When transmitting a plurality of audio signals separately, the amount of data transmitted to the receiving side is increased in comparison with the case of transmitting those audio signals in a mixed state. For suppressing such an increase in data amount, the amount of data for each of the plurality of audio signals must be reduced. This necessarily results in deterioration of sound quality of the audio signals reproduced on the receiving side.

On the other hand, a transmission rate (usable band) of a network varies depending on the line that is employed by a user for connection to the network, the number of users accessing the transmitting side such as a server, etc. In the case of receiving coded data of a plurality of audio signals, decoding the coded data and reproducing the audio signals in real time on the receiving side, therefore, it may become difficult to decode and reproduce the audio signals in real time because transmission of the coded data is relatively delayed due to the limited transmission rate of the network.

The present invention has been made in view of the state of art set forth above, and intends to transmit a plurality of highly versatile audio signals in an adaptable manner from the transmitting side so that the audio signals are decoded and reproduced in real time on the receiving side.

According to a first aspect of the present invention, in a transmitting apparatus for transmitting coded data resulted from encoding digital signals of multiple channels for each channel, the apparatus comprises dividing means for dividing each of the digital signals of multiple channels into one or more fragments, coding means for encoding respectively the fragments of multiple channels into coded fragment data, selecting means for selecting one of the coded fragment data of multiple channels at the predetermined time, and transmitting means for transmitting the coded fragment data selected by the selecting means.

According to a second aspect of the present invention, in a transmitting method for transmitting coded data resulted from encoding digital signals of multiple channels for each channel, the method comprises a dividing step of dividing each of the digital signals of multiple channels into one or more fragments, a coding step of encoding respectively the fragments of multiple channels into coded fragment data, a selecting step of selecting one of the coded fragment data of multiple channels at the predetermined time, and a transmitting step of transmitting the coded fragment data selected by the selecting step.

According to a third aspect of the present invention, in a providing medium for providing a computer program for rendering a computer to execute a process of transmitting coded data resulted from encoding digital signals of multiple channels for each channel, the computer program includes a dividing step of dividing each of the digital signals of multiple channels into one or more fragments, a coding step of encoding each of the fragments of multiple channels into coded fragment data, a selecting step of selecting one of the coded fragment data of multiple channels at the predetermined time, and a transmitting step of transmitting the coded fragment data selected by the selecting step.

According to a fourth aspect of the present invention, in a providing medium for providing coded data resulted from encoding digital signals of multiple channels for each channel, wherein the medium provides coded fragment data that is obtained by dividing each of the digital signals of multiple channels into one or more fragments, encoding respectively the fragments of multiple channels into coded fragment data, and selecting and outputting one of the coded fragment data of multiple channels at the predetermined time.

According to a fifth aspect of the present invention, in a receiving apparatus for receiving coded data resulted from encoding digital signals of multiple channels for each channel, the apparatus comprises receiving means for receiving the coded data, separating means for separating each of the coded data into coded fragment data that has been resulted from dividing each of the digital signals of multiple channels into one or more fragments and encoding each of the fragments into the coded fragment data, and time information related to the reproduction start time when reproduction of the coded fragment data is to be started, decoding means for decoding the coded fragment data into the fragment, and control means for controlling reproduction of the fragment outputted from the decoding means in accordance with the time information.

According to a sixth aspect of the present invention, in a receiving method for receiving coded data resulted from encoding digital signals of multiple channels for each channel, the method comprises a receiving step of receiving the coded data, a separating step of separating each of the coded data into coded fragment data that has been resulted from dividing the digital signals of multiple channels into one or more fragments and encoding each of the fragments into the coded fragment data, and time information related to the reproduction start time when reproduction of the coded fragment data is to be started, a decoding step of decoding the coded fragment data into the fragment, and a control step of controlling reproduction of the fragment obtained in the decoding step in accordance with the time information.

According to a seventh aspect of the present invention, in a providing medium for providing a computer program for rendering a computer to execute a process of receiving coded data resulted from encoding digital signals of multiple channels for each channel, the computer program included a receiving step of receiving the coded data, a separating step of separating each of the coded data into coded fragment data that has been resulted from dividing the digital signals of multiple channels into one or more fragments and encoding each of the fragments into the coded fragment data, and time information related to the reproduction start time when reproduction of the coded fragment data is to be started, a decoding step of decoding the coded fragment data into the fragment, and a control step of controlling reproduction of the fragment obtained from the decoding step in accordance with the time information.

According to an eighth aspect of the present invention, in a receiving apparatus for receiving coded data resulted from encoding digital signals of multiple channels for each channel, the apparatus comprises receiving means for receiving coded fragment data that is obtained by dividing each of the digital signals of multiple channels into one or more fragments, encoding respectively the fragments of multiple channels into coded fragment data, and selecting and transmitting one of the coded fragment data of multiple channels at the predetermined time; and decoding means for decoding the coded fragment data received by the receiving means into the fragment.

According to a ninth aspect of the present invention, in a receiving method for receiving coded data resulted from encoding digital signals of multiple channels for each channel, the method comprises a receiving step of receiving coded fragment data that is obtained by dividing each of the digital signals of multiple channels into one or more fragments, encoding respectively the fragments of multiple channels into coded fragment data, and selecting and transmitting one of the coded fragment data of multiple channels at the predetermined time; and a decoding step of decoding the coded fragment data received by the receiving means into the fragment.

According to a tenth first aspect of the present invention, in a providing medium for providing a computer program for rendering a computer to execute a process of receiving coded data resulted from encoding digital signals of multiple channels for each channel, the computer program includes a receiving step of receiving coded fragment data that is obtained by dividing each of the digital signals of multiple channels into one or more fragments, encoding respectively the fragments of multiple channels into coded fragment data, and selecting and transmitting one of the coded fragment data of multiple channels at the predetermined time; and a decoding step of decoding the coded fragment data received by the receiving means into the fragment.

FIG. 1 is a diagram showing an exemplified construction of one embodiment of a transmission system to which the present invention is applied.

FIG. 2 is a block diagram showing an exemplified construction of hardware of a server 1 in FIG. 1 .

FIG. 3 is a block diagram showing an exemplified construction of hardware of a client terminal 3 in FIG. 1 .

FIG. 4 is a block diagram showing an exemplified functional construction of the server 1 in FIG. 2 .

FIG. 5 is a diagram showing a format of data outputted from a controller 42 in FIG. 4 .

FIG. 6 is a block diagram showing an exemplified functional construction of the client terminal 3 in FIG. 3 .

FIGS. 7A to 7 C are diagrams for explaining processing executed by an audio signal output circuit 63 in FIG. 6 .

FIG. 8 is a flowchart for explaining a first embodiment of a transmitting process executed by the server 1 .

FIG. 9 is a chart for explaining a fragment selecting process executed in step S 5 of FIG. 8 .

FIG. 10 is a flowchart for explaining a first embodiment of the fragment selecting process executed in step S 5 of FIG. 8 .

FIG. 11 is a chart for explaining processing executed in step S 11 of FIG. 10 .

FIGS. 12A to 12 G are charts for explaining the fragment selecting process in FIG. 10 .

FIG. 13 is a flowchart for explaining a second embodiment of the fragment selecting process executed in step S 5 of FIG. 8 .

FIGS. 14A to 14 G are charts for explaining the fragment selecting process in FIG. 13 .

FIG. 15 is a flowchart for explaining a third embodiment of the fragment selecting process executed in step S 5 of FIG. 8 .

FIGS. 16A to 16 I are charts for explaining the fragment selecting process in FIG. 15 .

FIG. 17 is a flowchart for explaining a fourth embodiment of:the fragment selecting process executed in step S 5 of FIG. 8 .

FIGS. 18A to 18 E are charts for explaining the fragment selecting process in FIG. 17 .

FIGS. 19A to 19 I are charts for explaining the fragment selecting process executed in step S 5 of FIG. 8 when the client terminal 3 has a buffer 61 A.

FIG. 20 is a flowchart for explaining a second embodiment of the transmitting process executed by the server 1 .

FIG. 21 is a flowchart for explaining a fragment selecting process executed in step S 65 of FIG. 20 .

FIGS. 22A to 22 F are charts for explaining the fragment selecting process in FIG. 21 .

FIG. 23 is a flowchart for explaining a reproducing process executed by the client terminal 3 .

Embodiments of the present invention will be described below.

FIG. 1 shows an exemplified construction of one embodiment of a transmission system (the term “system” means a plurality of devices assembled together in a logical correlation regardless of whether the devices having their own constructions are all positioned in the same housing or not) to which the present invention is applied.

In the transmission system, when a request for audio signals of a song, for example, is issued from a client terminal 3 to a server 1 via a network 2 such as Internet, ISDN (Integrated Service Digital Network) or PSTN (Public Switched Telephone Network), the server 1 encodes the audio signals corresponding to the requested song with a predetermined coding method, and resulting coded data is transmitted to the client terminal 3 via the network 2 . After receiving the coded data from the server 1 , the client terminal 3 decodes the coded data and reproduces the original audio signals in real time, for example, (so-called streaming reproduction).

FIG. 2 shows an exemplified construction of hardware of the server 1 in FIG. 1 .

A ROM (Read Only Memory) 11 stores, for example, an IPL (Initial,Program Loading) program and so on. A CPU (Central Processing Unit) 12 executes an OS (Operating System), which is stored (recorded) in an external storage 16 , in accordance with the IPL program stored in the ROM 11 , and also executes various application programs, which are stored in the external storage 16 , under control of the OS, thereby carrying out a coding process of audio signals, a transmitting process of coded data resulted from the coding process to the client terminal 3 , etc. A RAM (Random Access Memory) 13 stores programs, data and so on which are necessary for operation of the CPU 12 . An input device 14 is constructed of, e.g., a keyboard, a mouse, a microphone or an external interface, and is operated when necessary data or command is inputted. The input device 14 is also constructed to function as an interface for accepting input of digital audio signals externally applied to the client terminal 3 . An output device 15 is constructed of, e.g., a display, a speaker or a printer, and displays or outputs necessary information. The external storage 16 comprises a hard disk, for example, and stores the OS and the application programs mentioned above. The external storage 16 also stores other data including data necessary for operation of the CPU 22 , etc. A communicating apparatus 17 carries out control required for communication via the network 2 .

FIG. 3 shows an exemplified construction of hardware of the client terminal 3 in FIG. 1 .

The client terminal 3 is made up of illustrated components 21 - 27 , i.e, a ROM 21 to a communicating apparatus 27 , and has a construction basically similarly to that of the server 1 made up of the ROM 11 to the communicating apparatus 17 .

Unlike the server 1 , an external storage 26 stores, as application programs, a program for decoding the coded data transmitted from the server 1 , programs for executing later-described processing, and so on. A CPU 22 executes those application programs to, for example, decode the coded data and perform a reproducing process.

FIG. 4 shows an exemplified functional construction of the server 1 in FIG. 2 . The illustrated construction is realized upon the CPU 12 executing the application programs stored in the external storage 16 .

Audio signals to be provided to the client terminal 3 are supplied to an audio signal input circuit 31 . In the audio signal input circuit 31 , analog audio signals are A/D-converted into digital audio signals. Then, the audio signal input circuit 31 separates the digital audio signals for each of channels, for example, and divides the audio signal of each channel into fragments each having a predetermined length.

More specifically, audio signals of a piece of music, for example, are separated for each of respective parts of guitar, piano, drum, song (words), etc. that constitute the piece of music together, and are inputted to the audio signal input circuit 31 in a synchronous relation. The audio signal input circuit 31 allocates a predetermined channel to the audio signal of each part. Further, the audio signal input circuit 31 divides the audio signal of each of the multiple channels into fragments each having a predetermined length, and then supplies the fragments to a coding circuit 32 .

Note that audio signals of two or more parts may be inputted to the audio signal input circuit 31 in a mixed condition. In this case, the audio signal input circuit 31 outputs fragments of the audio signals in the mixed condition.

Also, audio signals inputted to the audio signal input circuit 31 may be in the form of, e.g., MIDI (Musical Instrument Digital Interface) data rather than waveforms of the actual audio signals themselves.

The audio signal input circuit 31 can be constructed to divide each of digital signals of multiple channels into fragments each having a fixed length or a variable length. When dividing an audio signal into fragments each having a variable length, the fragments can be produced by detecting, for example, a soundless portion of the audio signal or a portion thereof where power (square sum of amplitude) is not larger than a predetermined value), and splitting the audio signal at every such a portion.

Alternatively, the audio signal may be divided into fragments having the same lengths for all the multiple channels (i.e, if fragments of the audio signal of one channel have lengths of L 1 , L 2 , etc. starting from the head, fragments of the audio signal of any other channel may also have lengths of L 1 , L 2 , etc. starting from the head).

