JP3286110B2 - Voice packet interpolation device - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

The present invention relates to, for example, televisions, VTs,
The present invention relates to an audio packet interpolation device used in digital equipment having a serial digital interface, such as R or a camera.

[0002]

2. Description of the Related Art In recent years, in transmission and recording of a video signal or an audio signal, digital equipment having little image quality or sound quality deterioration has begun to spread in place of conventional analog equipment. In particular, in studio equipment, video and audio signals are transmitted over a single transmission line using SM.
PTE 259M standard has been established. This is a standard for serial digital interfaces.

Therefore, a conventional technique for transmitting a voice digital signal as a serial digital signal will be described. S17.100 of SMPTE describes a standard for multiplexing audio digital data or additional data in an auxiliary data area of a serial digital video signal conforming to the SMPTE 259M standard. That is, this standard intends to transmit an audio digital signal during a blanking period of a video signal which is a main signal. For audio formats,
The rules of AES3-1991 (ANSI S4.40-1991) apply mutatis mutandis. This standard for transmitting audio digital signals as serial digital signals uses three packets: control packets, audio packets, and additional packets.

FIG. 3 shows the data format of the three types of packets. FIG. 3A is a diagram illustrating a data format of a control packet. That is, the control packet is 1
It consists of a 0-bit word string, which is fixed at 17 words in the composite system and 19 words in the component system. ANC data flag (AD
F) is one word (3FCh) in the composite system,
3 words in the component system (000h, 3FFh, 3F
Fh). The data ID (DID) indicates to which audio group among the audio groups 1, 2, 3, and 4 the control packet belongs. Data block number (DBN)
Is always fixed at 200h. Data count (DC)
Is always fixed at 20 ch (12 words). AFT1-2
Represents a voice frame number for ch1 and ch2.
AFT3-4 represents an audio frame number for ch3 and ch4. RATE specifies the sampling frequency of each audio channel pair and the like. ACT is the active channel
Is shown. DELA0-DELA2 or DELB0-DE
LB2 expresses a relative delay amount of audio with respect to the video signal by a multiple of an audio sampling interval. C
S is an error detection code of the control packet.

[0005] The control packet includes 1 at the beginning of each field.
Transmitted every time. The control packet indicates the sampling frequency, synchronization / asynchronism with the video signal, delay with respect to the video signal, presence / absence of an audio signal, and an audio frame number. If this control packet is not received, it is determined that the audio data is synchronized with the video signal at a sampling frequency of 48 kHz. The sampling frequency depends on the transmission speed of the voice packet.
It is set to either 2, 44.1 or 48 kHz.

FIG. 3B shows a data format of a voice packet. That is, the voice packet is composed of a 10-bit word string, and the number of words is variable. AN
The C data flag (ADF) is one word (3FCh) in the composite system and three words (000h, 3FFh, 3FFh) in the component system. The data ID (DID) indicates to which of the audio groups 1, 2, 3, or 4 the audio packet belongs. The data block number (DBN) is a serial number assigned to audio packets belonging to the same audio group when the MSB 20 bits of the audio digital data of the audio group are subdivided into audio packets. A numerical value that changes periodically in the range is assigned. The data count (DC) indicates the number of words of the user data. The user data is a sub-frame AD1 or AD2 composed of three words, and the maximum number of words of the user data is 255. Data relating to audio digital data of 20 bits on the MSB side is allocated to the subframe AD1 or AD2. The CS is assigned an error detection code of a voice packet.

FIG. 4A shows the contents of the 30-bit bit address of the subframe AD1 or AD2. That is, the block sync (Z) indicates whether the subframe is the beginning of a new channel status block. If it is the beginning of the channel status block, Z = 1, otherwise Z = 0. ch1 and ch2 are audio channels
This is for identifying 1-4. aud0-19
Is 20-bit audio digital data linearly represented by two's complement. The validity bit (V) indicates the validity of the audio sample. If the audio digital data of the subframe is suitable for conversion to an analog audio signal, V = 1, and if not, V = 0. . The user bit (U) is a bit for transmitting user data specified by the user. The channel status bit (C) is for transmitting information related to the audio channel, and one block is constituted by 192 channel status bits. The block sync of the subframe serving as the starting point of this block is Z = 1 as described above. The parity bit (P) is an even parity for 26 bits of the subframe excluding b9 in the first, second, and third columns.

FIG. 3C shows a data format of the additional packet. That is, the additional packet is 10
It consists of a word string of bits, and the number of words is variable. A
The NC data flag (ADF) is one word (3FCh) in the composite system and 3 words in the component system.
Word (000h, 3FFh, 3FFh). Data ID (DI
D) indicates that the additional packet is a voice group 1, 2, 3 or 4.
Indicates to which of the audio groups the audio group belongs. The data block number (DBN) is a serial number assigned to an additional packet belonging to the same audio group when the LSB 4 bits of the audio digital data of the audio group are subdivided into additional packets. A numerical value that changes periodically in the range is assigned. Data count (D
C) represents the number of words in the AUX column. An AUX consisting of 10 bits per word is assigned data related to 4-bit audio digital data on the LSB side. CS is
This is an error detection code of the additional packet.

