JP2007096451A - Ip phone and buffer control method - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide an IP phone which does not cause voice quality deterioration in the case of carrying out voice processing by software, and a buffer control method. <P>SOLUTION: The IP phone includes: a PCM input device for capturing input voice and converting it into PCM voice data, a voice CODEC for applying coding processing for an IP phone call to the PCM voice data, and a buffer for storing the PCM voice data, and a buffer monitor transfer means that stores the PCM voice data and transfers at least part of the stored PCM voice data from the buffer to the voice CODEC when the PCM voice data storage amount reaches a prescribed amount. The technology above includes the buffer control method of controlling write to/read from the buffer. <P>COPYRIGHT: (C)2007,JPO&INPIT

Description

  The present invention converts an input voice into PCM voice data, supplies the PCM voice data to a VoIP network after encoding it in conformity with the VoIP telephone system, and a buffer buffer for storing the PCM voice data It relates to a control method.

  Conventionally, such PCM audio data processing generally uses hardware such as a DMA controller and a buffer memory, as disclosed in Patent Document 1. Recently, however, attempts have been made to realize such processing by software.

As a realization form by such software, a VoIP telephone application including a software codec is usually mounted on a computer apparatus provided with a PCM input device. The VoIP phone application specifies a sampling rate such as 16 kHz, 44.1 kHz or 48 kHz for the PCM input device. The PCM input device captures input sound, generates PCM sound data at such a sampling rate, and writes it in a buffer memory on the memory. Next, the codec processing unit included in the VoIP telephone application reads the PCM audio data written from the buffer memory at a predetermined time interval, not the data amount, and performs the necessary encoding processing. . As an example of this time interval, in the case of a stereo PCM audio signal with a sampling rate of 16 kHz and a bit rate of 8 kHz, 640 kbytes of PCM audio data must be read every 10 ms.
Japanese Patent No. 0351670

  However, when PCM audio data processing is to be realized by software, the hardware performance of the PCM input device, the CPU speed of the computer device on which the VoIP telephone application operates, the performance of the control program, and the load of other applications that operate simultaneously An event occurs in which PCM audio data of a required size cannot be read from the buffer at a time interval always determined by a variable factor such as the above. Such an event causes a loss of audio information and causes a decrease in sound quality.

  An object of the present invention is to provide an IP telephone apparatus and a buffer control method that do not cause deterioration in sound quality when performing voice processing by software.

  An IP telephone apparatus according to the present invention is an IP telephone apparatus that includes a PCMPCM input device that takes input voice and converts it into PCM voice data, and a voice codec that performs encoding processing for the IP phone on the PCM voice data. Yes, the buffer for storing the PCM audio data, and the PCM audio data storage amount in the buffer are monitored, and when the PCM audio data storage amount reaches a predetermined amount, at least a part of the PCM audio data stored therein And a buffer monitoring transfer means for transferring the data from the buffer to the audio codec.

  The buffer control method according to the present invention is a buffer control method in a computer device that takes in an input voice, converts it into PCM voice data, and performs voice processing on the PCM voice data, and stores the PCM voice data in a buffer. Monitoring the amount of PCM audio data stored in the buffer and when the PCM data storage amount reaches a predetermined amount, at least a part of the stored PCM audio data is transferred from the buffer to the audio processing. And a monitoring transfer step.

  According to the IP telephone apparatus and the buffer control method of the present invention, an IP telephone apparatus and a buffer control method that do not cause deterioration in sound quality when performing voice processing by software are provided.

  Embodiments of the present invention will be described in detail with reference to the accompanying drawings.

  FIG. 1 shows an embodiment of the present invention and shows the entire configuration including an IP telephone apparatus. The IP telephone device 10 is realized as a single computer device, and includes a CPU 11 that is a central processing unit, a memory 12 that is a storage device (hereinafter referred to as MEM 12), a PCM input device 13, and a VoIP that accommodates an IP phone. A communication device 14 that can be connected to a network is included, and these are interconnected by a bus 15. Although not shown, the IP telephone device 10 naturally includes other units related to audio output in order to perform the IP telephone function. In the following description, the description relating to the sound output is omitted.

  The PCM input device 13 has a function of taking external input sound from a sound collecting device such as a microphone, converting it into PCM according to a preset operating condition setting value, and supplying it to the bus 15 as PCM sound data. The operating condition setting value includes setting values such as the number of quantization bits, the sampling rate, and the number of channels, and is set according to the device management function of the control program described later when the PCM input device 13 is initially set. As an operation condition setting value as an example, it is assumed that the number of quantization bits is 16 bits, the sampling rate is 16 kHz, and the number of channels is 2 for stereo sound. In this case, the bit rate is 512 kbps.