Further, the audio signal may be divided into fragments having independently different lengths for each of the channels. In other words, fragments of the audio signal of each channel may have different lengths depending on the channel.

The coding circuit 32 encodes the audio signals from the audio signal input circuit 31 at high efficiency.

More specifically, in the embodiment shown in FIG. 4, the coding circuit 32 comprises a number M of encoders 41 1 to 41 M , a controller 42 , and a multiplexer 43 .

The encoder 41 m (m=1, 2, . . . , M) is supplied with fragments of the audio signal of each corresponding channel (part) from the audio signal input circuit 31 . The encoder 41 m encodes the fragments of the audio signal from the audio signal input circuit 31 with a predetermined coding method, and supplies resulting data (referred to as coded fragment data hereinafter) to the controller 42 .

Here, examples of the coding method for the audio signal include linear PCM (Pulse Code Modulation), ADPCM (Adaptive Differential PCM), layers 1 , 2 , 3 of MPEG (Moving Picture Experts Group), ATRAC (Adaptive Transform Acoustic Coding), and ATRAC 2 . Incidentally, the coding methods employed in the encoders 41 1 to 41 M may be the same or different from one another.

The controller 42 starts reproducing the fragments by decoding the decoded fragment data supplied from the encoders 41 1 to 41 M . For example, the controller 42 adds the time with the head of a piece of music set to a reference (referred to as the reproduction start time hereinafter) to the corresponding coded fragment data. Furthermore, the controller 42 adds necessary control information to the coded fragment data, followed by supplying the coded fragment data to the multiplexer 43 .

The multiplexer 43 selects one of the coded fragment data of respective channels, supplied from the controller 42 , at the predetermined timing (time) based on importance indexes representing levels of importance of the respective channels set in an importance setting unit 34 , the reproduction start time added to each coded fragment data, the data amount of each coded fragment data, etc. Stated otherwise, at the time (counted from a reference set to, e.g., the time to start transmission of the head of music) when transmission of the previously selected coded fragment data is ended (referred to as the transmission end time hereinafter), the multiplexer 43 selects one of the coded fragment data of respective channels supplied from the controller 42 , and outputs the selected coded fragment data (referred to as the selected fragment data hereinafter) to a transmitting circuit 33 .

The transmitting circuit 33 converts the selected fragment data supplied from the multiplexer 43 into a format corresponding to the communication protocol adapted for the network 2 , and transmits resulting coded data to the client terminal 3 via the network 2 .

The importance indexes representing levels of importance of the respective channels for the audio signals coded by the encoders 41 1 to 41 M are set in the importance setting unit 34 . This setting is performed by, e.g., a person managing the server 1 who manipulates the importance setting unit 34 . Specifically, in the case of a piece of music comprising parts of guitar, piano, drum and song (words), for example, if it is thought that the parts of song, piano, guitar and drum have levels of importance descending in the order named, the importance setting unit, 34 is operated so as to set importance indexes to the channels corresponding to the parts of song, piano, guitar and drum in match with this order. Rather than relying on the operation of the importance setting unit 34 , the levels of importance of the respective channels may be set based on the magnitudes of power of the audio signals of respective channels or the number of soundless portions thereof. When the levels of importance of the respective channels are set with the operation of the importance setting unit 34 , they will not change until the next operation. On the other hand, when the levels of importance of the respective channels are set based on the magnitudes of power of the audio signals or the number of soundless portions thereof, they may change during transmission.

FIG. 5 shows a format of data outputted from the controller 42 in FIG. 4 .

As shown in FIG. 5, the controller 42 adds the relevant reproduction start time, control information, etc, as a header, to the coded fragment data outputted from the encoder 41 m , followed by outputting resulting coded data.

FIG. 6 shows an exemplified functional construction of the client terminal 3 in FIG. 3 . The illustrated construction is realized upon the CPU 22 executing the application programs stored in the external storage 26 .

The coded data transmitted from the server 1 via the network 2 is supplied to a receiving circuit 61 . After receiving the coded data, the receiving circuit 61 performs format conversion on the received coded data corresponding to the format conversion made in the transmitting circuit 33 in FIG. 4, and then supplies resulting coded data to a decoding circuit 62 . The decoding circuit 62 decodes the coded data supplied from the receiving circuit 61 .

More specifically, the decoding circuit 62 comprises a separator 71 , a number M of decoders 72 1 to 72 M , and a reproduction timing adjuster 73 . While the number of decoders 72 1 to 72 M constituting the decoding circuit 62 is set here to be the same as the number of encoders 41 1 to 41 M constituting the coding circuit 32 in FIG. 4, it is not always needed that both the numbers are identical to each other. If the number of decoders is less than the number of encoders, the audio signals of some one ore more channels (parts) are neither decoded nor reproduced in the client terminal 3 . Further, the decoders 72 1 to 72 M are here constructed to be, for example, capable of decoding the data coded respectively in the encoders 41 1 to 41 M in FIG. 4 .

The coded data outputted from the receiving circuit 61 is supplied to the separator 71 . The separator 71 separates the coded data into the coded fragment data and the header, followed by supplying the coded fragment data to corresponding channels of the decoders 72 1 to 72 M , respectively. In addition, the separator 71 supplies the reproduction start time placed in the header to the reproduction timing adjuster 73 , and also supplies the control information placed in the header to an audio signal output circuit 63 .

The decoder 72 m decodes the coded fragment data from the separator 71 into the original fragments which are supplied to the reproduction timing adjuster 73 . The reproduction timing adjuster 73 supplies the original fragments of the audio signal to the audio signal output circuit 63 while adjusting the output timing to the audio signal output circuit 63 so that the fragments of the audio signal supplied from the decoder 72 m are outputted (reproduced) from the audio signal output circuit 63 at the reproduction start time applied from the separator 71 .

The audio signal output circuit 63 performs D/A conversion of the fragments of the audio signal of each channel supplied from the reproduction timing adjuster 73 , and outputs resulting audio signals after mixing if required. The audio signals outputted from the audio signal output circuit 63 are then outputted from, e.g., a speaker constituting the output device 25 (FIG. 3 ).

FIGS. 7A to 7 C show exemplified constructions of the audio signal output circuit 63 in FIG. 6 .

In an embodiment of FIG. 7A, the audio signal output circuit 63 is constructed such that a number N (<M) of audio signals (fragments) S 1 to S N outputted from the reproduction timing adjuster 73 are outputted as audio signals ch 1 to ch N , respectively, as they are.

In an embodiment of FIG. 7B, the audio signal output circuit 63 has a mixer 63 A. The mixer 63 A mixes the number N of audio signals S 1 to S N outputted from the reproduction timing adjuster 73 in accordance with the control information from the separator 71 , and outputs resulting audio signals ch 1 and ch 2 . The audio signals ch 1 and ch 2 correspond to, for example, an L (left) channel and an R (right) channel, respectively. In this case, the control information includes information indicating from which one of the L and R channels each of the audio signals S 1 to S N is to be outputted. Further, in this case, the control information may include information for panning each of the audio signals S 1 to S N . Specifically, the control information may include information indicating at what percentage the audio signal S n (n=1, 2, . . . , N) is to be outputted from each of the L and R channels (i.e., indicating, for example, that the audio signal S n (n=1, 2, . . . , N) is outputted at levels of 30% and 70% from the L and R channels, respectively.

In an embodiment of FIG. 7C, the audio signal output circuit 63 has a matrix switcher 63 B. The matrix switcher 63 B is constructed so as to output each of the number N of audio signals S 1 to S N outputted from the reproduction timing adjuster 73 as any of audio signals ch 1 to ch N in accordance with the control information from the separator 71 . In this case, the control information includes information indicating that each of the audio signals S 1 to S N is to be outputted as which one of the audio signals ch 1 to ch N .

Rather than supplying the control information as being included in the coded data, the audio signal output circuit 63 may be given with the control information that is entered upon the user operating the input device 24 (FIG. 3 ). In this case, the user can freely set how the audio signals are outputted from the audio signal output circuit 63 .

Next, a transmitting process of the audio signals (in the coded form) executed by the server 1 will be described with reference to a flowchart of FIG. 8 .

Upon receiving a request for a piece of music from the client terminal 3 via the network 2 , the server 1 executes the transmitting process in accordance with the flowchart of FIG. 8, whereby audio signals of respective parts of the piece of music requested from the client terminal 3 are transmitted, as audio signals of separate independent channels, to the client terminal 3 via the network 2 so that they can be decoded and reproduced in real time.

More specifically, in the case of transmitting the audio signals of respective parts as the audio signals of separate independent channels without mixing them, as mentioned above, the amount of data is increased as compared with the case of transmitting the audio signals after mixing them. Accordingly, if the transmission rate of the network 2 is insufficient with respect to the amount of data, transmission of the audio signals of multiple channels is delayed relative to the start of reproduction thereof and the audio signals can no longer be reproduced in real time.

On the other hand, if the data amount of the audio signals of multiple channels is reduced so that the transmission rate of the network 2 is not insufficient with respect to the data amount of the audio signals of multiple channels transmitted, the audio signals having deteriorated sound quality are supplied to the client terminal 3 even when the amount of traffic in the network 2 is small and a large (high) transmission rate can be secured.

In view of the above, the embodiment ensures real-time reproduction of audio signals in the client terminal 3 by adaptively selecting some of fragments of audio signals of multiple channels which is transmitted to the client terminal 3 , i.e., by making selection such that the fragments of an audio signal of at least one channel having high importance are always transmitted while transmission of the fragments of audio signals of other channels is given up (or abandoned) as the occasion requires, when the transmission rate of the network 2 is insufficient (i.e., when real-time reproduction can no longer be realized if audio signals of all channels are transmitted).

More specifically, in the server 1 (FIG. 4 ), audio signals of respective parts making up a piece of music requested from the client terminal 3 are inputted to the audio signal input circuit 31 . In the audio signal input circuit 31 , the audio signals of respective parts inputted thereto are A/D-converted into digital audio signals of predetermined respective channels. Further, the audio signal input circuit 31 divides the audio signal of each channel into fragments each having a predetermined length and supplies the fragments to the coding circuit 32 . In the coding circuit 32 , the encoder 41 m encodes the fragments of the audio signal of the corresponding channel and supplies resulting coded fragment data to the controller 42 .

The controller 42 adds the relevant reproduction start time and other necessary control information to the coded fragment data supplied from the encoders 41 1 to 41 M , and supplies resulting data to the multiplexer 43 .

Upon receiving the coded fragment data (added with headers) from the controller 42 for the first time after a request for the piece of music has been issued from the client terminal 3 , the multiplexer 43 chooses, as selected fragment data, one of the coded fragment data that belongs to the channel set to the highest importance in the importance setting unit 34 and has the earliest reproduction start time, followed by supplying the selected fragment data to the transmitting circuit 33 . Correspondingly, in step S 1 of FIG. 8, the transmitting circuit 33 transmits the selected fragment data from the multiplexer 43 , i.e., the coded fragment data that, in this case, belongs to the channel set to the highest importance in the importance setting unit 34 and has the earliest reproduction start time, to the client terminal 3 via the network 2 .

The control flow then goes to step S 2 to determine whether transmission of the selected fragment data is fully completed. If the transmission of the selected fragment data is not completed, the control flow returns to step Si, causing the transmitting circuit 33 to continue transmitting the selected fragment data.

If it is determined in step S 2 that the transmission of the selected fragment data is completed, the control flow goes to step S 3 where the multiplexer 43 detects the transmission end time T e when the transmission of the selected fragment data is completed. The control flow then goes to step S 4 to determine whether any coded fragment data to be transmitted still exists. If it is determined in step S 4 that the coded fragment data to be transmitted still exists, the control flow goes to step S 5 to execute a fragment selecting process in which the multiplexer 43 selects the coded fragment data to be next transmitted to the transmitting circuit 33 . When the fragment selecting process is executed by the multiplexer 43 and the selected fragment data is given by selection of the coded fragment data, the control flow returns to step Si after supplying the selected fragment data to the transmitting circuit 33 . In the transmitting circuit 33 , the selected fragment data from the multiplexer 43 is transmitted to the client terminal 3 via the network 2 . Thereafter, the control flow goes to step S 2 to repeat the processing described above.

On the other hand, if it is determined in step S 4 that there is no coded fragment data to be transmitted, the transmitting process is ended.

The fragment selecting process executed in step S 5 of FIG. 8 will now be described in more detail.

It is here assumed that the piece of music requested from the client terminal 3 is made up of audio signals A, B, C, D of four channels, for example, and the importance setting unit 34 sets levels of importance descending in the order named, as shown in FIG. 9 . In an example of FIG. 9, therefore, the channel of the audio signal A has the highest importance and the channel of the audio signal D has the lowest importance. The channel having the highest importance channel (here the channel of the audio signal A) will be referred to also as the most importance channel hereinafter.