FIG. 4B shows the contents of a 10-bit AUX bit address. That is, four bits on the LSB side of the subframe AD1 are assigned to x0 to X3. The LSB enters X0. The four bits on the LSB side of the subframe AD2 are assigned to y0 to y3. The LSB enters y0. P is b0-b
7, even parity.

[0010] Devices fully compliant with this standard described above operate in two modes. That is, a 24-bit mode for transmitting 24-bit audio and a 20-bit mode for transmitting 20-bit audio. In the 24-bit mode, the data of 20 bits on the MSB side is transmitted by a voice packet, and the remaining 4 bits are transmitted by an additional packet. Therefore, in order to simplify the circuit for reconstructing the AES format, the voice packet and the additional packet must be transmitted within the same blanking period. Further, the voice packet and the additional packet need to be transmitted adjacently. In this mode, 4 bits of 24-bit audio
Up to 12 channels can be transmitted. On the other hand, in the 20-bit mode, since all 20 bits can be transmitted by the voice packet, no additional packet is used. In this mode, sound with 20-bit accuracy can be transmitted from 4 ch to 16 ch.

[0011]

However, according to such a conventional method, an error occurs on a transmission line.
When a voice packet is lost, the length of the variable-length packet cannot be estimated on the receiving side, so that the missing voice packet cannot be interpolated. For this reason, there has been a problem that sound signal information or the like is lost and large noise is generated.

SUMMARY OF THE INVENTION It is therefore an object of the present invention to provide a voice packet interpolating apparatus capable of preventing the occurrence of noise in a digital device having a serial digital interface in consideration of such problems.

[0013]

Means for Solving the Problems The first invention (claim 1)
Storage means for temporarily storing and holding voice packet data having voice data and data relating to other information, wherein the number of words of a predetermined bit is always fixed;
A data block number detector that detects a data block number that is a serial number assigned to the audio packet from the data of the audio packet read from the storage unit based on a predetermined order; Based on a lost voice packet detecting means for detecting a lost voice packet and a first-in first-out order.
The voice packet data read from the storage means.
From the video signal data and audio
Frequency and phase synchronization with the
The data position of the audio packet based on the video signal.
Clock that detects the clock phase information that holds the position
Phase information detecting means, and data included in the voice packet.
To detect the sampling frequency of the
Reading from the sampling means and the storage means.
Clock included in the data of the voice packet
Using the phase information and the sampling frequency, the reading is performed.
Clock phase of data for extracted voice packet
Estimated value generating means for generating an estimated value of information and the missing voice
If a packet is missing based on the detection result of the packet detection
The voice packet that can be read from the storage means.
Clock data and clock position generated by the estimated value generator
Using the estimated value of the phase information,
Corresponding interpolated data containing clock phase information data
Interpolation data generating means for generating the missing voice packet
Is missing based on the detection result of the
Is based on the order corresponding to the missing voice packets.
When the interpolation data is output and there is no missing data
Is the interpolation data interpolation for outputting the data of the voice packet.
An audio packet interpolating device is characterized in that a guide means.

[0014] The second of the present invention (corresponding to claim 2), the sound
Voice data and other information-related data,
Decoding of voice packets where the number of bits words is always fixed
Storage means for temporarily storing and holding over data, erroneous speech packet
Sound that detects an erroneous voice packet by the check code
Voice packet error detection means and based on first-in first-out order.
Data of the voice packet read from the storage means.
Data and audio of video signals that are independent of each other
Synchronize frequency and phase with packet data
Data of audio packets based on video signals
The clock that detects the clock phase information that holds the position
Phase information detecting means, and data included in the voice packet.
The sampling frequency for the audio signal of the data
Sampling frequency detection means, and reading from the storage means.
The clock included in the data of the audio packet
Using the phase information and the sampling frequency,
Clock position of data for read voice packet
An estimated value generating means for generating an estimate of the phase information, the voice
An error occurs based on the detection result of the packet error detection means
If it has been, it can be read from the storage means
Generated by voice packet data and estimated value generator
Using the estimated value of the clock phase information,
Of clock phase information corresponding to the voice packet
Generating interpolation data including: interpolation data generation means;
Based on the detection result of the voice packet error detection means,
If the audio packet has been lost,
Outputting the interpolation data based on the corresponding order,
If the voice packet has not been generated,
Data interpolation means for outputting data
Is a voice packet interpolation device .