  The MEM 12 includes a control program (OS) 20 that has a basic control function of the entire apparatus and a device management function of a device such as the PCM input device 13 after the IP telephone apparatus 10 is turned on, and a VoIP telephone that manages the IP telephone function. An application 30, a PCM input device driver 31, and a communication control unit 35 are loaded. The VoIP telephone application 30 can include a buffer 32, a buffer control unit 33, and a PCM voice codec unit 34.

  The PCM input device driver 31 has a function of acquiring PCM audio data from the PCM input device 13 via the bus 15 and writing it in the buffer 32. Also, the PCM input device driver 31 monitors the amount of PCM audio data stored in the buffer 32, generates a read permission notification when this amount reaches a predetermined amount, and supplies this to the buffer controller 33. It has a function.

  The buffer control unit 33 has a function of reading at least a part of the stored PCM audio data from the buffer 32 and transferring it to the PCM audio codec unit 34, triggered by this read permission notification.

  The PCM audio codec unit 34 has a function of encoding, that is, encoding, the PCM audio data transferred from the buffer control unit 33 using a codec method designated in advance based on the specifications of the VoIP telephone. The codec method may be a multi-codec method capable of selectively switching between a plurality of methods, and may be selected and changed dynamically in response to the setting of the operating condition for the PCM input device 13 by the device management function of the control program 20. good.

  The communication control unit 35 has a function of converting the encoded PCM audio data into a VoIP signal for IP telephone communication and sending it to the communication device 14 via the bus 15.

  FIG. 2 explains how data is written to and read from the buffer shown in FIG. The buffer 32 is divided into two buffer planes, blocks B1 and B2. The block B1 includes, for example, three records Buf1-1 to 3, and the block B2 includes, for example, three records Buf2-1 to 3. PCM audio data is stored in each of the plurality of records Buf1-1 to Buf1-1 to Buf1-1 to 3. Each of the blocks B1 and B2 is a permission unit that is a target of a read permission notification. For example, when the block B1 is permitted to be read, the block B2 is in a state where reading is not permitted. On the other hand, when the block B2 is permitted to be read, the block B1 is in a state where reading is not permitted.

  The PCM input device driver 31 stores the PCM audio data in each of the three records Buf1-1 to Buf1-1 and sequentially writes them in the buffer plane, that is, one block B1. When the PCM input device driver 31 finishes writing to one block B1, the PCM audio data is stored in each of the next three records Buf2-1 to 1-3 and sequentially written to the other one block B2. Here, the PCM input device driver 31 generates a read permission notice and transfers it to the buffer control unit 33 when writing is transferred from the block B1 to the block B2.

  On the other hand, the buffer control unit 33 that has been in the read waiting state sequentially reads the records Buf1-1 to 3 stored in the block B1 using the read permission notification as a trigger. The buffer control unit 33 that has finished reading is again in a reading waiting state. Note that the buffer control unit 33 in the read waiting state sequentially reads the records Buf2-1 to 3 stored in the block B2 using the next read permission notification as a trigger.

  In this way, reading and writing of PCM audio data is performed by alternately selecting a buffer plane, that is, a plurality of blocks B1 and B2. By appropriately selecting the sizes of the blocks B1 and B2, it is possible to give a predetermined amount for determining at what point the read permission notification is generated with respect to the amount of the PCM audio data accumulated in the buffer 32. it can.

  As described above, if the operating condition setting value is 16 bits for quantization bit rate, the sampling rate is 16 kHz, and the number of channels is 2 for stereo audio, the bit rate is 512 kbps, and 640 kbytes of PCM audio data every 10 ms. Will occur. Therefore, the size of each of the blocks B1 and B2 can be set to 640 kbytes, and a read permission notification can be generated regularly every 10 ms. The size of each of the blocks B1 and B2 is set by the device management function of the control program 20 when the VoIP telephone application 30 is initially set. Since taking a larger buffer size than necessary causes voice delay, adjustments are made in consideration of the performance of the PCM input device, the performance of the platform on which the VoIP telephone application operates, and the processing of other applications that operate simultaneously. Should be.

  In the example of this figure, a configuration in which three records are included in each of the two blocks B1 and B2 is shown, but the present invention is not limited to such a number. The number of records is appropriately selected from the bit rate of PCM audio data. Moreover, it is also possible to use two or more blocks instead of using the two blocks B1 and B2 alternately.

  FIG. 3 shows an operation procedure in the embodiment of the present invention. The operation procedure will be described with reference to the components shown in FIG. The operation procedure is shown from the point where the IP telephone apparatus is turned on and started to operate.