In the example of FIG. 9, the audio signal A is divided into two fragments A 1 and A 2 ; the audio signal B is divided into two fragments B 1 and B 2 ; the audio signal C is divided into four fragments C 1 , C 2 , C 3 and C 4 ; and the audio signal D is divided into three fragments D 1 , D 2 and D 3 . The fragments A 1 , B 1 , C 1 and D 1 have the same reproduction start time, and the fragments A 2 , B 2 , C 2 , C 3 , C 4 , D 2 and D 3 have respectively the reproduction start time T a2 , Tb b2 , T c2 , T c3 , T c4 , T d2 and t d3 . The relation of T c2 <T d2 <T b2 <T c3 <T a2 >t d3 <T c4 is set in the example of FIG. 9 .

As a first embodiment of the fragment selecting process, the multiplexer 43 is constructed to preferentially choose, as the selected fragment data, ( 1 ) the coded fragment data having the earlier reproduction start time, and ( 2 ) the coded fragment data having higher importance if the reproduction start time is the same. In this case, the fragment selecting process in step S 5 of FIG. 8 is executed in accordance with, for example, a flowchart shown in FIG. 10 .

In the fragment selecting process of this case, a candidate for the selected fragment data is first chosen in step S 11 from among the coded fragment data not yet transmitted for each of the channels except the most important channel. Specifically, as shown in FIG. 11 by way of example, assuming that the transmission end time of the coded fragment data previously transmitted as the selected fragment data is T e , the reproduction start time of the coded fragment data not yet transmitted for each of the channels except the most important channel is T s , the period of time required for transmitting that coded fragment data is T t , and the reproduction start time of the coded fragment data of the most important channel to be next transmitted (in FIG. 9, for example, if the fragment A 1 has been already transmitted, the fragment A 2 is the fragment of the most important channel to be next transmitted) is T sm , the coded fragment data, for which the reproduction start time T s is after (later than) the transmission end time T e and the period of time d (=T s −T e ) from the reproduction start time T s to the transmission end time T e is minimum, is chosen as the selected fragment data candidate.

The control flow then goes to step S 12 where, from among the selected fragment data candidates chosen from the respective channels except the most important channel, one for which the period of time d from the reproduction start time T s to the transmission end time T e is minimum is chosen as temporary selected fragment data.

Subsequently, the control flow goes to step S 13 to determine whether two or more of the selected fragment data candidates have the same period of time d. If it is determined in step S 13 that two or more of the selected fragment data candidates have the same period of time d, namely, if the coded fragment data are chosen as the temporary selected fragment data from two or more channels in step S 12 , the control flow goes to step S 14 where one of the coded fragment data chosen from the two or more channels, which belongs to the channel having the highest importance, is finally chosen as the temporary selected fragment data, followed by going to step S 15 .

If it is determined in step S 13 that no selected fragment data candidates have the same period of time d, namely, if the coded fragment data is chosen as the temporary selected fragment data from one channel in step S 12 , that coded fragment data is finally chosen as the temporary selected fragment data, following which the control flow goes to step S 15 while skipping step S 14 . In step S 15 , the transmission end time of the temporary selected fragment data is estimated on an assumption that it is transmitted (the estimated time will be referred to as the estimated end time hereinafter), and whether the estimated end time is before the reproduction start time T s of the temporary selected fragment data is determined.

Here, the estimated end time of the temporary selected fragment data is derived, for example, by adding the period of time T t required for transmitting the temporary selected fragment data to the transmission end time T e , i.e., (T t +T e ).

Also, the period of time T t required for transmitting the temporary selected fragment data is derived, for example, by dividing the data amount of the temporary selected fragment data by the transmission rate of the network 2 , i.e., ([data amount]/[transmission rate]).

The transmission rate of the network 2 can be derived, for example, as follows.

Generally, transmission between the server 1 and the client terminal 3 is performed such that a request for data is issued from the client terminal 3 , and in response to the request the server 1 transmits the data, or that the server 1 transmits data, and upon receiving the data, the client terminal 3 transmits an acknowledge signal ACK to the server 1 .

By using, for example, the time t 1 when the server 1 receives a request for data from the client terminal 3 , the bit amount n 1 of data transmitted from the server 1 in response to the request, the time t 2 when the server 1 receives a next request for data from the client terminal 3 , and the bit amount n 2 of data transmitted from the client terminal 3 along with the next request, therefore, the transmission rate B (bps) of the network 2 can be derived from the following formula:

B=(n 1 +n 2 )/(t 2 −t 1 )

The transmission rate B of the network 2 may be derived only upon the fragment selecting process of step S 15 is first executed after the start of communication, and the initially derived transmission rate B may be used until the end of the communication. It is however desired that the transmission rate B of the network 2 be derived whenever the fragment selecting process of step S 15 is first executed, because the transmission rate B basically changes with time depending on the traffic condition.

Where the network 2 comprises Internet, for example, the protocol called RTP (Real-time Transport Protocol), which is specified in RFC (Request For Comments) 1889 (issued by IFTE (Internet Engineering Task Force)), is usable. In this case, since the transmission time is assigned to each packet for use in RTP, the transmission rate B can be derived based on the packet transmission time.

Also, where a protocol of reserving the transmission rate (band) to be used as with RSVP (Resource Reservation Protocol) specified in RFC 2205 (issued by IFTF), for example, the reserved transmission rate can be used as it is.

If it is determined in step S 15 that the estimated end time, at which the transmission of the temporary selected fragment data is to be completed, is before (prior to in time base) the reproduction start time T s of the same, namely, if the transmission of the temporary selected fragment data is to be completed until the reproduction start time T s , the control flow goes to step S 16 to determine whether the estimated end time, at which the transmission of the temporary selected fragment data is to be completed, is before the reproduction start time T sm of the fragment of the most important channel to be next transmitted.

If it is determined in step S 16 that the estimated end time of transmission of the temporary selected fragment data is before the reproduction start time T sm of the fragment of the most important channel to be next transmitted, namely, if the transmission of the temporary selected fragment data is to be completed until the reproduction start time T sm of the fragment of the most important channel to be next transmitted, the control flow goes to step S 17 where the temporary selected fragment data is decided as the selected fragment data, followed by returning to the main control flow.

On the other hand, if it is determined in step S 15 that the estimated end time of transmission of the temporary selected fragment data is not before the reproduction start time T s of the same, namely, if the transmission of the temporary selected fragment data is not to be completed until the reproduction start time T s , the control flow goes to step S 18 to determine whether any of the selected fragment data candidates, which are chosen in step S 11 , remains not yet finally chosen as the temporary selected fragment data. If it is determined in step S 18 that any selected fragment data candidate remains not yet finally chosen as the temporary selected fragment data, the control flow returns to step S 12 , following which the above-described processing subsequent to step S 12 is repeated on the candidate (for the selected fragment data) not yet chosen as the temporary selected fragment data.

Also, if it is determined in step S 16 that the estimated end time of transmission of the temporary selected fragment data is not before the reproduction start time T sm of the fragment of the most important channel to be next transmitted, namely, if the transmission of the temporary selected fragment data is to be completed until the reproduction start time T sm of the fragment of the most important channel to be next transmitted, the control flow goes to step S 18 , following which the same processing as described above is carried out.

In this embodiment, since transmission of fragments of the most important channel is essential to ensure real-time reproduction of audio signals in the client terminal 3 as mentioned above, transmission of the selected fragment data must be completed before the reproduction start time of the fragment of the most important channel to be next transmitted. Accordingly, if the transmission of the temporary selected fragment data must be completed until the reproduction start time T sm of the fragment of the most important channel to be next transmitted, transmitting the temporary selected fragment data as the selected fragment data is abandoned, and it is checked whether any other candidate can become the selected fragment data.

Additionally, if transmission of the coded fragment data is started until the reproduction start time thereof, real-time reproduction of audio signals can be made theoretically. Taking an allowance into account, however, one condition for choosing the coded fragment data as the selected fragment data is set here to that the transmission of the coded fragment data is completed until the reproduction start time thereof.

On the other hand, if it is determined in step S 18 that there remains no selected fragment data candidate which is not yet finally chosen as the temporary selected fragment data, the control flow returns to step S 19 where the coded fragment data of the most important channel to be next transmitted is decided as the selected fragment data, followed by returning to the main control flow.

With the fragment selecting process of FIG. 10, the audio signals A to D of four channels shown in FIG. 9 are transmitted as shown in FIGS. 12B to 12 G.

More specifically, as described above in connection with step S 1 of FIG. 8, since the coded fragment data of the most important channel, which has the earliest reproduction start time, is given as the selected fragment data at the first, the coded fragment data A 1 is first transmitted as shown in FIG. 12B (step S 1 ).

After transmitting the coded fragment data A 1 , the coded fragment data B 2 , C 2 and D 2 , for each of which the reproduction start time T s is after the transmission end time T e of the coded fragment data A 1 and the period of time from the reproduction start time T s to the transmission end time T e is minimum, are chosen as candidates for the selected fragment data from the respective channels except the most important channel (step S 11 ). At this point in time, therefore, transmission of the coded fragment data B 1 , C 1 and D 1 is abandoned (given up) because the transmission of these data would be too late for the start of reproduction thereof.

Of the selected fragment data candidates B 2 , C 2 and D 2 , only C 2 of which reproduction starts at the time T c2 has the minimum period of time d (=T s −T e ) from the reproduction start time T s thereof to the transmission end time T e of the coded fragment data A 1 as seen from FIGS. 12A and 12B. Therefore, the coded fragment data C 2 is finally chosen as the temporary selected fragment data (steps S 12 to S 14 ).

On an assumption that the temporary selected fragment data C 2 is transmitted, as seen from FIG. 12C, the estimated end time of transmission of the temporary selected fragment data C 2 is before not only the reproduction start time T c2 thereof but also the reproduction start time T a2 of the coded fragment data A 2 of the most important channel to be next transmitted (steps S 15 and S 16 ). Accordingly, the temporary selected fragment data C 2 is decided as the selected fragment data (step S 17 ) and then transmitted (step S 1 ).

After transmitting the coded fragment data C 2 , the coded fragment data B 2 , C 3 and D 2 , for each of which the reproduction start time T s is after the transmission end time T e of the coded fragment data C 2 and the period of time from the reproduction start time T s to the transmission end time T e is minimum, are chosen as candidates for the selected fragment data from the respective channels except the most important channel (step S 11 ).

Of the selected fragment data candidates B 2 , C 3 and D 2 , only D 2 of which reproduction starts at the time T d2 has the minimum period of time d (=T s −T e ) from the reproduction start time T s there of to the transmission end time T e of the coded fragment data C 2 as seen from FIGS. 12A and 12C. Therefore, the coded fragment data D 2 is finally chosen as the temporary selected fragment data (steps S 12 to S 14 ).

On an assumption that the temporary selected fragment data D 2 is transmitted, as seen from FIG. 12D, the estimated end time of transmission of the temporary selected fragment data D 2 is before not only the reproduction start time T d2 thereof but also the reproduction start time T a2 of the coded fragment data A 2 of the most important channel to be next transmitted (steps S 15 and S 16 ). Accordingly, the temporary selected fragment data D 2 is decided as the selected fragment data (step S 17 ) and then transmitted (step S 1 ).

After transmitting the coded fragment data D 2 , the coded fragment data B 2 , C 3 and D 3 , for each of which the reproduction start time T s is after the transmission end time T e of the coded fragment data D 2 and the period of time from the reproduction start time T s to the transmission end time T e is minimum, are chosen as candidates for the selected fragment data from the respective channels except the most important channel (step S 11 ).

Of the selected fragment data candidates B 2 , C 3 and D 3 , only B 2 of which reproduction starts at the time T b2 has the minimum period of time d (=T s −T e ) from the reproduction start time T s thereof to the transmission end time T e of the coded fragment data D 2 as seen from FIGS. 12A and 12D. Therefore, the coded fragment data B 2 is finally chosen as the temporary selected fragment data (steps S 12 to S 14 ).

On an assumption that the temporary selected fragment data B 2 is transmitted, as seen from FIG. 12E, the estimated end time of transmission of the temporary selected fragment data B 2 is not before the reproduction start time T b2 thereof (step S 15 ). Then, of the selected fragment data candidates B 2 , C 3 and D 3 , the candidates C 3 and D 3 , which are not yet finally chosen as the temporary selected fragment data, are checked and subjected to the same processing as described above (step S 18 ).

