JP5046661B2 - Audio equipment - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide an "audio device" which interporates data so as not to generate noise in reproduction when data are missed due to an error in transfer by an isochronous protocol. <P>SOLUTION: When detecting a transfer error while another audio device 1 is connected by a USB cable 2 to transfer data by the isochronous protocol, data between the time B at which the code of the data is changed between positive and negative and the time A prior to the time B and at which the code is changed in the same direction in the data before the time P of transfer error are defined as interpolation data S, an interpolation part V is formed by deleting the data between the time B and the time C at which the code is changed first in the data before the error occurrence, and the interpolation data S are appropriately continuously copied in it. After the data, data W after the time C are connected. When the interpolation data S are longer than the interpolation part V, they are compressed so as to be fit in the length for interpolation. <P>COPYRIGHT: (C)2008,JPO&amp;INPIT

Description

  The present invention relates to an audio apparatus capable of capturing and reproducing compressed audio data such as MP3 from the outside, and in particular, at the time of data loss interpolation due to a communication error when performing digital communication of compressed audio data using a USB cable or the like. The present invention relates to an audio apparatus capable of preventing the generation of noise.

  In recent years, portable audio devices that can store a large amount of compressed audio data such as MP3 in a memory, HDD, or the like have become widespread, and a large amount of audio data can be carried and used all the time. When recording audio data in such portable audio equipment, audio data stored in a personal computer or audio data such as a CD is often converted into MP3 or the like.

  On the other hand, although such portable audio devices are convenient to carry, they are often listened with earphones or headphones that do not necessarily have sufficient acoustic effects. Hereinafter, when listening with a “car audio device or the like”, the portable audio device is connected to the car audio device or the like by wire or wirelessly. At this time, in addition to a method for inputting an analog audio signal reproduced by a portable audio device into a car audio device or the like in a car audio device, audio data from the portable audio device is converted into a digital signal as a method for reproducing high-quality music. The method of inputting as is also adopted. In that case, a portable audio device and a car audio device are connected with a USB cable, and digital data is transferred using a USB serial interface.

  When such a transfer is performed, an audio signal is transmitted as a digital signal to a car audio device or the like every 1 ms by a protocol called isochronous, and the number of data transmitted at this time is the audio reproduced by the portable audio device. For example, it is 178 bytes at 44.1 kHz. A car audio device or the like stores a received digital signal in a memory, reads out data from the memory at time intervals corresponding to the sampling frequency, converts it into an analog signal using a DAC (digital-analog converter), and outputs a reproduction output from a speaker. Is going.

  When digital data is transferred through such a USB serial interface, a communication error may occur due to various factors. At that time, for example, when an error occurs when transferring data to store data from a personal computer to a portable audio device, the data in the portion where the error has occurred is retransmitted from the personal computer to the portable audio device. Therefore, even if a communication error occurs, the audio file on the personal computer and the audio file on the portable audio device will be the same.

  However, if an error occurs when audio data is transmitted from a portable audio device to a car audio device or the like, the error portion is not retransmitted. This is because the protocol called isochronous guarantees the reproduction of audio and the like in real time, and therefore does not retransmit data when an error occurs.

  It is to be noted that the error of the reproduced PCM signal is checked, and when this error exceeds a specified value, the block of the signal causing the error is corrected using the block before or after that. A reproduction apparatus is described in Patent Document 1.

Also, if there is a possibility that an error has been added in the transmission path and the error of the encoded data received by the receiver can be corrected, error correction is performed. In the case of encoded data that cannot be corrected, and is combined with the encoded data of a frame that forms a certain section of speech before the encoded data that is currently received, The encoded data before the received frame is guided from the buffer to the audio composite means by the interpolation information by the interpolation information, and the frame of the audio is interpolated, and the encoded information of the continuous frame before and after the encoded data of the currently received frame by the interpolation information A technique for performing interpolation while maintaining continuity by using a waveform of a composite voice of data is described in Patent Document 2.
JP 58-155509 A Japanese Patent Laid-Open No. 4-100422