According to a third aspect of the present invention (corresponding to claim 3), a sound
Voice data and other information-related data,
Decoding of voice packets where the number of bits words is always fixed
Storage means for temporarily storing data, and a first-in first-out order.
The voice packet read from the storage means based on the preamble.
Data of video signals that are independent of each other
Frequency and phase between the data and the voice packet data.
Audio packets based on the video signal
Detects clock phase information that holds the data position
Clock phase information detecting means included in the voice packet.
The sampling frequency for the audio signal of the data
Sampling frequency detecting means for detecting, and the storage means
Included in the voice packet data that can be read from
Using the clock phase information and the sampling frequency
Data packet for the read voice packet.
Estimated value generating means for generating an estimated value of the lock phase information;
Detected by the estimated value and the clock phase information detecting means
Missing or incorrect based on the clock phase information
Missing voice packet detection that detects generated voice packets
Means based on the detection result of the missing voice packet detecting means.
If any errors or omissions have occurred,
Using the voice packet data that can be read from the
The voice packet in which the drop or error has occurred.
An interpolation data generating means for generating a response to interpolate data, before
Based on the detection result of the missing voice packet detector,
Or, if an error has occurred,
Outputs the interpolation data based on the order corresponding to the bracket
If no omissions or errors have occurred,
Interpolating data insertion means for outputting packet data
A voice packet interpolating apparatus characterized by the above.

According to a fourth aspect of the present invention (corresponding to claim 4), a sound
Voice data and other information-related data,
Decoding of voice packets where the number of bits words is always fixed
Storage means for temporarily storing and holding over data, based on a predetermined order
Data of the voice packet read from the storage means.
Data, which is a serial number assigned to the voice packet.
Data block number to detect tab block number
Based on the detection means and the data block number,
First missing voice packet detection for detecting dropped voice packets
Outgoing means and the first in first out order.
From the voice packet data read from the stage,
Independent video signal data and audio packet data
Video signal to synchronize the frequency and phase between
Holds the position of the audio packet data with respect to
Clock phase information detector that detects clock phase information
And a voice signal of data contained in the voice packet.
Sampling frequency to detect sampling frequency
Number detection means and a sound readable from the storage means
Clock phase information included in voice packet data and the
Using the sampling frequency and the read audio
An estimate of the clock phase information of the data for the packet
Means for generating an estimated value, the estimated value and the clock
Clock phase information detected by the phase information detecting means;
Voice packets that are missing or incorrect based on the
A second missing voice packet detecting means for outputting the first missing voice packet;
Voice packet detection means and the second missing voice packet detection
Missing or incorrect data is generated based on the detection result of
If there is a voice that can be read from the storage means
Using the data of the packet, the missing voice packet
Data generation means for generating interpolation data corresponding to
When the first lost audio packet detecting means and the second missing
Based on the detection result of the dropped voice packet detection means,
If an error has occurred, the missing voice packet
Output the interpolation data based on the order corresponding to the
If no dropouts or errors have occurred , the audio packet
And an interpolation data insertion means for outputting the data of the
An audio packet interpolating apparatus characterized in that:

A fifth aspect of the present invention (corresponding to claim 5) is that
Data and other information-related data.
Voice packet data where the number of words in the packet is always fixed
Storage means for temporarily storing data, and errors in voice packets
Speech to detect erroneous speech packets by check code
Complementary data generation means comprising packet error detection means
However, the detection result of the voice packet
Missing or incorrect based on the detection result of the
If an error has occurred, the
Using the data of the voice packet
Generates interpolation data corresponding to the voice packet in which
Voice packet interpolation according to the third aspect of the present invention.
Device.

According to a sixth aspect of the present invention ( corresponding to claim 6),
Data and other information-related data.
Voice packet data where the number of words in the packet is always fixed
Storage means for temporarily storing data, and errors in voice packets
Speech to detect erroneous speech packets by check code
Complementary data generation means comprising packet error detection means
The first missing voice packet detecting means and the second missing voice packet detecting means.
Detection result of voice packet detection means and voice packet error detection
Missing or incorrect data is generated based on the detection result of
If there is a voice packet that can be read from the storage
The missing or incorrect data was generated using the data of the
Generate interpolation data corresponding to audio packets.
5. The voice packet interpolator according to claim 4, wherein:
You.

According to a seventh aspect of the present invention (corresponding to claim 7),
The interpolation data generating means converts the voice packet data into
Using the included voice data, the missing voice packet is used.
Interpolation data corresponding to the audio data included in the data
Generating a first, second, third, fourth, fifth, or
Is a voice packet interpolation device according to the present invention.

An eighth aspect of the present invention (corresponding to claim 8) is that
The interpolation data generating means converts the voice packet data into
Using the included clock phase information data, the missing
Phase information included in the data of the voice packet
Generating interpolation data corresponding to the data of
The third, fourth, fifth, or sixth voice packet complement of the present invention
Device.

According to a ninth aspect of the present invention (corresponding to claim 9),
The sampling frequency detecting means detects the voice of the voice packet.
Control including data of sampling frequency for signal
The packet signal is used to detect the sampling frequency of the audio signal.
First, second, third, fourth, fifth, or
6 is a voice packet interpolation device of the present invention.

A tenth aspect of the present invention ( corresponding to claim 10) is:
The sampling frequency detecting means is the storage means;
Uses clock phase information of voice packets that can be read from
To detect the sampling frequency of the audio signal.
The third or fourth aspect of the present invention provides a voice packet compensator according to the present invention.
Device.

