JP4338083B2 - Digital audio playback device - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a digital audio reproducing device in which the occurrence of sound interruption or delay during audio data reproduction is reduced as much as possible in audio data communication. <P>SOLUTION: PCM data (a) received by a packet receiving section 1 are temporarily accumulated in a buffer memory 2. When a buffer accumulation amount (b) becomes equal to or greater than an initial delay amount Di, an initial delay judgment section 10 starts transferring data (c) in the buffer memory 2 to a sampling frequency varying section 7. A reproduced frequency determining section 3 determines a reproduced sampling frequency fsd based on a time difference &Delta;D from a target delay amount (a target buffer level)Dt in the buffer memory 2 for each fixed time and controls the sampling frequency of the sampling frequency varying section 7. Moreover, when the buffer accumulation amount (b) exceeds a maximum delay amount, the received PCM data are discarded. As a result, digital audio reproduction having reduced sound interruption and delay is realized. <P>COPYRIGHT: (C)2005,JPO&amp;NCIPI

Description

  The present invention relates to a digital audio reproducing apparatus, and more particularly to a digital audio reproducing apparatus that prevents interruption or delay of reproduced audio on the receiving side.

  In real-time communication typified by a digital cellular phone, an IP phone, a video phone, etc., digital audio data generated on the transmission side is transmitted, and the reception side that has received it reproduces it immediately. In digital encoding of an audio signal, the audio signal is sampled at a preset sampling frequency, and then quantization, band compression, and the like are performed to generate digital audio data.

  In the IP communication that has been generalized in recent years, the generated digital voice data forms a data packet by collecting data at regular intervals, and is transmitted / received in units of packets. On the receiving side, digital audio data is extracted from the received packet, decoded, and then reproduced based on the sampling frequency.

A technique for preventing the reception buffer from overflowing or underflowing due to a clock difference between the transmitting side and the receiving side or sound interruption due to packet jitter is disclosed in Patent Document 1 below. It is written.
JP 2001-045065 A

  At this time, the playback side may be interrupted or delayed on the receiving side for the following reasons.

  Since the apparatus is physically different between the transmission side and the reception side, there is a slight error in the sampling frequency between the apparatuses. At this time, if the sampling frequency on the receiving side is higher than that on the transmitting side, the time length of the audio data during playback is shorter than the time length of the generated audio data, so the received data is less than the specified playback time. It will be consumed quickly, resulting in sound interruptions.

  On the contrary, when the sampling frequency on the receiving side is lower than that on the transmitting side, the reproduction time of the audio data becomes longer than that at the time of generation, and the received data cannot be consumed within the specified time. In this case, if there are an abundance of buffers that are temporary storage locations for the data, the audio data accumulates in the buffer, so the audio transmission delay increases with time. If the buffer is small and cannot be stored, or if it is not stored intentionally in order to suppress delay, the received data is discarded and the reproduction of the audio data becomes discontinuous, resulting in sound interruption.

  Also, in the transmission of audio data, even if the sampling frequencies on the transmission side and reception side are completely the same, due to fluctuations in the transmission path state, sound interruptions during playback, increased delay, and audio data discontinuity Such a problem may occur.

  For example, when the network transmission rate changes sequentially, the packet transmission delay increases when the transmission rate drops temporarily from the specified value. Such a packet is accumulated in the transmission buffer. At this time, the arrival of the packet is delayed on the receiving side, resulting in a shortage of received data, resulting in sound interruption. After that, assuming that the transmission speed has recovered and the speed has become sufficiently high, in this case, the packets that have been stored in the transmission buffer until then are transmitted to the receiving side at once, and then the audio data is stored on the receiving side. The delay in voice transmission remains increased. Alternatively, when the buffer is small and cannot be stored, or when it is not stored intentionally to suppress delay, the data is discarded and the reproduction of the audio data becomes discontinuous.

  The present invention has been made in view of the above-described problems of the prior art, and an object of the present invention is to provide a packet arrival interval when there is an error in sampling frequency between the transmission side and the reception side in voice data communication. It is an object of the present invention to provide a digital audio reproducing apparatus that can reduce problems such as sound interruption or delay when reproducing audio data as much as possible in a case where there is a variation in the above or both of these situations.

