JP2001144694A - Sampling rate converter - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a sampling rate converter that can enhance tracking performance to a change in input audio data packets. SOLUTION: The sampling rate converter is provided with an unpacket processing means 1 that detects time stamp information and block size information from a received data packet and provides an output of them, an input frequency calculation means 5 that calculates a rate corresponding to a sampling rate of main body data, an arithmetic means 3 that conducts sampling rate conversion arithmetic operation in matching with a sampling frequency of outputted data, and a frequency ratio calculation means 6 that calculates frequency ratio information on the basis of the input frequency information and an output sampling clock signal corresponding to the sampling frequency of the outputted data and gives the result to the arithmetic means 3, and the sampling rate converter conducts sampling rate conversion arithmetic operation in response to the frequency ratio information.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a sampling rate converter, and more particularly to an asynchronous sampling rate converter in which a sampling rate of an input signal used for digital audio and the like is asynchronous with a sampling rate of an output signal, such as an IEEE 1394 system. ,
The present invention relates to a sampling rate converter suitable for application to a data receiving device that receives data and additional information via a digital interface.

[0002]

2. Description of the Related Art In recent years, IEEE has been used as a serial transmission method.
(The Institute of Electric
al and Electronics Engineering
ers, Inc. ) The 1394 system has attracted attention.
The IEEE 1394 system uses a conventional SCSI (small
computer system interface
e) Not only can it be used in place of the transmission of computer data by a method or the like, but also audio and video AV
It can also be used for data transmission. This is IEEE
This is because the 1394 system defines two communication methods, asynchronous (asynchronous) communication and isochronous (isochronous) communication.

[0003] Isochronous communication is a data transmission method that can be used for transmission of data such as AV data that requires real-time performance. Prior to the start of transmission, a band necessary for data transmission is acquired. Then, by transmitting data using the band, real-time data transmission is guaranteed. On the other hand, asynchronous communication is a transmission method used for transmission of data that does not require real-time processing and device control, such as transmission of computer data.

[0004] Various methods have been proposed as transmission protocols on IEEE 1394, and one of them is called an AV protocol. The AV protocol is based on IEC (International El
electrotechnical Commissio
n) It is standardized as 61883, and defines a method of transmitting and receiving AV data requiring real-time performance by isochronous communication, a method of transmitting and receiving an instruction given to a device by asynchronous communication, and the like.

When data is transmitted and received by isochronous communication, a transmission and reception system is defined for each data format such as video data and audio data. For example, CD
Transmission and reception of audio data such as Audio and Music DataT determined by the 1394 Trade Association (hereinafter referred to as 1394TA)
transmission Protocol Rev
1.0 (hereinafter referred to as AM protocol) shows the transmission method.

Next, transmission of clock information of data on the IEEE 1394 bus will be described.

FIG. 6 is a block diagram showing transmission of clock information between a transmitting node and a receiving node on the IEEE 1394 bus. In the figure, 6400 is a cycle master node that issues a cycle start packet including a reference count, 6100 is a transmitting node, 6200 is a receiving node,
Reference numeral 6300 denotes an IEEE 1394 bus. IEEE13
As described above, various packets are transmitted in a time-division manner on the 94 bus, but two types of data lines of a reference count and an isochronous packet are used in FIG. 6 for simplification of description. Further, the IEEE 1394 bus connects the nodes in a one-to-one manner, and does not make a T-shaped branch on the cable. Therefore, in an actual connection, for example, the cycle master node 6400 and the transmission node 6100 are connected by one cable,
100 and the receiving node 6200 will be connected by another cable. Here, a T-shaped connection is shown conceptually to conceptually show that the three nodes have a common reference count. .

The cycle master node 6400 has a reference clock generation means 609 and a counter 606. Similarly, transmitting node 6100 also has reference clock generating means 608 and counter 605, and receiving node 6200 also has reference clock generating means 610 and counter 607. Although the reference clock generating means 608, 609, and 610 are independent, counters 605, 606, 607
Is corrected by the reference count issued by the cycle master node 6400. Therefore, the count value of each node has only an error smaller than the reference clock resolution of each node.

The transmission node 6100 receives an audio clock (for example, 44.1 kHz for a CD) and a count from the counter 605 to generate a time stamp, and a time stamp generating means 603; There is a packetizing means 601 for inputting 16-bit stereo PCM data) and the time stamp to generate an isochronous packet.