[0023]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS Hereinafter, the present invention will be described in detail with reference to the drawings showing the embodiments.

FIG. 1 is a block diagram of one embodiment of the voice packet interpolation apparatus of the present invention. Before describing the present embodiment, the specification of a serial digital signal transmission system supported by the apparatus of the present embodiment will be described. The transmission method is
It is for transmitting a video digital signal and an audio digital signal as a serial digital signal. The video digital signal is processed as a main signal, and the audio digital signal is processed as an auxiliary signal independent of the video signal. For audio format, AES3-1
991 (ANSI S4.40-1991) applies mutatis mutandis. The device according to the present embodiment relates to the audio digital signal which is the auxiliary signal. This transmission method uses two packets, a control packet and a voice packet.

FIG. 5 is a diagram showing the data format of a control packet and a voice packet of the present transmission system. FIG.
(A) shows the data format of the control packet.
For simplicity, the control packet has the same specifications as the control packet in the conventional component system shown in FIG. Therefore, the control packet is composed of a 10-bit word string, and the number of words is always fixed at 19 words. Although the control packet of the present embodiment has the same specifications as the control packet of the conventional component system shown in FIG. 3A, the present invention is not limited to this. Any control packet containing a word may be used.

FIG. 5B shows a data format of a voice packet of the present transmission system. That is, the voice packet is 10
It consists of a word string of bits, and the number of words is always 21 fixed. The ANC data flag (ADF) consists of three words (000h, 3FFh, 3FFh). Data ID (DI
D) indicates to which of voice groups 1, 2, 3 or 4 the voice packet belongs. The data block number (DBN) is a serial number assigned to audio packets belonging to the same audio group when the audio digital data of the audio group is subdivided into audio packets, and is a periodic number in the range of 1 to 255. Is changed. The data count (DC) indicates the number of words of the user data. The user data is a sub-frame (AD1, AD1) consisting of three words.
2, AD3 or AD4) or clock phase information (P
I). Subframe (AD1, AD2, A
D3 or AD4) is assigned data relating to audio digital data. The clock phase information (PI) of the audio digital data is used to synchronize the frequency and the phase with the video signal independent of the audio packet, and the clock phase information set based on the video signal That is. The checksum (CS) is assigned an error detection code of a voice packet. Here, the necessary conditions for the specification of the serial digital signal transmission system supported by the device of the present embodiment are that the audio packet is composed of a data block number (DBN), clock phase information (P
I) and checksum (CS).

Next, the configuration of this embodiment will be described with reference to FIG. That is, the buffer memory 1 is for temporarily storing and holding audio packets. CS error detection circuit 2
Is to detect an erroneous voice packet by using a checksum (CS) of the voice packet. The data block number detection circuit 3 detects a data block number (DBN) from the data of the voice packet. The clock phase information detection circuit 4 detects clock phase information (PI) from data of an audio packet. The sampling frequency detection circuit 5 detects the sampling frequency of the digital audio data included in the audio packet from the RATE of the control packet. The estimated value generation circuit 6 generates an estimated value of clock phase information for a voice packet read from the buffer memory 1. The control circuit 7 performs interpolation control. The interpolation data generation circuit 8 generates interpolation data of an audio packet corresponding to an audio packet requiring interpolation. The replacement insertion circuit 9 outputs either a voice packet input from the buffer memory 1 or interpolation data input from the interpolation data generation circuit 8 under the control of the control circuit 7. The delay circuit 10 delays the voice packet input from the replacement insertion circuit 9 by one voice packet.

Next, the operation of this embodiment will be described.

The control packet is one for each field.
Times, the sampling frequency detection circuit 5 at the beginning of the field
Is input to The sampling frequency detection circuit 5 detects the value of the sampling frequency from the RATE of the control packet and outputs the value to the estimated value generation circuit 6. The estimated value generation circuit 6 determines whether or not the next sampling frequency value is input.
The sampling frequency is stored and held.

The voice packet input to the apparatus of this embodiment is first input to the buffer memory 1, where it is temporarily stored and held. The buffer memory 1 stores and holds a plurality of voice packets, and the input / output order is based on first-in first-out. By the way, a voice packet may not be input to the buffer memory 1 and may be lost during transmission for some reason. Hereinafter, the missing voice packet is referred to as a missing voice packet.

The buffer memory 1 supplies the CS error detection circuit 2, the data block number detection circuit 3 and the clock phase information detection circuit 4 to the CS signal detection circuit 4 in accordance with the control signal of the control circuit 7. Output data.

The CS error detection circuit 2 includes a buffer memory 1
When the voice packet data output from is input,
The error of the voice packet is detected using the checksum (CS). When an error is detected, interpolation of the voice packet is necessary, and thus the CS error detection circuit 2
Outputs an error occurrence signal to the control circuit 7.