To achieve the above objects, the present invention provides a digital audio reproducing apparatus for reproducing digitally encoded audio data, a buffer memory for temporarily storing digital audio data, the amount of data accumulated in the buffer memory An initial delay determination unit that determines whether or not the initial delay amount has been exceeded , and at least a target delay based on the amount of data stored in the buffer memory, the sampling frequency on the transmission side, and the number of quantization bits of the digital audio data A reproduction time determining unit for determining a reproduction time of the digital audio data temporarily stored in the buffer memory every predetermined time so as to obtain a time difference from the volume and eliminating the time difference; and the reproduction time determining unit A data reproducing unit for reproducing the digital audio data in accordance with the reproduction time determined by A maximum delay determination unit for determining whether the amount of data stored in the memory exceeds a maximum delay amount, and the buffer memory when the amount of data stored in the buffer memory exceeds the initial delay amount; The data reproduction unit reproduces audio data according to the reproduction time sequentially determined by the reproduction time determination unit, and the amount of data stored in the buffer memory exceeds the maximum delay amount. In this case, the received voice data is discarded.

  According to the present invention, in audio data communication, if there is an error in the reproduction frequency or reproduction speed such as the sampling frequency between the transmitting side and the receiving side, problems such as sound interruption or delay during audio data reproduction are minimized. Can be reduced.

  In addition, when there is a change in the packet arrival interval, it is possible to reduce problems such as sound interruption, delay, or change in sound pitch when reproducing audio data.

  Hereinafter, the present invention will be described in detail with reference to the drawings. FIG. 1 is a block diagram showing a configuration of an embodiment of the present invention.

  The packet receiver 1 receives a packet of digital voice data (digitally encoded musical tone data, conversation data, etc.) from an external network such as an IP network, and extracts digital voice data (PCM data) from the packet. If the digital audio data is compressed, it is decoded by a decoder corresponding to the compression method, and uncompressed audio PCM data a is output. The PCM data a is temporarily stored in the buffer memory 2, and the accumulated amount b is input to the reproduction frequency determining unit 3. The accumulated amount b is obtained by subtracting the value of the read number counter 5 from the value of the write number counter 4 by the subtractor 6, for example.

  The reproduction frequency determination unit 3 calculates an optimum reproduction frequency at regular intervals from a preset target delay amount Dt and buffer memory accumulation amount b. The sampling frequency variable unit 7 which is a data reproducing unit D / A converts the PCM data c in the buffer memory 2 with the reproduction sampling frequency fsd given from the reproduction frequency determining unit 3 and outputs it as an analog signal. The analog audio signal d is reproduced by a reproduction device 9 such as a speaker or a headphone through an amplification amplifier 8.

  The reproduction sampling frequency fsd of the reproduction frequency determination unit 3 is assumed to be the sampling frequency fs on the transmission side in the initial state. However, even when fsd = fs, there is a physical error for each device, so the actual reproduction frequency is not completely the same on the transmission side and the reception side.

  When the reception of the digital audio data packet is started in the packet receiver 1 and the decoded PCM data a starts to be stored in the buffer memory 2, the count value of the write counter 4 is counted up and increased. When the buffer accumulation amount b reaches the initial delay amount Di, the buffer memory 2 starts outputting the PCM data c therein. That is, when the buffer accumulation amount b becomes larger than the predetermined initial delay amount Di, the initial delay determination unit (for example, the comparator) 10 outputs an H (high) level signal and turns on the switch 11. As a result, output of the PCM data c in the buffer memory 2 to the variable sampling frequency D / A converter 7 is started.

  After that, the reproduction frequency determination unit 3 determines the reproduction sampling frequency at a predetermined interval Ti that is set in the following procedure.

  That is, if the average buffer memory storage amount in the past Tp in the buffer memory 2 is Bp, the quantization bit number of the PCM data a is Qb, and the number of audio channels is Cn, the target delay amount (target buffer level) in the buffer memory 2 The time difference ΔD from Dt is as follows.

  △ D = Bp / (fs × Cn × Qb / 8) −Dt

  If this difference is eliminated at time Tx, the reproduction sampling frequency fsd to be set is expressed by the following equation.

  fsd = fs × (1 + △ D / Tx)

  The reproduction sampling frequency fsd is sent from the reproduction frequency determination unit 3 to the sampling frequency variable unit 7. The sampling frequency variable unit 7 varies the reproduction sampling frequency in accordance with the reproduction sampling frequency fsd, and performs D / A conversion.

  In addition, when a large network jitter occurs, a large amount of data can be stored in the buffer memory at a time. In this case, the average buffer memory accumulation amount Bp in the above equation becomes very large, and the reproduction frequency fsd increases significantly. As a result, the reproduction sampling frequency fsd changes suddenly, which may cause unnatural hearing, so if the buffer accumulation amount exceeds the maximum buffer accumulation amount (or maximum delay amount) Dm, it will be received. The packet is discarded. That is, when the buffer accumulation amount b exceeds the maximum buffer accumulation amount Dm, the output of the maximum delay determination unit (for example, the comparator) 12 becomes H level, and the switch 13 is switched to the discard side.