In this case, of the selected fragment data candidates C 3 and D 3 , only C 3 of which reproduction starts at the time t c3 has the minimum period of time from the reproduction start time T s thereof to the transmission end time T e of the coded fragment data D 2 as seen from FIGS. 12A and 12D. Therefore, the coded fragment data C 3 is finally chosen as the temporary selected fragment data (steps S 12 to S 14 ).

On an assumption that the temporary selected fragment data C 3 is transmitted, as seen from FIG. 12F, the estimated end time of transmission of the temporary selected fragment data C 3 is before not only the reproduction start time T c3 thereof but also the reproduction start time T a2 of the coded fragment data A 2 of the most important channel to be next transmitted (steps S 15 and S 16 ). Accordingly, the temporary selected fragment data C 3 is decided as the selected fragment data (step S 17 ) and then transmitted (step S 1 ).

After transmitting the coded fragment data C 3 , the coded fragment data C 4 and D 3 , for each of which the reproduction start time T e is after the transmission end time T e of the coded fragment data C 3 and the period of time from the reproduction start time T s to the transmission end time T e is minimum, are chosen as candidates for the selected fragment data from the respective channels except the most important channel (step S 11 ).

Of the selected fragment data candidates C 4 and D 3 , only D 3 of which reproduction starts at the time t d3 has the minimum period of time d (=T s −T e ) from the reproduction start time T s thereof to the transmission end time T e of the coded fragment data C 3 as seen from FIGS. 12A and 12F. Therefore, the coded fragment data D 3 is finally chosen as the temporary selected fragment data (steps S 12 to S 14 ).

On an assumption that the temporary selected fragment data D 3 is transmitted, the estimated end time of transmission of the temporary selected fragment data D 3 is not before the reproduction start time t d3 thereof (step S 15 ). Then, of the selected fragment data candidates C 4 and D 3 , the candidate C 4 , which is not yet finally chosen as the temporary selected fragment data, is checked and subjected to the same processing as described above (step S 18 ). Thus, in this case, the coded fragment data C 4 is finally chosen as the temporary selected fragment data. On an assumption that the temporary selected fragment data C 4 is transmitted, the estimated end time of transmission of the temporary selected fragment data C 4 is before the reproduction start time T c4 thereof (step S 15 ), but not before the reproduction start time T a2 of the coded fragment data A 2 of the most important channel to be next transmitted (step S 16 ). Further, there remains no candidate which is not yet finally chosen as the temporary selected fragment data (step S 18 ). Accordingly, the coded fragment data A 2 of the most important channel to be next transmitted is decided as the selected fragment data (step S 19 ) and then transmitted (step S 1 ).

As seen from the above description, when real-time reproduction of the audio signal of the most important channel is essential and the fragment selecting process is executed as shown in FIG. 10, transmission of the coded fragment data B 1 , B 2 , C 1 , C 4 , D 1 and D 3 is abandoned.

Next, as a second embodiment of the fragment selecting process, the multiplexer 43 is constructed to preferentially choose, as the selected fragment data, the coded fragment data that has a smaller numeral (referred to as a fragment number hereinafter) in a combination of alphabets and numerals (i.e., A 1 , A 2 , B 1 , B 2 , C 1 to C 4 , and D 1 to D 3 ), and that has higher importance among other channels than the most important channel. In this case, the fragment selecting process in step S 5 of FIG. 8 is executed in accordance with, for example, a flowchart shown in FIG. 13 .

More specifically, in the fragment selecting process of this case, a candidate for the selected fragment data is first chosen in step S 21 from among the coded fragment data for each of the channels except the most important channel as with step S 11 of FIG. 10 .

The control flow then goes to step S 22 where, from among the selected fragment data candidates chosen from the respective channels except the most important channel, one having the highest importance and the other having the second highest importance are chosen, followed by going to step S 23 . In step S 23 , of the two candidates (for the selected fragment data) chosen in step S 22 , one having a smaller fragment number is finally chosen as temporary selected fragment data, followed by going to step S 24 . Further, in step S 23 , if the two candidates chosen in step S 22 have the same fragment number, one belonging to the channel having higher importance, for example, is finally chosen as temporary selected fragment data.

In step S 24 , on an assumption that the temporary selected fragment data is transmitted, the estimated end time (T t +T e ) of transmission of the temporary selected fragment data is estimated as with step S 15 of FIG. 10, and whether the estimated end time is before the reproduction start time T s of the temporary selected fragment data is determined.

If it is determined in step S 24 that the estimated end time of transmission of the temporary selected fragment data is before the reproduction start time T s of the same, the control flow goes to step S 25 to determine whether the estimated end time of transmission of the temporary selected fragment data is before the reproduction start time T sm of the fragment of the most important channel to be next transmitted, as with step S 16 of FIG. 10 .

If it is determined in step S 25 that the estimated end time of transmission of the temporary selected fragment data is before the reproduction start time T sm of the fragment of the most important channel to be next transmitted, the control flow goes to step S 26 where the temporary selected fragment data is decided as the selected fragment data, followed by returning to the main control flow.

On the other hand, if it is determined in step S 24 that the estimated end time of transmission of the temporary selected fragment data is not before the reproduction start time T s of the same, the control flow goes to step S 27 to determine whether any of the two candidates, which are chosen in step S 22 , remains not yet finally chosen as the temporary selected fragment data. If it is determined in step S 27 that any of the two candidates chosen in step S 22 remains not yet finally chosen as the temporary selected fragment data, the control flow returns to step S 28 where the selected fragment data candidate not yet finally chosen as the temporary selected fragment data is finally chosen as the temporary selected fragment data, followed by returning to step S 24 to repeat the subsequent processing in a like manner.

If it is determined in step S 27 that there remains no candidate which is chosen in step S 22 and not yet finally chosen as the temporary selected fragment data, the control flow goes to step S 29 to determine whether any of the selected fragment data candidates chosen in step S 21 remains not yet finally chosen as the temporary selected fragment data. If it is determined in step S 29 that any selected fragment data candidate remains not yet finally chosen as the temporary selected fragment data, the control flow returns to step S 22 , following which the above-described processing subsequent to step S 22 is repeated on the candidate(s) not yet chosen as the temporary selected fragment data.

Also, if it is determined in step S 25 that the estimated end time of transmission of the temporary selected fragment data is not before the reproduction start time T sm of the fragment of the most important channel to be next transmitted, the control flow goes to step S 27 , following which the same processing as described above is carried out.

On the other hand, if it is determined in step S 29 that there remains no selected fragment data candidate which is not yet finally chosen as the temporary selected fragment data, the control flow goes to step S 30 where the coded fragment data of the most important channel to be next transmitted is decided as the selected fragment data, followed by returning to the main control flow.

With the fragment selecting process of FIG. 13, the audio signals A to D of four channels shown in FIG. 9 are transmitted as shown in FIGS. 14B to 14 G.

More specifically, as described above in connection with step S 1 of FIG. 8, since the coded fragment data of the most important channel, which has the earliest reproduction start time, T s is given as the selected fragment data at the first, the coded fragment data A 1 is first transmitted as shown in FIG. 14B (step S 1 ).

After transmitting the coded fragment data A 1 , the coded fragment data B 2 , C 2 and D 2 , for each of which the reproduction start time T s is after the transmission end time T e of the coded fragment data A 1 and the period of time from the reproduction start time T s to the transmission end time T e is minimum, are chosen as candidates for the selected fragment data from the respective channels except the most important channel (step S 21 ). At this point in time, therefore, transmission of the coded fragment data B 1 , C 1 and D 1 is abandoned because the transmission of these data would be too late for the start of reproduction thereof.

Of the selected fragment data candidates B 2 , C 2 and D 2 , the candidates B 2 and C 2 have the highest importance and the second highest importance, respectively; hence these two candidates are chosen (step S 22 ). Then, of the candidates B 2 and C 2 , one having a smaller fragment number is finally chosen as the temporary selected fragment data (step S 23 ). Here, as mentioned above, if the two candidates have the same fragment number, one having higher importance is chosen. In this case, therefore, the candidate B 2 is finally chosen as the temporary selected fragment data.

On an assumption that the temporary selected fragment data B 2 is transmitted, as seen from FIG. 14A and 14C, the estimated end time of transmission of the temporary selected fragment data B 2 is before not only the reproduction start time T b2 thereof but also the reproduction start time T a2 of the coded fragment data A 2 of the most important channel to be next transmitted (steps S 24 and S 25 ). Accordingly, the temporary selected fragment data B 2 is decided as the selected fragment data (step S 26 ) and then transmitted (step S 1 ).

After transmitting the coded fragment data B 2 , the coded fragment data C 3 and D 2 , for each of which the reproduction start time T s is after the transmission end time T e of the coded fragment data B 2 and the period of time from the reproduction start time T s to the transmission end time T e is minimum, are chosen as candidates for the selected fragment data from the respective channels except the most important channel (step S 21 ) (FIGS. 14 A and 14 C). At this point in time, therefore, transmission of the coded fragment data C 2 is abandoned because the transmission of that data would be too late for the start of reproduction thereof.

Since only two C 3 and D 2 are the selected fragment data candidates, both are chosen as the candidates having the highest importance and the second highest importance (step S 22 ). Then, of the candidates C 3 and D 2 , one D 2 having a smaller fragment number is finally chosen as the temporary selected fragment data (step S 23 ).

On an assumption that the temporary selected fragment data D 2 is transmitted, as seen from FIG. 14A and 14D, the estimated end time of transmission of the temporary selected fragment data D 2 is not before the reproduction start time T d2 thereof (step S 24 ). Accordingly, of the two candidates C 3 and D 2 chosen in step S 22 , the other C 3 not yet finally chosen as the temporary selected fragment data is finally chosen as the temporary selected fragment data (steps S 27 and S 28 ).

Further, on an assumption that the temporary selected fragment data C 3 is transmitted, as seen from FIG. 14A and 14E, the estimated end time of transmission of the temporary selected fragment data C 3 is before not only the reproduction start time T c3 thereof but also the reproduction start time T a2 of the coded fragment data A 2 of the most important channel to be next transmitted (steps S 24 and S 25 ). Accordingly, the temporary selected fragment data C 3 is decided as the selected fragment data (step S 26 ) and then transmitted (step S 1 ).

After transmitting the coded fragment data C 3 , the coded fragment data C 4 and D 3 , for each of which the reproduction start time T s (FIG. 14A) is after the transmission end time T e (FIG. 14E) of the coded fragment data C 3 and the period of time from the reproduction start time T s to the transmission end time T e is minimum, are chosen as candidates for the selected fragment data from the respective channels except the most important channel (step S 21 ).

Since only two C 4 and D 3 are the selected fragment data candidates, both are chosen as the candidates having the highest importance and the second highest importance (step S 22 ). Then, of the candidates C 4 and D 3 , one D 3 having a smaller fragment number is finally chosen as the temporary selected fragment data (step S 23 ).

On an assumption that the temporary selected fragment data D 3 is transmitted, as seen from FIG. 14A and 14F, the estimated end time of transmission of the temporary selected fragment data D 3 is before not only the reproduction start time t d3 thereof but also the reproduction start time T a2 of the coded fragment data A 2 of the most important channel to be next transmitted (steps S 24 and S 25 ). Accordingly, the temporary selected fragment data D 3 is decided as the selected fragment data (step S 26 ) and then transmitted (step S 1 ).

After transmitting the coded fragment data D 3 , the coded fragment data C 4 , for which the reproduction start time T s (FIG. 14A) is after the transmission end time T e (FIG. 14F) of the coded fragment data C 3 and the period of time from the reproduction start time T s to the transmission end time T e is minimum, is chosen as a candidate for the selected fragment data from the respective channels except the most important channel (step S 21 ).

Since only one C 4 is the selected fragment data candidate, it is chosen as the candidate having the highest importance (step S 22 ), and then chosen as the candidate having a smaller fragment number (step S 23 ). In other words, the candidate C 4 is finally chosen as the temporary selected fragment data.

On an assumption that the temporary selected fragment data C 4 is transmitted, the estimated end time of transmission of the temporary selected fragment data C 4 is before the reproduction start time t c4 thereof (step S 24 ), but not before the reproduction start time T a2 of the coded fragment data A 2 of the most important channel to be next transmitted (step S 25 ). In addition, there remains no candidate which is not yet finally chosen as the temporary selected fragment data (step S 29 ). Accordingly, the coded fragment data A 2 of the most important channel to be next transmitted is decided as the selected fragment data (step S 30 ) and then transmitted (step S 1 ) as shown in FIG. 14 G.

As seen from the above description, when real-time reproduction of the audio signal of the most important channel is essential and the fragment selecting process is executed as shown in FIG. 13, transmission of the coded fragment data B 1 , C 1 , C 2 , C 4 , D 1 and D 2 is abandoned.