The receiving node 6200 compares the count of the counter 607 with the unpacketizing means 602 for receiving the isochronous packet and extracting the audio data and the time stamp and the time stamp, and generates a pulse when the two match. Means 604 and a PLL 611 for reproducing an audio clock based on the pulse are included.

By the way, when converting a digital audio signal into an analog signal due to a problem in sound quality, a high-purity reference clock is generated using a crystal oscillator, and digital-analog conversion (DA conversion) is performed using this clock. Is often done. In data transmission using IEEE 1394, transmission of clock information is a problem in terms of purity. In clock reproduction by the PLL 611, sufficient clock purity may not be obtained in some cases. Then, it is conceivable to connect a sampling rate converter. Alternatively, even when the audio clock of the transmitting node 6100 does not match the sampling frequency that the receiving node 6200 can handle, the sampling frequency is converted by the sampling rate converter.

FIG. 7 is a block diagram showing an example of a configuration for converting a sampling frequency in a conventional sampling rate converter. In FIG. 7, a receiving node 701 is shown.
Has the same function as the receiving node 6200 in FIG. 6, and a description thereof will be omitted. The frequency comparing means 705 outputs an audio clock reproduced by the PLL 611 output from the receiving node 701 and a D clock configured using a crystal oscillator.
The frequency ratio data is generated by comparing the output sampling clock from the AC clock generator 706 with the sampling clock output from the AC clock generator 706 and supplied to the sampling rate converter 702.

A sampling rate conversion unit 702 performs input sampling rate conversion based on the frequency ratio data obtained by the frequency ratio detection means 705, and
FO (first-in first-out) 703
Output to Sampling rate converter 7 here
For 02, for example, a conventionally known one may be used, and thus the details are omitted.

The DAC 704 includes a DAC clock generator 7
FIFO 70 according to the output sampling clock
3 is converted into an analog signal using a DAC clock synchronized with an output sampling clock generated by a DAC clock generator 706. Thus, the PL included in receiving node 701
A much higher quality analog audio signal can be obtained than when only L611 is used for reproduction.

FIG. 8 is a block diagram showing a configuration example of a frequency ratio detecting means in a conventional sampling rate converter. In this example, a PLL 611 of a receiving node 701 is used.
A counter 802 measures the audio clock reproduced by the counter 802, and fetches the reference clock output from the reference clock generator 706 into a register 803 using a signal obtained by dividing the reference clock by 4096 using a frequency divider 801. Since the previously acquired measurement value (output of the register 803) is written in the register 804, if this difference is obtained using the subtractor 806, frequency ratio data can be obtained. The output of the subtractor 806 is written into a register 807, which is used as frequency ratio data. The delay unit 805 delays the output of the frequency divider 801 by one machine cycle, and secures a subtraction time of the subtractor 806. This output is output as a data update pulse.

In this way, the problem of clock transmission in IEEE 1394 is solved to some extent, and when converting a digital audio signal to an analog signal, a high-purity clock signal is generated using a crystal oscillator. , Digital-analog conversion (DA conversion).

[0017]

However, in such a configuration, since the audio clock is reproduced once by the PLL and the sampling frequency is converted again using the audio clock, the PLL settling time and the sampling rate converter settling time are required. However, there is a problem that the ability to follow the change of the input audio data packet is deteriorated, and that the hardware scale is increased and the cost of the apparatus is increased.

The present invention solves the above-mentioned conventional problems, and makes it possible to measure the input frequency and the frequency of the output sampling clock on the same time base, thereby achieving high-speed response and stabilization of the sampling rate converter. It is an object of the present invention to provide a sampling rate converter that can significantly improve the ability to follow changes in input audio data packets and does not require audio clock reproduction by a PLL.

[0019]

SUMMARY OF THE INVENTION A sampling rate converter according to the present invention represents, from a received data packet, time stamp information and the size of data associated with the time stamp information as body data and attribute data of the body data. Unpacketizing means for detecting and outputting the block size information, and time stamp information and block size information output from the unpacketizing means,
Input frequency calculating means for calculating input frequency information corresponding to a sampling rate of input main body data, and calculating means for performing a sampling rate conversion operation for converting a sampling rate of input data to a sampling rate of output data And frequency ratio calculating means for calculating frequency ratio information from the input frequency information output by the input frequency calculating means and the output sampling clock signal corresponding to the sampling frequency of the output data, and supplying the information to the calculating means, The calculation means performs a sampling rate conversion calculation according to the frequency ratio information.