When the data of the voice packet output from the buffer memory 1 is input, the data block number detection circuit 3 detects a data block number (DBN). In parallel with the detection, the data block number detection circuit 3 also detects the data block number (DBN) of the audio packet immediately before the audio packet input from the delay circuit 10. The data block number detection circuit 3 detects the two data block numbers (DB
Based on N), a missing voice packet is detected, and if detected, a missing signal notifying the fact is output to the control circuit 7.

Here, the data block number (DB
N) is a serial number assigned to audio packets belonging to the same audio group when audio digital data of the same audio group is subdivided into audio packets, and a numerical value that periodically changes in the range of 1 to 255 It is. The voice packets are transmitted in the order in which the data block numbers (DBN) are given. Therefore, for example, when the DBN of the voice packet input from the delay circuit 10 to the data block number detection circuit 3 is 5, the DBN of the voice packet input from the buffer memory 1 to the data block number detection circuit 3 has an error. 6 if not
It is. In this case, if the DBN of the audio packet input from the buffer memory 1 to the data block number detection circuit 3 is 7, it is detected that the audio packet with the DBN of 6 is missing. Thus, by using the data block number (DBN),
It is possible to detect missing voice packets.

When the data of the voice packet output from the buffer memory 1 is input, the clock phase information detection circuit 4 detects the clock phase information (PI) from the data and outputs it to the control circuit 7.

In parallel with the process for outputting the clock phase information (PI), the estimated value generation circuit 6
Clock phase information (P
I) is detected.

Here, the clock phase information (PI) is based on the assumption that one field (including a blanking period) of a video signal as a main signal is regarded as a kind of coordinate, and the clock frequency of the video signal is used for the coordinate. It indicates at which position the audio signal exists. Therefore, for the sake of explanation, although the audio packet is not of fixed length, the clock frequency T of the video signal as the main signal is 27 MHz and the audio signal of the auxiliary signal is taken as an example in the case of the SMPTE 259M component system. Assuming that the sampling frequency is 48 kHz, the interval between two continuously transmitted audio packets is represented by the value of this clock phase information (PI), {clock frequency of video signal / sampling frequency of audio signal}. Is 562 or 563. In this case, the range of the clock phase information (PI) is from 0 to 1715.
become. Therefore, if the clock phase information (P
Assuming that the value of I) is 100, the value of the clock phase information (PI) of the audio packet input from the buffer memory 1 to the clock phase information detection circuit 4 is 662 or 66
3 is estimated.

As described above, based on the clock phase information (PI) of the voice packet input from the delay circuit 10 and the stored sampling frequency, the estimated value generation circuit 6 sends the clock phase information detection circuit 4 from the buffer memory 1.
And an estimated value of clock phase information corresponding to the clock phase information (PI) of the data of the voice packet input to. Then, the estimated value generation circuit 6 outputs the generated estimated value to the control circuit 7 and the interpolation data generation circuit 8.

When the clock phase information (PI) detected by the clock phase information detection circuit 4 is compared with the estimated value generated by the estimated value generation circuit 6, if the difference between the two is large, a voice packet is lost. It is considered that either the clock phase information (PI) has an error or an error has occurred. If the difference between the two is almost the same as one voice packet, it can be estimated that one voice packet is missing. Therefore, the estimated value generation circuit 6 has a function of detecting both a missing voice packet and a clock phase information (PI) error.

The control circuit 7 controls the interpolation based on the signals input from the CS error detection circuit 2, the data block number detection circuit 3, the clock phase information detection circuit 4 and the estimated value generation circuit 6.

When the control circuit 7 receives an interpolation request based on the error occurrence signal input from the CS error detection circuit 2,
The voice packet is output from the buffer memory 1 to the replacement insertion circuit 9. At the time of this output, the control circuit 7 causes the buffer memory 1 to erase the data of the audio packet.
Then, the control circuit 7 controls the interpolation data generation circuit 8 to generate interpolation data corresponding to the voice packet output from the buffer memory 1 and output the interpolation data to the replacement insertion circuit 9. The interpolation data generation circuit 8 generates interpolation data corresponding to the audio packet output from the buffer memory 1 using the audio packet input from the delay circuit 10 under the control of the control circuit 7. Replacement insertion circuit 9
Outputs the interpolation data input from the interpolation data generation circuit 8 without outputting the voice packet input from the buffer memory 1 under the control of the control circuit 7.

The control circuit 7 receives the notification of the detection of the missing voice packet based on the missing signal input from the data block number detecting circuit 3 and / or receives the clock phase information (input) from the clock phase information detecting circuit 4. P
I) and the estimated value input from the estimated value generation circuit 6,
When a missing voice packet is detected, the output of the voice packet stored and held in the buffer memory 1 on which the detection was made is delayed by one voice packet. In the meantime, the control circuit 7 controls the interpolation data generation circuit 8 to generate interpolation data corresponding to the missing voice packet and output the generated data to the replacement insertion circuit 9. The interpolation data generation circuit 8 generates interpolation data corresponding to the missing voice packet using the voice packet input from the delay circuit 10 under the control of the control circuit 7. The replacement insertion circuit 9 outputs the interpolation data generated by the interpolation data generation circuit 8 under the control of the control circuit 7.