Next, as a third embodiment of the fragment selecting process, the multiplexer 43 is constructed to preferentially choose, as the selected fragment data, the coded fragment data that has a smaller fragment number and has higher importance among other channels than the most important channel, and in addition, when transmission of the coded fragment data of the most important channel to be next transmitted would be too late if the coded fragment data chosen based on the above conditions is chosen as the selected fragment data, to preferentially choose, as the selected fragment data, the coded fragment data that does not impede the transmission of the coded fragment data of the most important channel to be next transmitted. In this case, the fragment selecting process in step S 5 of FIG. 8 is executed in accordance with, for example, a flowchart shown in FIG. 15 .

More specifically, in the fragment selecting process of this case, the same processing as in steps S 21 to S 29 of FIG. 13 is executed in steps S 31 to S 39 . Then, if it is determined in step S 39 that there remains no selected fragment data candidate which is not yet finally chosen as the temporary selected fragment data, the control flow goes to step S 40 where one of the coded fragment data not yet transmitted, for which the reproduction start time T s is after the transmission end time T e and the period of time T t required for transmitting that data is minimum (i.e., the data amount thereof is minimum), is chosen as the temporary selected fragment data.

Subsequently, the control flow goes to step S 41 to determine, as with step S 16 of FIG. 10, whether the estimated end time of transmission of the temporary selected fragment data is before the reproduction start time T sm of the fragment of the most important channel to be next transmitted.

If it is determined in step S 41 that the estimated end time of transmission of the temporary selected fragment data is before the reproduction start time T sm of the fragment of the most important channel to be next transmitted, the control flow goes to step S 42 where the temporary selected fragment data is decided as the selected fragment data, followed by returning to the main control flow. In other words, when the transmission of the fragment of the most important channel to be next transmitted is in time for the start of reproduction thereof even if the coded fragment data having the minimum period of time T t required for the transmission thereof is transmitted, that coded fragment data having the minimum period of time T t required for the transmission thereof is chosen as the selected fragment data and then transmitted.

On the other hand, if it is determined in step S 41 that the estimated end time of transmission of the temporary selected fragment data is not before the reproduction start time T sm of the fragment of the most important channel to be next transmitted, namely, when the transmission of the fragment of the most important channel to be next transmitted would be too late if the coded fragment data having the minimum period of time T t required for the transmission thereof is transmitted, the control flow goes to step S 43 where the fragment of the most important channel to be next transmitted is decided as the selected fragment data, followed by returning to the main control flow.

With the fragment selecting process of FIG. 15, audio signals of multiple channels are transmitted, for example, as shown in FIGS. 16A to 16 I.

For convenience of description, it is supposed that a piece of music made up of audio signals A, B, C, D and E of five channels, shown in FIG. 16A, are transmitted.

In an example of FIG. 16A, therefore, the channel of the audio signal A has the highest importance, and the channels of the audio signals B, C, D have importance descending in this order. The channel of the audio signal E has the lowest importance.

Also, the audio signals A, C and D are each divided similarly to the case of FIG. 9 . The audio signal B is divided into three fragments B 1 , B 2 and B 3 . The fragments B 1 and B 2 are the same as those in the case of FIG. 9 . Accordingly, the audio signal B shown in FIG. 16A is constructed by adding the fragment B 3 to the fragments B 1 and B 2 . Note that the fragment B 3 has the reproduction start time T b3 (<T e5 ).

Furthermore, in FIG. 16A, the audio signal E is divided into five fragments E 1 , E 2 , E 3 , E 4 and E 5 . The reproduction start time of the fragment E 1 is coincident with the reproduction start time of the fragment A 1 (B 1 , C 1 , D 1 ), and the fragments E 2 to E 5 have the reproduction start time T e2 to T e5 , respectively. As shown, the relation of T e2 <T e3 <e t4 <T t5 is set in the example of FIG. 16 A. Additionally, the time T e3 is coincident with the time t c2 f T e4 is coincident with the time T e4 , and T e5 is coincident with the time T b3 , respectively.

With the fragment selecting process of FIG. 15, the audio signals A to E of five channels shown in FIG. 16A are transmitted as shown in FIGS. 16B to 16 I.

More specifically, as described above in connection with step S 1 of FIG. 8, since the coded fragment data of the most important channel, which has the earliest reproduction start time, is given as the selected fragment data at the first, the coded fragment data A 1 is first chosen as the selected fragment data and then transmitted as shown in FIG. 16B (step S 1 ).

After transmitting the coded fragment data A 1 , the coded fragment data B 2 , C 2 , D 2 and E 3 , for each of which the reproduction start time T s is after the transmission end time T e of the coded fragment data A 1 and the period of time from the reproduction start time T s to the transmission end time T e is minimum, are chosen as candidates for the selected fragment data from the respective channels except the most important channel (step S 31 ). At this point in time, therefore, transmission of the coded fragment data B 1 , C 1 , D 1 , E 1 and E 2 is abandoned because the transmission of these data would be too late for the start of reproduction thereof.

Of the selected fragment data candidates B 2 , C 2 , D 2 and E 3 , the candidates B 2 and C 2 have the highest importance and the second highest importance, respectively; hence these two candidates are chosen (step S 32 ). Then, of the candidates B 2 and C 2 , one having a smaller fragment number is finally chosen as the temporary selected fragment data (step S 33 ). Here, as mentioned above, if the two candidates have the same fragment number, one having higher importance is chosen. In this case, therefore, the candidate B 2 is finally chosen as the temporary selected fragment data.

On an assumption that the temporary selected fragment data B 2 is transmitted, as seen from FIGS. 16A and 16C, the estimated end time of transmission of the temporary selected fragment data B 2 is before not only the reproduction start time T b2 thereof but also the reproduction start time T a2 of the coded fragment data A 2 of the most important channel to be next transmitted (steps S 34 and S 35 ). Accordingly, the temporary selected fragment data B 2 is decided as the selected fragment data (step S 36 ) and then transmitted (step S 1 ).

After transmitting the coded fragment data B 2 , the coded fragment data B 3 , C 2 , D 2 and E 3 , for each of which the reproduction start time T s (FIG. 16A) is after the transmission end time T e (FIG. 16C) of the coded fragment data B 2 and the period of time from the reproduction start time T s to the transmission end time T e is minimum, are chosen as candidates for the selected fragment data from the respective channels except the most important channel (step S 31 ).

Of the selected fragment data candidates B 3 , C 2 , D 2 and E 3 , the candidates B 3 and C 2 have the highest importance and the second highest importance, respectively; hence these two candidates are chosen (step S 32 ). Then, of the candidates B 3 and C 2 , one C 2 having a smaller fragment number is finally chosen as the temporary selected fragment data (step S 33 ).

On an assumption that the temporary selected fragment data C 2 is transmitted, the estimated end time of transmission of the temporary selected fragment data C 2 is not before the reproduction start time T c2 thereof (step S 34 ). Accordingly, of the two candidates B 3 and C 2 chosen in step S 32 , the other B 3 not yet finally chosen as the temporary selected fragment data is finally chosen as the temporary selected fragment data (steps S 37 and S 38 ).

Further, on an assumption that the temporary selected fragment data B 3 is transmitted, as seen from FIGS. 16A and 16D, the estimated end time of transmission of the temporary selected fragment data B 3 is before not only the reproduction start time T b3 thereof but also the reproduction start time T a2 of the coded fragment data A 2 of the most important channel to be next transmitted (steps S 34 and S 35 ). Accordingly, the temporary selected fragment data B 3 is decided as the selected fragment data (step S 36 ) and then transmitted (step S 1 ).

After transmitting the coded fragment data B 3 , the coded fragment data C 3 , D 2 and E 4 , for each of which the reproduction start time T s (FIG. 16A) is after the transmission end time T e (FIG. 16D) of the coded fragment data B 3 and the period of time from the reproduction start time T s to the transmission end time T e is minimum, are chosen as candidates for the selected fragment data from the respective channels except the most important channel (step S 31 ). At this point in time, therefore, transmission of the coded fragment data C 2 and E 3 is abandoned because the transmission of these data would be too late for the start of reproduction thereof.

Of the selected fragment data candidates C 3 , D 2 and E 4 , the candidates C 3 and D 2 have the highest importance and the second highest importance, respectively; hence these two candidates are chosen (step S 32 ). Then, of the candidates C 3 and D 2 , one D 2 having a smaller fragment number is finally chosen as the temporary selected fragment data (step S 33 ).

On an assumption that the temporary selected fragment data D 2 is transmitted, as seen from FIGS. 16A and 16E, the estimated end time of transmission of the temporary selected fragment data D 2 is before not only the reproduction start time T d2 thereof but also the reproduction start time T a2 of the coded fragment data A 2 of the most important channel to be next transmitted (steps S 34 and S 35 ). Accordingly, the temporary selected fragment data. D 2 is decided as the selected fragment data (step S 36 ) and then transmitted (step S 1 ).

After transmitting the coded fragment data D 2 , the coded fragment data C 3 , D 3 and E 4 , for each of which the reproduction start time T s (FIG. 16A) is after the transmission end time T e (FIG. 16E) of the coded fragment data D 2 and the period of time from the reproduction start time T s to the transmission end time T e is minimum, are chosen as candidates for the selected fragment data from the respective channels except the most important channel (step S 31 ).

Of the selected fragment data candidates C 3 , D 3 and E 4 , the candidates C 3 and D 3 have the highest importance and the second highest importance, respectively; hence these two candidates are chosen (step S 32 ). Then, since the candidates C 3 and D 3 have the same fragment number, one C 3 of the channel having higher importance is finally chosen as the temporary selected fragment data (step S 33 ).

On an assumption that the temporary selected fragment data C 3 is transmitted, as seen from FIGS. 16A and 16F, the estimated end time of transmission of the temporary selected fragment data C 3 is before not only the reproduction start time T c3 thereof but also the reproduction start time T a2 of the coded fragment data A 2 of the most important channel to be next transmitted (steps S 34 and S 35 ). Accordingly, the temporary selected fragment data C 3 is decided as the selected fragment data (step S 36 ) and then transmitted (step S 1 ).

After transmitting the coded fragment data C 3 , the coded fragment data C 4 , D 3 and E 4 , for each of which the reproduction start time T s (FIG. 16A) is after the transmission end time T e (FIG. 16F) of the coded fragment data C 3 and the period of time from the reproduction start time T s to the transmission end time T e is minimum, are chosen as candidates for the selected fragment data from the respective channels except the most important channel (step S 31 ).

Of the selected fragment data candidates C 4 , D 3 and E 4 , the candidates C 4 and D 3 have the highest importance and the second highest importance, respectively; hence these two candidates are chosen (step S 32 ). Then, of the candidates C 4 and D 3 , one D 3 having a smaller fragment number is finally chosen as the temporary selected fragment data (step S 33 ).

On an assumption that the temporary selected fragment data D 3 is transmitted, as seen from FIGS. 16A and 16G, the estimated end time of transmission of the temporary selected fragment data D 3 is before not only the reproduction start time t d3 thereof but also the reproduction start time T a2 of the coded fragment data A 2 of the most important channel to be next transmitted (steps S 34 and S 35 ). Accordingly, the temporary selected fragment data D 3 is decided as the selected fragment data (step S 36 ) and then transmitted (step S 1 ).

After transmitting the coded fragment data D 3 , the coded fragment data C 4 and E 4 , for each of which the reproduction start time T s (FIG. 16A) is after the transmission end time T e (FIG. 16G) of the coded fragment data D 3 and the period of time from the reproduction start time T e to the transmission end time T e is minimum, are chosen as candidates for the selected fragment data from the respective channels except the most important channel (step S 31 ).

The selected fragment data candidates are now only two C 4 and E 4 ; hence these two candidates are chosen as having the highest importance and the second highest importance (step S 32 ). Then, since the candidates C 4 and D 4 have the same fragment number, one C 4 having higher importance is finally chosen as the temporary selected fragment data (step S 33 ).

On an assumption that the temporary selected fragment data C 4 is transmitted, the estimated end time of transmission of the temporary selected fragment data C 4 is before the reproduction start time t c4 thereof (step S 34 ), but not before the reproduction start time T a2 of the coded fragment data A 2 of the most important channel to be next transmitted (step S 35 ). Accordingly, of the two candidates C 4 and E 4 chosen in step S 32 , the other E 4 not yet finally chosen as the temporary selected fragment data is finally chosen as the temporary selected fragment data (steps S 37 and S 38 ).