According to the present invention, the ability to follow changes in input audio data packets can be greatly improved, and the scale of hardware can be reduced.

[0021]

Embodiments of the present invention will be described below with reference to the drawings.

(Embodiment 1) FIG. 1 is a block diagram showing a configuration of a sampling rate converter according to Embodiment 1 of the present invention, FIG. 2 is an explanatory diagram of a data packet of the AM protocol, and FIG. 3 is a sampling rate of the present invention. FIG. 3 is a block diagram illustrating a configuration example of a difference SYT calculating unit according to the first embodiment of the converter.

In FIG. 1, the unpacketizing means 1 comprises:
It has the same function as the unpacketizing means 602 in FIG.
The audio data portion included in the input data packet is separated from the attribute data. An example of a data packet in the case of the AM protocol is as shown in FIG. 2, in which an 8-byte isochronous packet header, an 8-byte common isochronous packet header (CIP header), and m samples of stereo audio data are followed. And the data CRC. The CIP header includes a transmission node ID (SID), a size (DBS) of a data block (a set of data, for example, in the case of a CD, Lch and Rch are combined into one data block), and the number of data blocks. (D
BC), the format type (FMT) of the isochronous packet, and information (EV
T and SFC), a time stamp (SYT), and the like. EVT indicates the format of the data portion following the CIP header. The SFC specifies a sampling frequency. The PAC of the data portion is a synchronization signal in the digital audio transmission system defined by the IEC60958 standard, and P, C, U, and V are additional information also defined by the IEC60958 standard. In the case of stereo audio data, the data block is subframe 1 (su in the figure).
b-frame1) corresponds to the L channel, and subframe 2 (sub-frame2 in the figure) corresponds to the R channel.

For the sake of simplicity, it is assumed that one packet includes a fixed number of blocks, that is, a method called a blocking method in which audio samples are included. In the case of the blocking method, the number of audio samples included in one packet, that is, the SYT interval is
It is fixed for each sampling frequency. For example, the sampling frequency is 32 kHz, 44.1 kHz or 48 kHz.
In the case of kHz, there are eight samples. Therefore, SY
The time stamp information indicated by T is time information for every eight samples.

Returning to FIG. 1, reference numeral 2 denotes SYT from the SYT separated by the unpacketizing means and the SYT eight samples before.
This is a difference SYT calculating means for calculating an increment of T, and is configured, for example, as shown in FIG. In FIG.
Is a register for storing the SYT of the previous packet, and 22 is a subtractor. The SYT of the previous packet is calculated by the subtractor 22.
And the current SYT difference is calculated. The sample number calculation means 8 in FIG. 1 decodes from the SFC included in the data packet described in FIG. 2 how many samples the SYT is transmitted once. In the case of 48 kHz, a value of "8" is obtained. The input frequency calculation means 5 includes the difference SYT calculation means 2
Is divided by the number of samples output from the number-of-samples calculating means 8 to obtain sampling period information TIN of the input audio data, which is supplied to the frequency ratio calculating means 6. The output frequency calculation means 7 includes a DAC
The period TOUT of the output sampling clock signal of the output audio data synchronized with the clock used in the (digital-analog converter) is determined by the count output of the counter 9 synchronized by the reference count from the cycle master node described in FIG. It is gained by counting. The reference clock generation means 10 is equivalent to the reference clock generation means 608, 609, 610 described in FIG.
The frequency ratio calculating means 6 divides the output of the input frequency calculating means 5 by the output of the output frequency calculating means 7 to obtain the frequency ratio data. The FS conversion operation unit converts the input sampling rate based on the frequency ratio data output from the frequency calculation means 7 and outputs the result to the FIFO 4. As the FS conversion operation unit 3 here, for example, a conventionally known one may be used, and thus the details are omitted. The FIFO 4 takes in the output of the FS conversion operation unit 3 and outputs the output timing in synchronization with the output clock.