  As described above, when a transfer error occurs when digital data is transferred between a portable audio device and a car audio device using the USB serial interface using the isochronous protocol, the data contained in that portion is stored. Missing. Here, since the length of the missing data is 1 ms, when this portion is reproduced, the pop sound is heard as noise. In order to reduce this noise, it is conceivable to employ a method of interpolating lost data by connecting the missing data with a straight line. However, if this method is always used when data is missing, the frequency component is large when normal audio signals are output. Therefore, the DC component is superimposed when the front and back of the missing data are connected by a straight line. Due to the relationship with the frequency components before and after, the noise is reduced more than the noise at the time of simple data loss, but the interpolated portion is heard as another noise.

  Therefore, according to the present invention, when digital audio data is connected by a USB serial interface and transferred by an isochronous protocol, if data is lost due to a transfer error, interpolation can be performed so as not to generate noise during reproduction. The main object is to provide an audio device.

  In order to solve the above-described problem, an audio device according to the present invention detects a memory for storing transferred data and an error in the transferred data in an audio device that performs data transfer with other audio devices in an isochronous manner. When a transfer error is detected by the transfer error detection unit and the transfer error detection unit, the data before the transfer error occurrence time, when the data changes between positive and negative, and the data after the transfer error end time A data code change time point detection unit for detecting a first code change time point at which the sign changes in the same direction as the change in the data, and a data point before the transfer error occurrence time and a time point before the data sign change time Interpolating data creating unit that uses the data between the time point when the sign changes in the same direction as the interpolating data, An unnecessary data deletion unit that deletes data between the first time point at which the sign changes in the data before the error occurrence time and the data code change time after the error end time, and creates an interpolation part; and the interpolation When the interpolation data is equal to or shorter than the interpolation portion, the interpolation data created by the interpolation data creation unit is continuously connected within a range in which all of the interpolation data enters after the transfer error occurs, and the data It is characterized by comprising a data connection processing unit for connecting data after the first same direction code change time in the data after the transfer error end time to the connected interpolation data.

  In another audio device according to the present invention, in the audio device, the interpolation data creation unit may include data before the time when the sign of the data first changes in the data before the transfer error occurs. The interpolation data is generated when the time when the code changes in the same direction is within a first predetermined time from the time when the transfer error occurs.

Further, in another audio device according to the present invention, in the audio device, the interpolation data generation unit is the same as immediately before the time when the sign of the data first changes in the data before the transfer error occurrence time. When the time interval between the time when the sign changes in the direction is less than the second predetermined time shorter than the first predetermined time, the time point when the sign changes in the same direction from the previous data is detected in order, The data between the time when the code changes at the end within the first predetermined time from the time when the transfer error occurs and the time when the first code changes is used as interpolation data.

  In the audio device, the length of the interpolation data created by the interpolation data creation unit is longer than the length of the interpolation part created by the unnecessary data deletion unit. Further, the interpolation is performed by compressing the interpolation data so as to be adapted to the interpolation portion.

  Since the present invention is configured as described above, when digital audio data is transferred by an isochronous protocol, such as by connecting with a USB serial interface, noise is not generated during reproduction when data is lost due to a transfer error. Can be interpolated.

  The present invention relates to the problem of performing interpolation so that noise is not generated during reproduction when data is lost due to a transfer error when transferring using an isochronous protocol. In the memory that stores the transferred data, the transfer error detection unit that detects an error of the transferred data, and when the transfer error is detected by the transfer error detection unit, in the data before the transfer error occurrence time, A data code change point detection unit for detecting a first code change point at which the sign changes in the same direction as the change in the data when the data is positive and negative and the sign changes, and the data after the transfer error end point, and a transfer error In the data before the time of occurrence, the data at which the data sign changes first and the data before it An interpolation data creation unit that uses data between the time when the sign changes in the same direction as the interpolation data, the first time when the sign changes in the data before the error occurrence time, and the error end time Created by an unnecessary data deletion unit that deletes data between the subsequent data code change points and creates an interpolation part, and when the interpolation data is equal to or shorter than the interpolation part, the interpolation data generation unit creates The interpolation data is continuously connected within a range in which all of the interpolation data enters after the transfer error occurrence time, and the first same direction code in the data after the transfer error end time is connected to the data connected interpolation data. This is realized by including a data connection processing unit that connects data after the change point.