Further, on an assumption that the temporary selected fragment data E 4 is transmitted, the estimated end time of transmission of the temporary selected fragment data E 4 is before the reproduction start time T e4 thereof (step S 34 ), but not before the reproduction start time T a2 of the coded fragment data A 2 of the most important channel to be next transmitted (step S 35 ). In addition, there remains at this point in time no candidate which is not yet finally chosen as the temporary selected fragment data (step S 39 ). Then, one of the coded fragment data not yet transmitted, for which the reproduction start time T s is after the transmission end time T e (FIG. 16G) of the coded fragment data D 3 and the period of time T t required for the transmission thereof is minimum; namely, the coded fragment data E 5 in this case, is chosen from among the fragments C 4 , E 4 and E 5 shown in FIG. 16A (step S 40 ). At this point in time, therefore, transmission of the coded fragment data C 4 and E 4 is abandoned.

On an assumption that the coded fragment data E 5 is transmitted, as seen from FIGS. 16A and 16H, the estimated end time of transmission of the coded fragment data E 5 is before the reproduction start time T a2 of the coded fragment data A 2 of the most important channel to be next transmitted (step S 41 ). Accordingly, the coded fragment data E 5 is decided as the selected fragment data (step S 42 ) and then transmitted (step S 1 ).

Stated otherwise, in this case, the coded fragment data E 5 , which can be transmitted during the period of time from the transmission end time T e of the coded fragment data D 3 to the reproduction start time T a2 of the coded fragment data A 2 of the most important channel to be next transmitted, is transmitted prior to the transmission of the coded fragment data A 2 .

After transmitting the coded fragment data E 5 , there remains no coded fragment data E 4 of which estimated end time is before the reproduction start time T a2 of the coded fragment data A 2 of the most important channel to be next transmitted (step S 41 ). Accordingly, the coded fragment data A 2 of the most important channel to be next transmitted is decided as the selected fragment data (step S 43 ) and then-transmitted (step S 1 ) as shown in FIG. 16 I.

As seen from the above description, when real-time reproduction of the audio signal of the most important channel is essential and the fragment selecting process is executed as shown in FIG. 15, transmission of the coded fragment data B 1 , C 1 , C 2 , C 4 , D 1 and E 1 to E 4 is abandoned.

Next, as a fourth embodiment of the fragment selecting process, in addition to preferentially choose, as the selected fragment data, ( 1 ) the coded fragment data having the earlier reproduction start time and ( 2 ) the coded fragment data having higher importance if the reproduction start time is the same, the multiplexer 43 is constructed not to choose, as the selected fragment data, the coded fragment data having importance lower than the importance of the coded fragment data that has been chosen as the selected fragment data, after the coded fragment data of the most important channel having the earliest reproduction start time was transmitted as the selected fragment data. In this case, the fragment selecting process in step S 5 of FIG. 8 is executed in accordance with, for example, a flowchart shown in FIG. 18 .

More specifically, in the fragment selecting process of this case, the same processing as in steps S 11 to S 19 of FIG. 10 is executed in steps S 51 to S 56 . Then, if it is determined in step S 56 that the estimated end time of transmission of the coded fragment data is before the reprouction start time T sm of the fragment of the most important channel to be next transmitted, the control flow goes to step S 57 to determine whether the importance of channel of the temporary selected fragment data is not lower than the lowest of importance levels of respective channels of the coded fragment data chosen as the selected fragment data (referred to as the lowest importance) after the coded fragment data of the most important channel having the earliest reproduction start time was transmitted as the selected fragment data.

If it is determined in step S 57 that the importance of channel of the temporary selected fragment data is not lower than the lowest importance, the control flow goes to step S 58 where the temporary selected fragment data is decided as the selected fragment data as with step S 17 of FIG. 10, followed by returning to the main control flow.

On the other hand, if it is determined in step S 57 that the importance of channel of the temporary selected fragment data is lower than the lowest importance, the control flow goes to step S 59 to determine whether any of the selected fragment data candidates chosen in step S 51 remains not yet finally chosen as the temporary selected fragment data. If it is determined in step S 59 that any selected fragment data candidate remains not yet finally chosen as the temporary selected fragment data, the control flow returns to step S 52 , following which the processing subsequent to step S 52 is repeated on the candidate not yet chosen as the temporary selected fragment data.

If it is determined in step S 59 that there remains no selected fragment data candidate which is not yet finally chosen as the,temporary selected fragment data, the control flow goes to step S 60 where the coded fragment data of the most important channel to be next transmitted is decided as the selected fragment data, followed by returning to the main control flow.

Thus, the coded fragment data of the channel having lower importance than the lowest importance is not chosen as the selected fragment data and hence not transmitted.

With the fragment selecting process of FIG. 17, the audio signals A to D of four channels shown in FIG. 9 are transmitted as shown in FIGS. 18A to 18 E.

More specifically, the coded fragment data A 1 of the most important channel, which has the earliest reproduction start time, is first chosen as the selected fragment data and then transmitted as shown in FIG. 18B (step S 1 ).

After transmitting the coded fragment data A 1 , the coded fragment data B 2 , C 2 and D 2 , for each of which the reproduction start time T s is after the transmission end time T e of the coded fragment data A 1 and the period of time from the reproduction start time T s to the transmission end time T e is minimum, are chosen as candidates for the selected fragment data from the respective channels except the most important channel (step S 51 ). At this point in time, therefore, transmission of the coded fragment data B 1 , C 1 and D 1 is abandoned because the transmission of these data would be too late for the start of reproduction thereof.

Of the selected fragment data candidates B 2 , C 2 and D 2 , only C 2 of which reproduction starts at the time TC 2 has the minimum period of time d (=T s −T e ) from the reproduction start time T s thereof to the transmission end time T e of the coded fragment data A 1 as seen from FIG. 18 A. Therefore, the coded fragment data C 2 is finally chosen as the temporary selected fragment data (steps S 52 to S 54 ).

On an assumption that the temporary selected fragment data C 2 is transmitted, as seen from FIG. 18C, the estimated end time of transmission of the temporary selected fragment data C 2 is before not only the reproduction start time T c2 thereof but also the reproduction start time T a2 of the coded fragment data A 2 of the most important channel to be next transmitted (steps S 55 and S 56 ). Accordingly, the temporary selected fragment data C 2 is decided as the selected fragment data (step S 58 ) and then transmitted (step S 1 ).

Immediately after transmitting the coded fragment data A 1 of the most important channel which has the earliest reproduction start time, the lowest of importance levels of the channels A to D is set to the lowest importance. The coded fragment data C 2 of the channel C is therefore chosen as the selected fragment data because the importance thereof is not lower than the lowest importance (S 57 ).

Also, the lowest importance is changed to the importance of the channel C after the coded fragment data C 2 of the channel C has been chosen as the selected fragment data.

After transmitting the coded fragment data C 2 , the coded fragment data B 2 , C 3 and D 2 , for each of which the reproduction start time T s is after the transmission end time T e of the coded fragment data C 2 and the period of time from the reproduction start time T s to the transmission end time T e is minimum, are chosen as candidates for the selected fragment data from the respective channels except the most important channel (step S 51 ).

Of the selected fragment data candidates B 2 , C 3 and D 2 , only the coded fragment data D 2 of which reproduction starts at the time T d2 has the minimum period of time d (=T s −T e ) from the reproduction start time T s thereof to the transmission end time T e of the coded fragment data C 2 as seen from FIG. 18 A. Therefore, the coded fragment data D 2 is finally chosen as the temporary selected fragment data (steps S 52 to S 54 ).

On an assumption that the temporary selected fragment data D 2 is transmitted, the estimated end time of transmission of the temporary selected fragment data D 2 is before not only the reproduction start time T d2 thereof but also the reproduction start time T a2 of the coded fragment data A 2 of the most important channel to be next transmitted (steps S 55 and S 56 ). However, the importance of the coded fragment data D 2 is lower than the lowest importance, i.e., the importance of the channel C, at this point in time. Accordingly, the temporary selected fragment data D 2 is not chosen as the selected fragment data (step S 57 ).

Then, of the selected fragment data candidates B 2 , C 3 and D 2 , the candidate B 2 and C 3 , which are not yet finally chosen as the temporary selected fragment data, are checked and subjected to the same processing as described above (step S 59 ). In this case, of the selected fragment data candidates B 2 and C 3 , the candidate B 2 having the minimum period of time d from the reproduction start time T 3 thereof to the transmission end time T e of the coded fragment data B 2 is chosen as the temporary selected fragment data (steps S 52 to S 54 ).

On an assumption that the temporary selected fragment data B 2 is transmitted, as seen from FIGS. 18D, the estimated end time of transmission of the temporary selected fragment data B 2 is before not only the reproduction start time T b2 thereof but also the reproduction start time T a2 of the coded fragment data A 2 of the most important channel to be next transmitted (steps S 55 and S 56 ). In addition, the importance of the coded fragment data B 2 is not lower than the lowest importance, i.e., the importance of the channel C (step S 57 ). Accordingly, the temporary selected fragment data B 2 is decided as the selected fragment data (step S 58 ) and then transmitted (step S 1 ).

Here, the lowest importance is changed to the importance of the channel B after the coded fragment data B 2 of the channel B has been chosen as the selected fragment data.

After transmitting the coded fragment data B 2 , the coded fragment data C 3 and D 3 , for each of which the reproduction start time T s is after the transmission end time T e of the coded fragment data B 2 and the period of time from the reproduction start time T s to the transmission end time T e is minimum, are chosen as candidates for the selected fragment data from the respective channels except the most important channel (step S 51 ).

However, the importance of each coded fragment data C 3 , D 3 is lower than the lowest importance, i.e., the importance of the channel B, at this point in time. Accordingly, the temporary selected fragment data C 3 and D 3 are not chosen as the selected fragment data (step S 57 ).

Further, since there remains no more coded fragment data which is not yet transmitted and belongs to any of the channels, except the most important channel, having the importance not lower than that of the channel B as the lowest importance, i.e., belongs to the channel B here, the coded fragment data A 2 of the most important channel to be next transmitted is decided as the selected fragment data (step S 60 ) and then transmitted (step S 1 ) as shown in FIG. 18 E.

In the fragment selecting process of FIG. 17, therefore, only A 1 , C 2 , B 2 and A 2 of the coded fragment data shown in FIG. 9 are transmitted and the number of the coded fragment data transmitted is smaller than in the above-described cases of FIGS. 10, 13 and 15 , but the coded fragment data can be transmitted with an allowance increased correspondingly.

In the above-described fragment selecting processes, it is supposed that the network 2 if free from any delay, and at the same time as when the coded fragment data A 1 of the most important channel, which has the earliest reproduction start time, is transmitted, the coded fragment data A 1 is received and decoded in the client terminal 3 , i.e., that the transmission start time of the coded fragment data A 1 is identical to the reproduction start time thereof. Hence, the coded fragment data B 1 , C 1 and D 1 having the same reproduction start time as the coded fragment data A 1 in FIG. 9, are not transmitted for the reason that the transmission of those data would be too late for the start of reproduction thereof. However, when the client terminal 3 is constructed, for example, such that a buffer 61 A is incorporated in the receiving circuit 61 as indicated by dotted lines in FIG. 6 and the coded fragment data (coded data) received from the network 2 is supplied to the decoding circuit 62 after being temporarily stored in the buffer 61 A, the coded fragment data B 1 , C 1 and D 1 having the same reproduction start time as the coded fragment data A 1 can also be transmitted in time for the start of reproduction thereof.

Stated otherwise, in this case, since the coded fragment data is decoded and reproduced in the client terminal 3 after being temporarily stored in the buffer 61 A, the reproduction start time of the coded fragment data A 1 is delayed from the transmission start time thereof by a time for which the received coded fragment data is held stored in the buffer 61 A (referred to as a buffering time hereinafter), as shown in FIG. 19 A. By setting the buffering time to be longer than the period of time required for transmitting the coded fragment data A 1 , therefore, it becomes possible to transmit even the coded fragment data B 1 , C 1 and D 1 having the same reproduction start time as the coded fragment data A 1 in time for the start of reproduction thereof.

More specifically, the fragment selecting process is executed as follows when the buffering time is set as shown in FIG. 19A, for example, and appropriate one of the coded fragment data is chosen as the selected fragment data by the fragment selecting process described above in connection with FIG. 10 . After transmitting the coded fragment data A 1 as shown in FIG. 19B, the coded fragment data B 1 , C 1 and D 1 , for each of which the reproduction start time T e is after the transmission end time T e of the coded fragment data C 2 and the period of time from the reproduction start time T s to the transmission end time T e is minimum, are chosen as candidates for the selected fragment data from the respective channels except the most important channel (step S 11 ).

Of the selected fragment data candidates B 1 , C 1 and D 1 , all these candidates have the same minimum period of time d (=T s −T e ) from the reproduction start time T s thereof to the transmission end time T e of the coded fragment data. A 1 as seen from FIGS. 19A and 19B. Therefore, the coded fragment data B 1 having the highest importance among B 1 , C 1 and D 1 is finally chosen as the temporary selected fragment data (steps S 12 to S 14 ).