  First, the PCM input device driver 31 sets operating conditions of the PCM input device (step S1). The operation condition is set based on a command from the device management function of the control program. On the other hand, the buffer control unit 33 reserves an area of a necessary size in the buffer 32 in the MEM (step S2).

Next, the PCM input device driver 31 selects one block among the plurality of blocks (step S3). Next, while acquiring the PCM audio data from the PCM input device, the PCM audio data is stored in each record in parallel, and this is sequentially written into one previously selected block (step S4). At this time, the PCM input device driver 31 determines whether or not one block is finished (step S5). That is, the PCM input device driver 31 is constantly monitoring the buffer accumulation amount of PCM audio data. If one block has not been completed (N), the PCM input device driver 31 continues to write, while if one block has been completed (Y), the writing is interrupted and A block is designated and reading start permission is notified to the buffer control unit 33 (step S6). There are various methods for realizing such notifications, such as using events, messages, or CallBack functions. However, the notification should be selected optimally depending on the specifications of the control program to be implemented and the specifications of the platform that constitutes the IP telephone device. It is.
With this read start permission notification as a trigger, that is, a trigger, the buffer control unit 33 reads the PCM audio data of the block from the buffer 32 for each record (step S7). Next, PCM audio data is extracted from each read record and supplied to the audio codec unit (step S8).

  On the other hand, the PCM input device driver 31 selects another block different from the block selected in step S3 (step S9), resumes writing to the block, and continues the writing operation.

  As a modification of the above-described embodiment, the buffer control method according to the present invention can be applied to a computer apparatus that processes an acoustic signal that is an audio signal in a broad sense. In this case, the audio signal is acquired by the audio device in a certain unit time data amount, converted into PCM audio data, and stored in the buffer. The accumulated PCM voice data is transferred to the acoustic application and used for the processing.

  In the above embodiments and modifications, the use of the IP telephone apparatus and the buffer control method of the present invention enables the processing of PCM audio data at a fixed time interval without being affected by the fluctuation of the processing on the PCM input device side. It becomes possible. Further, even if a deviation occurs between the writing timing and the reading timing of the PCM audio data, the deviation is absorbed by the buffer, and as a result, it is possible to reproduce sound with a good quality sound.

1 is a block diagram illustrating an overall configuration including an IP telephone device according to an embodiment of the present invention. It is explanatory drawing explaining the mode of writing to and reading from the buffer shown in FIG. It is a sequence diagram which shows the operation | movement procedure in the Example of this invention.

Explanation of symbols

10 IP telephone device 11 CPU
12 Memory (MEM)
13 PCM input device 14 Communication device 15 Bus 20 Control program (OS)
30 VoIP telephone application 31 PCM input device driver 32 buffer 33 buffer control unit 34 voice data codec unit 35 communication control unit B1, B2 block

Claims (5)

An IP telephone apparatus including a PCMPCM input device that captures input voice and converts it into PCM voice data, and a voice codec that performs encoding processing for the IP phone on the PCM voice data,
A buffer for storing the PCM audio data;
Monitor the PCM audio data accumulation amount in the buffer, and transfer the at least part of the accumulated PCM audio data from the buffer to the audio codec when the PCM audio data accumulation amount reaches a predetermined amount Transfer means;
An IP telephone apparatus comprising: The buffer monitoring transfer means includes:
A PCM input device driver that writes the PCM audio data to the buffer, monitors the PCM audio data accumulation amount of the PCM audio data, and generates a read permission notice when the PCM audio data accumulation amount reaches a predetermined amount When,
Triggered by the generation of the read permission notification, at least a part of the stored PCM audio data is read from the buffer and transferred to the audio codec.
The IP telephone apparatus according to claim 1, comprising:   2. The IP telephone apparatus according to claim 1, wherein the buffer is divided into a plurality of blocks which are permission units based on the read permission notification. A buffer control method in a computer device that takes in input sound, converts it into PCM sound data, and performs sound processing on the PCM sound data,
An accumulation step of accumulating the PCM audio data in a buffer;
Monitor and transfer step of monitoring the PCM audio data accumulation amount of the buffer and transferring at least a part of the accumulated PCM audio data from the buffer to the audio processing when the PCM data accumulation amount reaches a predetermined amount When,
A buffer control method comprising:   The monitoring transfer step writes the PCM audio data to the buffer and monitors the PCM audio data accumulation amount of the PCM audio data. When the PCM audio data accumulation amount reaches a predetermined amount, a read permission notification is issued. 5. The step of generating and reading at least a part of the stored PCM audio data from the buffer upon generation of the read permission notification and transferring the read data to the audio processing. The buffer control method described.

Images