After the interpolation data is output, if there is no interpolation request by the CS error detection circuit 2 for a voice packet whose output from the buffer memory 1 has been delayed by one voice packet, the control circuit 7 Is output from the buffer memory 1 to the replacement insertion circuit 9. At the time of this output,
The control circuit 7 causes the buffer memory 1 to erase the data of the audio packet. Then, the control circuit 7 causes the replacement insertion circuit 9 to output the voice packet to the delay circuit 10. After the interpolation data output, if there is an interpolation request by the CS error detection circuit 2 for a voice packet whose output from the buffer memory 1 has been delayed by one voice packet, the control circuit 7 stores the voice packet in the buffer memory. 1 to the replacement insertion circuit 9. At the time of this output, the control circuit 7 causes the buffer memory 1 to erase the data of the audio packet. Then, the control circuit 7 controls the interpolation data generation circuit 8 to generate interpolation data corresponding to the audio packet output from the buffer memory 1 and output the interpolation data to the replacement insertion circuit 9. The interpolation data generation circuit 8 generates interpolation data corresponding to the audio packet output from the buffer memory 1 using the audio packet input from the delay circuit 10 under the control of the control circuit 7. The replacement insertion circuit 9 outputs the interpolation data input from the interpolation data generation circuit 8 without outputting the voice packet input from the buffer memory 1 under the control of the control circuit 7.

The control circuit 7 does not need to interpolate a voice packet based on signals input from the CS error detection circuit 2, the data block number detection circuit 3, the clock phase information detection circuit 4 and the estimated value generation circuit 6. If so, a voice packet is output from the buffer memory 1 to the replacement insertion circuit 9. At the time of this output, the control circuit 7 causes the buffer memory 1 to erase the data of the audio packet. Then, the control circuit 7 causes the replacement insertion circuit 9 to output the voice packet to the delay circuit 10.

In this embodiment, the interpolation data generating circuit 8 generates the interpolation data corresponding to the audio packet to be interpolated. May be generated. Alternatively, interpolation data corresponding to clock phase information (PI) of an audio packet to be interpolated may be generated.

In this embodiment, the RA of the control packet is
Although the sampling frequency was detected using TE, the present invention is not limited to this, and the clock phase information (P
I) may be used to estimate the audio sampling frequency (such as 48 kHz for 562 or 563).
In the case of this method, the sampling frequency detection circuit 5 is configured to input the audio packet from the delay circuit 10 or the buffer memory 1 instead of the control packet, and use the clock phase information (PI) of the audio packet to obtain the sampling frequency. May be estimated.

In this embodiment, the data block number detecting circuit 3 includes the data block number detecting means and the missing voice packet detecting means or the first missing voice packet detecting means. However, the present invention is not necessarily limited to this configuration. Even if the data block number detection circuit 3 includes a data block number detection unit and the control circuit 7 includes the missing voice packet detection unit or the first missing voice packet detection unit according to the first aspect. Good. However, in this case, the signal line input from the delay circuit 10 to the data block number detection circuit 3 needs to be input not to the data block number detection circuit 3 but to the control circuit 7.

In this embodiment, the data block number detection circuit 3 reads the immediately preceding voice packet from the delay circuit 10 to detect a missing voice packet. However, the present invention is not limited to this configuration. The block number detecting circuit 3 (the missing voice packet detecting means or the first missing voice packet detecting means in claim 1) has a means for updating to the latest data block number (DBN) and storing the updated data block number (DBN). It is good. in this case,
The signal line input from the delay circuit 10 to the data block number detection circuit 3 is unnecessary.

Also, in this embodiment, the storage means is divided into two blocks, a buffer memory 1 and a delay circuit 10, and the voice packets after the voice packet being processed are read from the buffer memory 1, and the voice packet being processed is read out. 1 from packet
The previous voice packet was read out from the delay circuit 10, but is not necessarily limited to this configuration. As shown in FIG. 2, the delay circuit may be included in the buffer memory 1 as a delay memory. Alternatively, the voice packet may be read from 1. However, in this case, in order to be able to refer to the voice packets read out in the first-in first-out order even after the reading, the first-in first-out order is maintained and the voice packets stored in the buffer memory 1 are It is necessary to delay the erase timing.

Further, in the present embodiment, the voice packet referred to for interpolation is limited to the voice packet immediately before the voice packet being processed. However, the configuration of FIG. 2 is not necessarily limited to this. A plurality of voice packets prior to the voice packet being processed and / or a plurality of voice packets subsequent to the voice packet being processed may be referred to for interpolation.

In this embodiment, the audio packet of the serial digital signal transmission system corresponding to the device of this embodiment includes a data block number (DBN), clock phase information (PI), and a checksum (CS). Although it is assumed that the packet is a packet, the present invention is not limited to this. The clock phase information (PI) or the data block number (DB
N) and a packet including a checksum (CS). In short, the voice packet of the serial digital transmission system to which the device of the present invention corresponds is fixed length or variable length, and is treated as fixed length according to the rules at the time of transmission, and the clock phase information (PI) and And / or any packet containing a data block number (DBN) word.