On an assumption that the temporary selected fragment data B 1 is transmitted, as seen from FIG. 19C, the estimated end time of transmission of the temporary selected fragment data B 1 is before not only the reproduction start time T d2 thereof but also the reproduction start time T a2 of the coded fragment data A 2 of the most important channel to be next transmitted (steps S 15 and S 16 ). Accordingly, the temporary selected fragment data B 1 is decided as the selected fragment data (step S 17 ) and then transmitted (step S 1 ).

After transmitting the coded fragment data B 1 , the coded fragment data B 2 , C 1 and D 1 , for each of which the reproduction start time T s is after the transmission end time T e of the coded fragment data B 1 and the period of time from the reproduction start time T s to the transmission end time T e is minimum, are chosen as candidates for the selected fragment data from the respective channels except the most important channel (step S 11 ).

Of the selected fragment data candidates B 2 , C 1 and D 1 , two C 1 and D 1 have the same minimum period of time d (=T s −T e ) from the reproduction start time T s thereof to the transmission end time T e of the coded fragment data B 1 as seen from FIGS. 19A and 19C. Therefore, the coded fragment data. C 1 having higher importance among them is finally chosen as the temporary selected fragment data (steps S 12 to S 14 ).

On an assumption that the temporary selected fragment data C 1 is transmitted, as seen from FIG. 19D, the estimated end time of transmission of the temporary selected fragment data C 1 is before not only the reproduction start time T c1 thereof but also the reproduction start time T a2 of the coded fragment data A 2 of the most important channel to be next transmitted (steps S 15 and S 16 ). Accordingly, the temporary selected fragment data C 1 is decided as the selected fragment data (step S 17 ) and then transmitted (step S 1 ).

After transmitting the coded fragment data C 1 , the coded fragment data B 2 , C 2 and D 1 , for each of which the reproduction start time T s is after the transmission end time T e of the coded fragment data C 1 and the period of time from the reproduction start time T s to the transmission end time T e is minimum, are chosen as candidates for the selected fragment data from the respective channels except the most important channel (step S 11 ).

Of the selected fragment data candidates B 2 , C 2 and D 1 , only D 1 has the minimum period of time d (=T s −T e ) from the reproduction start time T s thereof to the transmission end time T e of the coded fragment data C 1 as seen from FIGS. 19A and 19D. Therefore, the candidate D 1 is finally chosen as the temporary selected fragment data (steps S 12 to S 14 ).

On an assumption that the temporary selected fragment data D 1 is transmitted, as seen from FIG. 19D, the estimated end time of transmission of the temporary selected fragment data D 1 is not before the reproduction start time thereof. Accordingly, of the selected fragment data candidates B 2 , C 3 and D 1 chosen in step S 11 , the candidates B 2 and C 2 , which are not yet finally chosen as the temporary selected fragment data, are checked and subjected to the same processing as described above. Hence, at this point in time, transmission of the coded fragment data D 1 is abandoned.

Of the selected fragment data candidates B 2 and C 2 , only C 2 of which reproduction starts at the time T c2 has the minimum period of time d from the reproduction start time T s thereof to the transmission end time T e of the coded fragment data C 1 as seen from FIGS. 19A and 19D. Therefore, the candidate C 2 is finally chosen as the temporary selected fragment data (steps S 12 to S 14 ).

On an assumption that the temporary selected fragment data C 2 is transmitted, as seen from FIG. 19E, the estimated end time of transmission of the temporary selected fragment data C 2 is before not only the reproduction start time T c2 thereof but also the reproduction start time T a2 of the coded fragment data A 2 of the most important channel to be next transmitted (steps S 15 and S 16 ). Accordingly, the temporary selected fragment data C 2 is decided as the selected fragment data (step S 17 ) and then transmitted (step S 1 ).

After transmitting the coded fragment data C 2 , the coded fragment data B 2 , C 3 and D 2 , for each of which the reproduction start time T s is after the transmission end time T e of the coded fragment data C 2 and the period of time from the reproduction start time T s to the transmission end time T e is minimum, are chosen as candidates for the selected fragment data from the respective channels except the most important channel (step S 11 ).

Of the selected fragment data candidates B 2 , C 3 and D 2 , only D 2 of which reproduction starts at the time T d2 has the minimum period of time d from the reproduction start time T s thereof to the transmission end time T e of the coded fragment data C 2 as seen from FIGS. 19A and 19E. Therefore, the coded fragment data D 2 is finally chosen as the temporary selected fragment data (steps S 12 to S 14 ).

On an assumption that the temporary selected fragment data D 2 is transmitted, as seen from FIG. 19F, the estimated end time of transmission of the temporary selected fragment data D 2 is before not only the reproduction start time T d2 thereof but also the reproduction start time T a2 of the coded fragment data A 2 of the most important channel to be next transmitted (steps S 15 and S 16 ). Accordingly, the temporary selected fragment data D 2 is decided as the selected fragment data (step S 17 ) and then transmitted (step S 1 ).

After transmitting the coded fragment data D 2 , the coded fragment data B 2 , C 3 and D 3 , for each of which the reproduction start time T s is after the transmission end time T e of the coded fragment data D 2 and the period of time from the reproduction start time T s to the transmission end time T e is minimum, are chosen as candidates for the selected fragment data from the respective channels except the most important channel (step S 11 ).

Of the selected fragment data candidates B 2 , C 3 and D 3 , only B 2 of which reproduction starts at the time T b2 has the minimum period of time d from the reproduction start time T e thereof to the transmission end time T e of the coded fragment data D 2 as seen from FIGS. 19A and 19F. Therefore, the coded fragment data B 2 is finally chosen as the temporary selected fragment data (steps S 12 to S 14 ).

On an assumption that the temporary selected fragment data B 2 is transmitted, as seen from FIG. 19G, the estimated end time of transmission of the temporary selected fragment data B 2 is before not only the reproduction start time T b2 thereof but also the reproduction start time T a2 of the coded fragment data A 2 of the most important channel to be next transmitted (steps S 15 and S 16 ). Accordingly, the temporary selected fragment data B 2 is decided as the selected fragment data (step S 17 ) and then transmitted (step S 1 ).

After transmitting the coded fragment data B 2 , the coded fragment data C 3 and D 3 , for each of which the reproduction start time T s is after the transmission end time T e of the coded fragment data B 2 and the period of time from the reproduction start time T s to the transmission end time T e is minimum, are chosen as candidates for the selected fragment data from the respective channels except the most important channel a(step S 11 ).

Of the selected fragment data candidates C 3 and D 3 , only C 3 of which reproduction starts at the time T c3 has the minimum period of time d (=T s −T e ) from the reproduction start time T s thereof to the transmission end time T e of the coded fragment data B 2 as seen from FIGS. 19A and 19G. Therefore, the coded fragment data C 3 is finally chosen as the temporary selected fragment data (steps S 12 to S 14 ).

On an assumption that the temporary selected fragment data C 3 is transmitted, as seen from FIG. 19H, the estimated end time of transmission of the temporary selected fragment data C 3 is before not only the reproduction start time T c3 thereof but also the reproduction start time T a2 of the coded fragment data A 2 of the most important channel to be next transmitted (steps S 15 and S 16 ). Accordingly, the temporary selected fragment data C 3 is decided as the selected fragment data (step S 17 ) and then transmitted (step S 1 ).

After transmitting the coded fragment data C 3 , the coded fragment data C 4 and D 3 , for each of which the reproduction start time T s is after the transmission end time T e of the coded fragment data C 3 and the period of time from the reproduction start time T s to the transmission end time T e is minimum, are chosen as candidates for the selected fragment data from the respective channels except the most important channel (step S 11 ).

Of the selected fragment data candidates C 4 and D 3 , only D 3 has the minimum period of time d from the reproduction start time T s thereof to the transmission end time T e of the coded fragment data B 2 as seen from FIGS. 19A and 19H. Therefore, the coded fragment data D 3 is finally chosen as the temporary selected fragment data (steps S 12 to S 14 ).

On an assumption that the temporary selected fragment data D 3 is transmitted, the estimated end time of transmission of the temporary selected fragment data D 3 is not before the reproduction start time t d3 thereof (step S 15 ). Accordingly, of the selected fragment data candidates C 4 and D 3 , the candidate C 4 , which is not yet finally chosen as the temporary selected fragment data, is checked and subjected to the same processing as described above (step S 18 ).

Here, the coded fragment data C 4 is finally chosen as the temporary selected fragment data. On an assumption that the temporary selected fragment data C 4 is transmitted, the estimated end time of transmission of the temporary selected fragment data C 4 is before the reproduction start time t c4 thereof (step S 15 ), but not before the reproduction start time T a2 of the coded fragment data A 2 of the most important channel to be next transmitted (step S 16 ). Further, since there remains no more candidate for the coded fragment data which is not yet finally chosen as the temporary selected fragment data (step S 18 ), the coded fragment data A 2 of the most important channel to be next transmitted is decided as the selected fragment data (step S 19 ) and then transmitted (step S 1 ) as shown in FIG. 19 I.

Thus, setting the buffering time makes it possible to increase the number of fragments which can be reproduced in real time.

The buffering time can be determined depending on, e.g., how long time can be allowed from the time when transmission of the coded fragment data for a piece of music requested from the client terminal 3 to the time when reproduction of the transmitted data is started in the client terminal 3 . It is however desired that the actual buffering time is longer than the buffering time estimated by the server 1 .

While transmission of the coded fragment data D 1 is abandoned in the embodiment of FIG. 19 as described above, the buffering time may be prolonged to such an extent that the coded fragment data D 1 can also be transmitted and reproduced in real time.

Next, another embodiment of the transmitting process for audio signals executed by the server will be described with reference to a flowchart of FIG. 20 .

As with the transmitting process described above, when a piece of music is requested to the server 1 from the client terminal 3 , audio signals of respective channels making up the piece of music are each divided into fragments each having a predetermined length, and are encoded into coded fragment data. The coded fragment data is then added with the relevant reproduction start time and other necessary control information, followed by being supplied to the multiplexer 43 .

Upon receiving the coded fragment data (added with headers) from the controller 42 for the first time after a request for the piece of music has been issued from the client terminal 3 , the multiplexer 43 chooses, as selected fragment data, one of the coded fragment data that belongs to the channel set to the highest importance in the importance setting unit 34 and has the earliest reproduction start time, followed by supplying the selected fragment data to the transmitting circuit 33 . Correspondingly, in step S 61 of FIG. 20, the transmitting circuit 33 transmits the selected fragment data from the multiplexer 43 , i.e., the coded fragment data that, in this case, belongs to the channel set to the highest importance in the importance setting unit 34 and has the earliest reproduction start time, to the client terminal 3 via the network 2 .

The control flow then goes to step S 62 to determine whether transmission of the selected fragment data is fully completed. If the transmission of the selected fragment data is completed, the control flow goes to step S 63 , following which similar processes as those in steps S 3 to S 5 of FIG. 8 are executed in steps S 63 to S 65 , respectively.

On the contrary, if it is determined in step S 62 that the transmission of the selected fragment data is not completed, namely, if the selected fragment data is under transmission, the control flow returns to step S 66 to determine whether the current time (counted from, e.g., the head of music being as a reference) T n is before the reproduction start time T sm of the coded fragment data of the most important channel to be next transmitted. If it is determined in step S 66 that the current time T n is before the reproduction start time T sm of the coded fragment data of the most important channel to be next transmitted, namely, when transmission of the coded fragment data of the most important channel to be next transmitted is in time for the start of reproduction thereof even if the coded fragment data under transmission continues to be transmitted, the control flow returns to step S 61 to continue the transmission of the coded fragment data under transmission.

On the other hand, if it is determined in step S 66 that the current time T n is not before the reproduction start time T sm of the coded fragment data of the most important channel to be next transmitted, namely, when transmission of the coded fragment data of the most important channel to be next transmitted would be too late for the start of reproduction thereof if the coded fragment data under transmission continues to be transmitted, the control flow goes to step S 67 to cease the transmission of the coded fragment data under transmission. The control flow then goes to step S 63 to detect the time when the transmission of the selected fragment data is ended (ceased) in step S 67 , i.e., the transmission end time T e . Thereafter, the control flow goes to step S 64 to execute similar processing as described above.