  The sampling frequency changing method in the sampling frequency variable unit 7 is a simple method using a VCO (voltage controlled oscillator) or the like, or a memory write speed and a read speed disclosed in Japanese Patent No. 2853147. It is possible to apply a known method such as a method of changing the reproduction frequency by changing.

  According to this embodiment, since the data reading from the buffer memory is started when the amount of data stored in the buffer memory exceeds the initial delay amount, sound interruption can be prevented as much as possible. In addition, since the sampling frequency is varied in the sampling frequency variable unit so as to approach the target delay amount, it is possible to prevent sound interruptions and delays when reproducing audio data as much as possible. In addition, since the received audio data is discarded when the amount of data stored in the buffer memory exceeds the maximum delay amount, it is possible to prevent problems such as a change in sound pitch as much as possible.

  Next, a second embodiment of the present invention will be described with reference to FIG. This embodiment is characterized in that the reproduction speed is changed instead of changing the sampling frequency. In addition, while attaching | subjecting the same code | symbol to the function part same or equivalent to FIG. 1, the description is abbreviate | omitted.

  In the present embodiment, the reproduction speed coefficient Vd in the initial state is set to 1.0 so as to be the same as that on the transmission side. However, even when Vd = 1.0, there is a physical error for each device, so the actual reproduction speed is not completely the same as the input speed on the transmission side.

  After the start of reception, when the buffer accumulation amount b reaches the initial delay amount Di, the output of the PCM data c in the buffer memory 2 is started (the switch 11 is turned on when the output of the initial delay determination unit 10 becomes H). ). Thereafter, the playback speed determination unit 21 determines the playback speed at predetermined intervals Ti set in the following procedure.

  When the average buffer memory storage amount in the past Tp in the buffer memory 2 is Bp, the number of quantization bits of PCM data is Qb, and the number of audio channels is Cn, the time difference ΔD from the target delay amount Dt in the buffer memory 2 is The following equation is obtained. Note that fs is the sampling frequency on the transmission side.

  △ D = Bp / (fs × Cn × Qb / 8) −Dt

  When this difference is eliminated at time Tx, the reproduction speed coefficient Vd to be revised is expressed by the following equation.

  Vd = 1.0 + △ D / Tx

  The reproduction speed coefficient Vd is sent from the reproduction speed determination unit 21 to a reproduction speed (speech speed) conversion unit 22 which is a data reproduction unit. The playback speed conversion unit 22 varies the playback speed according to the playback speed coefficient Vd. As a method of changing the playback speed in the playback speed conversion unit 22, a known method such as a speech speed conversion apparatus of Japanese Patent No. 3187242 that changes only the speed without changing the pitch can be applied.

  The packet discard based on the maximum delay amount is the same as that in the first embodiment.

  Also in this embodiment, the same effect as that of the first embodiment can be obtained.

It is a block diagram which shows the structure of 1st Embodiment of this invention. It is a block diagram which shows the structure of 2nd Embodiment of this invention.

Explanation of symbols

  DESCRIPTION OF SYMBOLS 1 ... Packet receiving part, 2 ... Buffer memory, 3 ... Reproduction frequency determination part, 7 ... Sampling frequency variable part, 10 ... Initial delay determination part, 12 ... Maximum delay determination , 21... Playback speed determination section, 22... Playback speed (speech speed) conversion section.

Claims (3)

In a digital audio reproducing apparatus for reproducing digitally encoded audio data,
A buffer memory for temporarily storing digital audio data;
An initial delay determination unit that determines whether the amount of data stored in the buffer memory exceeds the initial delay amount;
At least the time difference from the target delay amount is obtained based on the amount of data stored in the buffer memory, the sampling frequency on the transmission side, and the number of quantization bits of the digital audio data, and so as to eliminate the time difference. A playback time determination unit for determining the playback time of digital audio data temporarily stored in the buffer memory every predetermined time ;
A data reproduction unit for reproducing the digital audio data according to the reproduction time determined by the reproduction time determination unit;
A maximum delay determination unit that determines whether the amount of data stored in the buffer memory exceeds the maximum delay amount;
When the amount of data stored in the buffer memory exceeds the initial delay amount, data reading from the buffer memory is started, and the data reproducing unit is operated according to the reproduction time sequentially determined by the reproduction time determining unit. A digital audio reproducing apparatus for reproducing audio data and discarding the received audio data when the amount of data stored in the buffer memory exceeds the maximum delay amount. 2. The digital audio reproduction apparatus according to claim 1, wherein the reproduction time determination unit determines a reproduction frequency for every predetermined time, and the data reproduction unit changes the reproduction frequency. 2. The digital audio reproducing apparatus according to claim 1, wherein the reproduction time determination unit determines a reproduction speed for every predetermined time, and the data reproduction unit changes the reproduction speed.

Images