Therefore, the data block number (DB
N) and / or the SMPTE 259M serial digital signal transmission method is not limited to the fixed-length voice packet including the clock phase information (PI). If the voice packet is fixed, the missing voice packet can be interpolated using the data block number (DBN). Furthermore, if the fixed-length clock phase information (PI) is added to the user data as shown in FIG. 5B, the data block number (D
BN) and clock phase information (PI) can be used to interpolate missing voice packets.

[0053]

As is apparent from the above description, according to the present invention, when an error or omission of a voice packet occurs, the voice packet can be replaced or inserted by interpolation data. Even if an error or omission occurs, it is possible to reproduce a corrected good music signal without noise.

[Brief description of the drawings]

FIG. 1 is a configuration diagram of an embodiment according to a voice packet interpolation device of the present invention.

FIG. 2 is a configuration diagram of another embodiment according to the voice packet interpolation apparatus of the present invention.

FIGS. 3A and 3B are diagrams showing a data format of a packet in a conventional serial digital signal transmission method, wherein FIG. 3A shows a control packet, FIG.
(C) is a diagram showing the data format of the additional packet.

FIG. 4A is a diagram showing the contents of a 30-bit bit address of the subframe AD1 or AD2 of the voice packet of FIG. 3B, and FIG. 4B is a diagram showing the contents of FIG. )
FIG. 10 is a diagram showing the contents of a 10-bit AUX bit address of the additional packet of FIG.

FIGS. 5A and 5B are diagrams showing a data format of a packet in a serial digital signal transmission system supported by the device of the present embodiment, wherein FIG. 5A shows a control packet, and FIG. It is a figure showing a format.

[Explanation of symbols]

DESCRIPTION OF SYMBOLS 1 Buffer memory ... Storage means 2 CS error detection circuit 3 Data block number detection circuit ... Data block number detection means, missing voice packet detection means, first missing voice packet detection means of claim 1 4 Clock phase information detection circuit ... Clock phase Information detecting means 5 Sampling frequency detecting circuit ... Sampling frequency detecting means 6 Estimated value generating circuit ... Estimated value generating means 7 Control circuit ... Missing packet detecting means, second missing voice packet detecting means of claim 2 8 Interpolation data generating circuit ... Interpolation Data generation means 9 Replacement insertion circuit ... Interpolation data insertion means 10 Delay circuit ... Storage means

Continuation of the front page (58) Field surveyed (Int.Cl. ⁷ , DB name) H04L 1/00 H04L 12/56 H04N 7/00

Claims (10)