Thus, with the transmitting process of FIG. 20, when the transmission of the coded fragment data of the most important channel to be next transmitted would be too late for the start of reproduction thereof, the transmission of the coded fragment data under transmission is ceased and the fragment selecting process is executed in step S 65 , that is to say, the selected fragment data is newly chosen. In the fragment selecting process of step S 65 , therefore, it is no longer required to derive the estimated end time of transmission of the temporary selected fragment data on an assumption that it is transmitted, and to determine whether the estimated end time of transmission of the temporary selected fragment data is before the reproduction start time thereof, and whether it is before the reproduction start time T sm of the coded fragment data of the most important channel to be next transmitted (e.g., steps S 15 and S 16 of FIG. 10 ).

More specifically, for example, the multiplexer 43 is constructed to preferentially choose, as the selected fragment data, (1) the coded fragment data having the earlier reproduction start time, and (2) the coded fragment data having higher importance if the reproduction start time is the same. In this case, the fragment selecting process in step S 65 of FIG. 10 is executed in accordance with, for example, a flowchart shown in FIG. 21 .

In the fragment selecting process of this case, a candidate for the selected fragment data is first chosen in step S 71 from among the coded fragment data not yet transmitted for each of all the channels including the most important channel. Specifically, by way of example, the coded fragment data, for which the reproduction start time T s is after the transmission end time T e and the period of time from the reproduction start time T s to the transmission end time T e is minimum, is chosen as the selected fragment data candidate from the respective channels including the most important channel.

The control flow then goes to step S 72 where, from among the selected fragment data candidates chosen from the respective channels including the most important channel, one for which the period of time d (=T s −T e ) from the reproduction start time T s to the transmission end time T e is minimum is chosen as temporary selected fragment data.

Subsequently, the control flow goes to step S 73 to determine whether two or more of the selected fragment data candidates have the same period of time d. If it is determined in step S 13 that two or more of the selected fragment data candidates have the same period of time d, namely, if the coded fragment data are chosen as the temporary selected fragment data from two or more channels in step S 72 , the control flow goes to step S 74 where one of the coded fragment data chosen from the two or more channels, which belongs to the channel having the highest importance, is finally chosen as the temporary selected fragment data, followed by going to step S 75 .

If it is determined in step S 73 that no selected fragment data candidates have the same period of time d, namely, if the coded fragment data is chosen as the temporary selected fragment data from one channel in step S 72 , that coded fragment data is finally chosen as the temporary selected fragment data, following which the control flow goes to step S 75 while skipping step S 74 . In step S 75 , the coded fragment data chosen as the temporary selected fragment data at that time is decided as the temporary selected fragment data, followed by returning to the main control flow.

With the transmitting process of FIG. 20, the audio signals A to D of four channels shown in FIG. 9 are transmitted as shown in FIGS. 22A to 22 F. It is here assumed that the buffering time is not taken into consideration and the fragment selecting process of step S 65 is executed in accordance with the flowchart shown in FIG. 21 .

More specifically, the coded fragment data A 1 of the most important channel, which has the earliest reproduction start time, is first chosen as the selected fragment data and then transmitted as shown in FIG. 22B (step S 61 ).

After transmitting the coded fragment data A 1 , the coded fragment data A 2 , B 2 , C 2 and D 2 , for each of which the reproduction start time T s is after the transmission end time T e of the coded fragment data A 1 and the period of time from the. reproduction start time T s to the transmission end time T e is minimum, are chosen as candidates for the selected fragment data from the respective channels including the most important channel (step S 71 ).

Of the selected fragment data candidates A 2 , B 2 , C 2 and D 2 , only C 2 of which reproduction starts at the time t c2 has the minimum period of time d (=T e −T e ) from the reproduction start time T s thereof to the transmission end time T e of the coded fragment data A 1 as seen from FIGS. 22A and 22B. Therefore, the coded fragment data C 2 is finally chosen as the temporary selected fragment data (steps S 72 to S 74 ).

Subsequently, the temporary selected fragment data C 2 is decided as the selected fragment data (step S 75 ) and then transmitted as shown in FIG. 22 C.

Since the current time T n does not reach the reproduction start time T a2 of the coded fragment data A 2 of the most important channel to be next transmitted, as seen from FIGS. 22A and 22C, during the transmission of the selected fragment data C 2 , the selected fragment data C 2 is normally transmitted to the last without being ceased during the transmission thereof.

After transmitting the coded fragment data C 2 , the coded fragment data A 2 , B 2 , C 3 and D 3 , for each of which the reproduction start time T s is after the transmission end time T e of the coded fragment data C 2 and the period of time from the reproduction start time T s to the transmission end time T e is minimum, are chosen as candidates for the selected fragment data from the respective channels including the most important channel (step S 71 ).

Of the selected fragment data candidates A 2 , B 2 , C 3 and D 3 , only B 2 of which reproduction starts at the time t c2 has the minimum period of time d (=T s −T e ) from the reproduction start time T s thereof to the transmission end time T e of the coded fragment data C 2 as seen from FIGS. 22A and 22C. Therefore, the coded fragment data B 2 is finally chosen as the temporary selected fragment data (steps S 72 to S 74 ).

Subsequently, the temporary selected fragment data B 2 is decided as the selected fragment data (step S 75 ) and then transmitted as shown in FIG. 22 D.

Since the current time T n reaches the reproduction start time T a2 of the coded fragment data A 2 of the most important channel to be next transmitted, as seen from FIGS. 22A and 22D, during the transmission of the selected fragment data B 2 , the transmission of the selected fragment data B 2 is ceased and ended at the reproduction start time T a2 as shown in FIG. 22D (steps S 66 and S 67 ). As a result, the selected fragment data B 2 is transmitted just midway as shown in FIG. 22 E. In this case, the time at which the transmission of the selected fragment data B 2 has been ceased is detected as the transmission end time T e thereof (step S 63 ).

After ending (ceasing) the transmission of the coded fragment data B 2 , the coded fragment data A 2 , C 4 and D 3 , for each of which the reproduction start time T s is after the transmission end time T e of the coded fragment data B 2 and the period of time from the reproduction start time T s to the transmission end time T e is minimum, are chosen as candidates for the selected fragment data from the respective channels including the most important channel (step S 71 ).

Of the selected fragment data candidates A 2 , C 4 and D 3 , only A 2 of which reproduction starts at the time T a2 has the minimum period of time d (=T s −T e ) from the reproduction start time T s thereof to the transmission end time T e of the coded fragment data B 2 as seen from FIGS. 22A and 22C. Therefore, the coded fragment data A 2 is finally chosen as the temporary selected fragment data (steps S 72 to S 74 ).

Subsequently, the temporary selected fragment data A 2 is decided as the selected fragment data (step S 75 ) and then transmitted as shown in FIG. 22 F.

Thus, by ceasing the transmission of the selected fragment data when the transmission of the coded fragment data of the most important channel to be next transmitted would be too late for the start of reproduction thereof, it is no longer required to derive the estimated end time at which the transmission of the temporary selected fragment data under transmission is to be completed. The transmission of the coded fragment data of the most important channel to be next transmitted can be avoided from becoming too late for the start of reproduction thereof due to an error possibly caused by the need of estimation. In other words, the transmitting process is surely adaptable for the transmission rate of the network 2 that varies with time.

Next, a reproducing process of (coded) audio signals executed by the client terminal 3 will be described with reference to a flowchart of FIG. 23 .

When a piece of music is requested to the server 1 from the client terminal 3 and coded data is transmitted from the server 1 as described above, the coded data is received by the receiving circuit 61 of the client terminal 3 (FIG. 6 ). The received coded data is subjected to format conversion in the receiving circuit 61 and then supplied to the separator 71 of the decoding circuit 62 .

In the separator 71 , as shown in step S 81 , the coded fragment data and the headers are extracted from the coded data, the extracted coded fragment data are supplied to corresponding channels of the decoders 72 1 to 72 M , respectively, and the reproduction start time put in each header is supplied to the reproduction timing adjuster 73 .

The coded fragment data from the separator 71 is decoded to the original fragment in the decoder 72 m and then supplied to the reproduction timing adjuster 73 . In step S 82 , the reproduction timing adjuster 73 determines whether the current time T n is coincident with the reproduction start time supplied from the separator 71 . If it is determined in step S 82 that the current time T n is not coincident with the reproduction start time, the control flow returns to step S 82 . On the other hand, if it is determined that the current time T n is coincident with the reproduction start time supplied from the separator 71 , the reproduction timing adjuster 73 outputs the audio signal fragment, which has been added with the relevant reproduction start time and resulted from decoding the coded fragment data, to the audio signal output circuit 63 . The audio signal output circuit 63 applies a predetermined process to the audio signal fragment from the reproduction timing adjuster 73 , followed by outputting it. The audio signal fragment is then reproduced in step S 83 . Thereafter, the control flow returns to step S 81 to repeat similar processing as described above.

In the case of the receiving circuit 61 having the buffer 61 A, the coded data received by the receiving circuit 61 is supplied to the separator 71 after being temporarily stored in the buffer 61 A.

As described above, digital signals of multiple channels are each divided into one or more fragments, and the fragments of multiple channels are each encoded into coded fragment data. At the transmission end time of the coded fragment data previously transmitted, one of the coded fragment data of multiple channels is selected, and the selected coded fragment data is transmitted. It is therefore possible to realize scalability in units of channel (part in the illustrated embodiments) in such a manner, for example, that the coded fragment data of more channels are transmitted when the transmission rate is sufficiently large, and the coded fragment data of less channels are transmitted when the transmission rate is not sufficiently large. As a result, when the transmission rate is sufficiently large, a piece of music can be provided with high sound quality. Also, even when the transmission rate is insufficient, a piece of music can be provided in real time although sound quality deteriorates somewhat.

Further, in the illustrated embodiments, since the server 1 transmits audio signals while assigning a channel for each part, the user at the client terminal 3 can adjust a volume level of the sound of certain one musical instrument or the singing voice alone, listen to only a vocal part or an accompaniment part, or listen to the sounds of several desired musical instruments in a mixed state. This makes it possible to provide service with high user interactivity.

While the illustrated embodiments execute processing on audio signals, the present invention is also applicable to the case of providing other digital signals such as video signals or text data, for example. In addition to the case of providing only one kind of digital signals such as audio signals or video signals, the present invention is further applicable to the case of providing different kinds of signals in a combination of video signals and audio signals in synch therewith. In this case, for example, real-time reproduction of at least video signals-can be ensured by setting the importance of channels of the video signals to be higher than the importance of channels of the audio signals.

Moreover, the network 2 may be a wire or wireless system so long as it is a bidirectional network.

The server 1 shown in FIG. 4 and the client terminal 3 shown in FIG. 6 can be realized by rendering a CPU (computer) to execute a computer program (software), or by dedicated hardware.

When the processing executed by the server 1 and the client terminal 3 is realized with a computer program, the computer program may be provided in the form of a recording medium such as a CD-ROM (Compact Disc—Read Only Memory) on which it is recorded, or may be provided by transmission via Internet or any other suitable transmitting medium.

While the coded data outputted from the server 1 is provided in the illustrated embodiments via the network 2 as a transmitting medium, the coded data may be provided to the client terminal 3 in the form of a recording medium such as a CD-ROM on which the coded data is recorded.

Furthermore, the manner of selecting a particular fragment in the fragment selecting process is not limited to those described above.

According to the transmitting apparatus and method and the providing medium as described in the above embodiments, digital signals of multiple channels are each divided into one or more fragments, and the fragments of multiple channels are each encoded into coded fragment data. At the predetermined time, one of the coded fragment data of multiple channels is selected, and the selected coded fragment data is transmitted. The digital signals can be therefore transmitted in such a manner as to be able to decode and reproduce them in real time, for example.

Also, according to the providing medium as described in the above embodiments, coded fragment data is provided which is obtained by dividing each of digital signals of multiple channels into one or more fragments, encoding each of the fragments of multiple channels into coded fragment data, and at the predetermined time, selecting and then outputting one of the coded fragment data of multiple channels. The digital signals can be therefore decoded and reproduced in real time, for example.

Further, according to the receiving apparatus and method and the providing medium as described in the above embodiments, coded data is separated into coded fragment data that is obtained by dividing each of digital signals of multiple channels into one or more fragments and encoding each of the fragments of multiple channels, and time information related to the reproduction start time at which reproduction of the coded fragment data is to be started. Then, the coded fragment data is decoded into the original fragment and reproduction of the original fragment is controlled in accordance with the time information. The digital signals can be therefore decoded and reproduced in real time, for example.

Moreover, according to the receiving apparatus and method and the-providing medium as described in the above embodiments, coded fragment data is received which is obtained by dividing each of digital signals of multiple channels into one or more fragments, encoding each of the fragments of multiple channels into coded fragment data, and at the predetermined time, selecting and then outputting one of the coded fragment data of multiple channels. The coded fragment data is then decoded into the original fragment. The digital signals can be therefore decoded and reproduced in real time, for example.