As described above, according to the present embodiment, the input frequency obtained from the information of the time stamp and the number of samples added at the transmitting node and the frequency of the output sampling clock are measured on the same time base. As a result, an accurate frequency ratio can be calculated immediately, and as a result, a high-speed response and stabilization of the sampling rate converter can be achieved, so that the ability to follow changes in the input audio data packet is greatly improved, and a PLL is also required. Therefore, the hardware scale of the apparatus can be reduced.

(Embodiment 2) In this embodiment, the difference S
Since the configuration is the same as that of the first embodiment except for the configuration of the YT calculation means, the description of the same parts will be omitted, and the SYT calculation means, which is the feature thereof, will be mainly described.

FIG. 4 is a block diagram showing a configuration example of the difference SYT calculating means in the sampling rate converter according to the second embodiment of the present invention. FIG. 5 is a block diagram showing the difference SYT calculating means in the sampling rate converter according to the second embodiment of the present invention. FIG.

In FIG. 4, reference numeral 401 denotes an accumulator for accumulating the sampling period information TIN of the input audio data output from the input frequency calculating means 5 of FIG. 1 by the number of samples, a register 402 and a subtractor 403. Figure 3
Are the same as the register 21 and the subtractor 22, respectively. FIG.
The difference between the difference SYT calculation means 2 of the first embodiment shown in FIG. 1 and the difference SYT calculation means 40 used in the present embodiment is that the difference SYT calculation means 2 of FIG. While the sampling period of the input audio data is obtained by division, in the present embodiment, the difference between the accumulation result of the past sampling period information TIN of the input audio data and the current SYT is represented by the number of samples. The sampling period of the input audio data is obtained by the division. In this way, it is possible to remove the influence of a minute calculation error in the difference SYT calculation means. That is, in the first embodiment, the sampling error information TIN
Contains a slight error with respect to the ideal value, and this error may cause an overflow or an underflow of the number of samples to be input.

In this embodiment, as shown in FIG. 5, instead of SYT (n) which is the SYT of the first embodiment, SYT which is the accumulation result of sampling period information TIN is used.
By using (n) ', the difference between SYT (N + 1) and SYT (n)', that is, .DELTA.SYT 'is output. In FIG. 5, for simplicity, the SYT interval is set to 4, so that the sampling period information TIN of the input audio data output from the input frequency calculation means 5 is as shown in (Equation 1).

[0031]

## EQU1 ## TIN = ΔSYT (n) ′ 累 4 This is further accumulated to the past accumulated value of the accumulator 401 for four samples to obtain SYT (n + 1) ′. By doing so, the time error DI for the calculation error is instantaneously obtained.
Even if F occurs, the error is canceled out and thus is not accumulated, and there is no fear of overflow or underflow of the number of input samples.

As described above, according to the present embodiment, in addition to the features of the first embodiment, even if a time error DIF corresponding to a calculation error occurs due to a calculation error in the difference SYT calculating means, the error is reduced. Since they are canceled and not accumulated, there is no fear of overflow or underflow of the number of input samples, and the high-speed response and stabilization of the sampling rate converter are further improved.

In each of the above embodiments, for simplification of the description, the description has been made only on the blocking method of the AM protocol for audio transmission in IEEE1394, but the present invention is not limited to the block (sample) included in one packet. Number) is variable, the time stamp information (SYT) and the number of samples included in the interval between the time stamp information are transmitted.
Be adaptable. In addition, other than audio transmission, for example, I
It goes without saying that the present invention can be similarly applied to the MPEG transport and the DVCR transport defined in EC61883.

[0034]

As described above, according to the present invention, it is possible to measure the input frequency and the frequency of the output sampling clock on the same time base, thereby improving the high-speed response and stabilization of the sampling rate converter. Accordingly, the ability to follow the change of the input audio data packet can be greatly improved, and the audio clock reproduction by the PLL is not required. Therefore, the advantageous effect that the hardware scale can be reduced accordingly is obtained. Can be

[Brief description of the drawings]

FIG. 1 is a block diagram illustrating a configuration of a sampling rate converter according to a first embodiment of the present invention.

FIG. 2 is an explanatory diagram of a data packet of the AM protocol.