(57) [Claims] A storage means for temporarily storing and holding audio packet data having audio data and data relating to other information, wherein the number of words of predetermined bits is always fixed; Data block number detecting means for detecting a data block number which is a serial number assigned to the voice packet from the data of the voice packet read from the storage means; and detecting a missing voice packet based on the data block number. The lost voice packet is detected from the storage unit based on the lost voice packet detection unit and the first-in first-out order.
Independent of the voice packet data
Between the video signal data and audio packet data
In order to synchronize the wave number and phase,
Clock that holds the position of the data of the voice packet
Clock phase information detecting means for detecting phase information ;
Sampling frequency detection to detect sampling frequency
Means, and data of a voice packet readable from the storage means.
Clock phase information included in the
Using the wave number and the
Estimate to generate an estimate of the clock phase information of the data
Generating means , based on a detection result of the missing voice packet detecting means,
If it has been dropped, it can be read from the storage means.
Voice packet data and the
The missing sound is estimated using the estimated value of the clock phase information.
Contains clock phase information data corresponding to voice packets.
An interpolation data generating means for generating interpolation data; and a missing data based on a detection result of the missing voice packet detecting means.
If the voice packet has been dropped,
Output the interpolation data based on the order in which
If not, output the data of the voice packet
An audio packet interpolating device comprising interpolation data inserting means . 2. Data relating to audio data and other information.
The number of words of a given bit is always fixed
A storage means for temporarily storing and holding voice packet data, and a voice packet having an error due to an error check code of the voice packet.
Voice packet error detecting means for detecting a packet , and reading from the storage means based on a first-in first-out order.
Independent of the voice packet data
Between the video signal data and audio packet data
In order to synchronize the wave number and phase,
Clock that holds the position of the data of the voice packet
Clock phase information detecting means for detecting phase information ;
Sampling frequency detection to detect sampling frequency
Means, and data of a voice packet readable from the storage means.
Clock phase information included in the
Using the wave number and the
Estimate to generate an estimate of the clock phase information of the data
Generating means and an error based on the detection result of the voice packet error detecting means.
Read out from the storage means if
Voice packet data and the estimated value generation means
By using the estimated value of the generated clock phase information, the error
Clock phase information corresponding to the voice packet in which
Interpolation data generation that generates interpolation data including the data of
Means based on the detection result of the voice packet error detection means.
If the audio packet has been lost,
Outputting the interpolation data based on the corresponding order,
If the voice packet has not been generated,
Data interpolation means for outputting data
Voice packet interpolation device . 3. Data relating to audio data and other information
The number of words of a given bit is always fixed
Storage means for temporarily storing and holding voice packet data; and reading from the storage means based on a first-in first-out order.
Independent of the voice packet data
Between the video signal data and audio packet data
In order to synchronize the wave number and phase,
Clock that holds the position of the data of the voice packet
Clock phase information detecting means for detecting phase information ;
Sampling frequency detection to detect sampling frequency
Means, and data of a voice packet readable from the storage means.
Clock phase information included in the
Using the wave number and the
Estimate to generate an estimate of the clock phase information of the data
Generating means, detected by the estimated value and the clock phase information detecting means
Missing or incorrect based on the clock phase information
Missing voice packet detection that detects generated voice packets
Means based on the detection result of the missing voice packet detecting means.
If a drop or error has occurred,
Using the readable voice packet data,
Complementary information for voice packets with missing or incorrect
Based on the detection result of the interpolated data generating means for generating the interim data and the missing voice packet detecting means.
If dropped or incorrect, the missing audio
The interpolation data is output based on the order corresponding to the packet.
If no dropouts or errors have occurred,
Interpolation data insertion means for outputting voice packet data.
An audio packet interpolation device, comprising: 4. Data relating to audio data and other information
The number of words of a given bit is always fixed
A storage unit for temporarily storing and holding voice packet data, and read from the storage unit based on a predetermined order.
From the data of the voice packet,
Data for detecting the data block number, which is the serial number
A block number detecting means and a missing audio pattern based on the data block number.
First lost voice packet detecting means for detecting a packet, and reading from the storage means based on a first-in first-out order.
Independent of the voice packet data
Circumference between the data of the data and voice packets that video signal
In order to synchronize the wave number and phase,
Clock that holds the position of the data of the voice packet
Clock phase information detecting means for detecting phase information ;
Sampling frequency detection to detect sampling frequency
Means, and data of a voice packet readable from the storage means.
Clock phase information included in the
Using the wave number and the
Estimate to generate an estimate of the clock phase information of the data
Generating means, detected by the estimated value and the clock phase information detecting means
Missing or incorrect based on the clock phase information
Second missing voice packet for detecting generated voice packet
Detecting means, the first missing voice packet detecting means, and the second missing sound
Based on the detection result of the voice packet detection means,
Read out from the storage means if
Using the data of the voice packet
Interpolation data for generating interpolation data corresponding to voice packets
Data generating means, the first missing voice packet detecting means and the second missing sound.
Based on the detection result of the voice packet detection means,
If the audio packet has been lost,
The interpolation data is output based on the corresponding order,
If no error or error has occurred, the voice packet
Interpolating data insertion means for outputting the data of
A voice packet interpolation device characterized by the above-mentioned. 5. Data relating to audio data and other information
The number of words of a given bit is always fixed
Storage means for temporarily storing and holding voice packet data; Voice packet with error due to error check code of voice packet
Voice packet error detecting means for detecting the packet, The supplementary data generation means detects the missing voice packet detection means.
Based on the output result and the detection result of the voice packet error detection means.
Is missing or missing?
Using the data of the voice packet that can be read from
Corresponding to the voice packet in which the missing or error occurred
Generating interpolation data , Claim 3
Voice packet interpolator. 6. Data relating to audio data and other information
The number of words of a given bit is always fixed
Storage means for temporarily storing and holding voice packet data; Voice packet with error due to error check code of voice packet
Voice packet error detecting means for detecting the packet, Supplementary data generation means for detecting the first missing voice packet
Means and the detection result of the second missing voice packet detecting means;
Based on the detection result of the voice packet error detection means,
Or if an error has occurred, read from the storage
Using the data of the voice packet that is possible,
Or interpolation data corresponding to the voice packet in which the error occurred.
5. The voice packet according to claim 4, wherein
Interpolator. 7. The audio data generating means according to claim 1, wherein
The voice data contained in the data of the
Audio data included in the dropped audio packet data
Generating corresponding interpolated data.
Voice packet interpolation according to 1, 2, 3, 4, 5, or 6
apparatus. 8. The audio data generating means according to claim 1, wherein
The clock phase information data included in the
Used in the data of the missing voice packet.
Generate interpolation data corresponding to clock phase information data
The method according to claim 3, 4, 5, or 6,
Voice packet interpolator. 9. The apparatus according to claim 1, wherein
Of the sampling frequency for the voice signal of the voice packet
Sampling of audio signals using control packets containing data
2. The method according to claim 1, wherein a ring frequency is detected.
The audio packet interpolation device according to 2, 3, 4, 5, or 6
Place. 10. The sampling frequency detecting means,
The audio packet clock readable from the storage means
Using the phase information, the sampling frequency of the audio signal
The method according to claim 1, wherein the detecting is performed.
7. The voice packet interpolator according to 4, 5, or 6.