FIG. 3 is a block diagram showing a configuration example of a difference SYT calculating means in the sampling rate converter according to the first embodiment of the present invention;

FIG. 4 is a block diagram showing a configuration example of a difference SYT calculating means in a sampling rate converter according to a second embodiment of the present invention;

FIG. 5 is an explanatory diagram of a difference SYT calculating means in a sampling rate converter according to a second embodiment of the present invention;

FIG. 6 is a block diagram showing transmission of clock information between a transmitting node and a receiving node on an IEEE 1394 bus;

FIG. 7 is a block diagram showing a configuration example in a case where a sampling frequency is converted in a conventional sampling rate converter.

FIG. 8 is a block diagram showing a configuration example of a frequency ratio detecting means in a conventional sampling rate converter.

[Explanation of symbols]

 REFERENCE SIGNS LIST 1 unpacketizing means 2 difference SYT calculating means 3 FS conversion calculating unit 4 FIFO 5 input frequency calculating means 6 frequency ratio calculating means 7 output frequency calculating means 8 sample number calculating means 9 counter 10 reference clock generating means

Claims (6)

[Claims] 1. An unpacket for detecting and outputting, from a received data packet, time stamp information and block size information representing the size of data associated with the time stamp information as body data and attribute data of the body data. Means for calculating input frequency information corresponding to the sampling rate of the input main data from the time stamp information and block size information output from the unpacketizing means, and the sampling rate of the input data Calculating means for performing a sampling rate conversion operation for converting the output frequency into a sampling rate of output data; and input frequency information output from the input frequency calculating means and a frequency obtained from an output sampling clock signal corresponding to the sampling frequency of the output data. Ratio information A sampling rate converter, comprising: frequency ratio calculating means for calculating and supplying the calculated ratio to the calculating means, wherein the calculating means performs a sampling rate conversion calculation according to the frequency ratio information. 2. An unpacket for detecting and outputting, from a received data packet, time stamp information and block size information representing the size of data associated with the time stamp information as body data and attribute data of the body data. Means for calculating input frequency information corresponding to the sampling rate of the input main data from the time stamp information and block size information output by the unpacketizing means; Synchronization timing signal generating means for obtaining a timing signal synchronized with the reference timing signal used for generating the stamp, an output sampling clock signal corresponding to the sampling frequency of the data to be output, and an output from the synchronization timing signal generating means From the timing signal, Output frequency calculating means for calculating force frequency information, calculating means for performing a sampling rate conversion operation for converting a sampling rate of input data to a sampling rate of output data, and an output of the input frequency calculating means. Frequency ratio calculating means for calculating frequency ratio information from input frequency information and output frequency information output from the output frequency calculating means, and supplying the calculated frequency ratio information to the calculating means, wherein the calculating means performs sampling in accordance with the frequency ratio information. A sampling rate converter for performing a rate conversion operation. 3. The input frequency calculation means calculates difference information between the previous time stamp information and the current time stamp information, and calculates the sample number information of data to be reproduced between the previous time stamp information and the current time stamp information. 2. A sampling rate converter according to claim 1, wherein the sampling rate converter calculates and substantially divides the difference information by the sample number information. 4. An input frequency calculating means calculates difference information between the previous time stamp information and the current time stamp information, and calculates information on the number of samples of data to be reproduced between the previous time stamp information and the current time stamp information. 3. The sampling rate according to claim 2, wherein the calculation is performed, and the difference information is substantially divided by the sample number information to obtain period information of the input data, thereby obtaining input frequency information. converter. 5. An input frequency calculating means calculates difference information between the previous time stamp information and the current time stamp information, and obtains information on the number of samples of data to be reproduced between the previous time stamp information and the current time stamp information. By calculating and substantially dividing the difference information by the sample number information, the period information of the input data is obtained, whereby the input frequency information is obtained, and the period information of the input data is accumulated by the number of samples. 2. The sampling rate converter according to claim 1, wherein the calculated time stamp information is used as previous time stamp information. 6. The input frequency calculation means calculates difference information between the previous time stamp information and the current time stamp information, and calculates information on the number of samples of data to be reproduced between the previous time stamp information and the current time stamp information. By calculating and substantially dividing the difference information by the sample number information, the period information of the input data is obtained, whereby the input frequency information is obtained, and the period information of the input data is accumulated by the number of samples. 2. The sampling rate converter according to claim 1, wherein the calculated time stamp information is used as previous time stamp information.