  The present invention will be described with reference to the drawings. FIG. 1 is a functional block diagram for implementing the present invention. For example, a portable audio device 1 storing digital audio data compressed by MP3 or the like is connected to a vehicle audio device or a general household audio device by a USB cable 2. The example which connects to the audio apparatuses 3, such as these, and outputs from the speaker 4 is shown.

  The illustrated audio apparatus 3 includes a USB controller 5 that can be connected to the portable audio apparatus 1 via a USB cable 2 and can exchange data and monitor received data. There is also provided a transfer error detection unit 6 for detecting whether or not there is a communication error such as missing received data in the digital audio data received serially from the audio apparatus 1 by isochronous transfer.

  The audio device 3 includes a memory 8 that temporarily stores audio data input to the USB controller 5, and audio data can be written to and read from the memory 8 by the memory access unit 9 in the central processing unit 7. . Further, the audio data stored in the memory 8 can be read sequentially or by an arbitrary portion by the memory access unit 9, and the read data can be subjected to various processes as described later in the central processing unit 7. I am doing so.

  In the central processing unit 7, compressed digital audio data such as MP3 stored in the memory 8 is converted into predetermined audio data and output to the audio output signal creation processing unit 19, and the audio output signal creation processing unit 19 performs digital audio data. Is converted into analog audio data, and necessary processing is performed and output from the speaker 4. The central processing unit 7 includes a data interpolation processing unit 10. When the error detection unit 6 of the USB controller 5 detects that an error has occurred in the received audio data as described above, the error part is replaced with other data. Perform interpolation processing.

  In the data interpolation processing unit 10 shown in FIG. 1, a main data interpolation processing unit 11 that performs main data interpolation processing of the present invention, and auxiliary data interpolation that performs auxiliary interpolation processing when appropriate main data interpolation processing is not performed. The auxiliary data interpolation processing unit 12 linearly interpolates the missing data portion. The main data interpolation processing unit 11 is connected to an interpolation data generation unit 13 for generating interpolation data, an unnecessary data deletion unit 16 for deleting unnecessary data in the audio data, and various data in the interpolation process. And a data compression processing unit 18 that performs processing for compressing the interpolation data under specific conditions as will be described later.

  The interpolation data creation unit 13 includes a data sign change point detection unit 14 that detects a point in time when the sign of audio data changes from positive to negative, or from negative to positive, and for creating appropriate interpolation data as described later. And a code change point interval measuring unit 15 for calculating intervals between a plurality of data code change points. Each functional unit as described above is a representative functional unit for operating the present invention and the embodiments thereof, and in addition, functional units that perform various functions can be provided.

  2 and 3 are operation flowcharts showing an example of data interpolation processing implemented in the functional block diagram of FIG. In the example of the data interpolation processing when a transfer error occurs as shown in FIG. 2, the portable audio device 1 and the audio device 3 shown in FIG. 1 are first connected by the USB cable 2 and the audio data is transferred using the isochronous protocol. Is performed (step S1), and the transfer data is temporarily stored in the memory 8 of FIG. 1 on the receiving side (step S2).

  Next, it is determined whether or not a transfer error is detected during the transfer of the audio data (step S3). For example, this check is performed in synchronization with data transfer every 1 ms. For checking the occurrence of transfer error at this time, various methods conventionally used can be used. For example, it is possible to determine the occurrence of transfer error based on the fact that one frame of transfer data is missing. This determination is made by the transfer error detection unit 6 of the USB controller 6 in the example of FIG.

  If no transfer error is detected in step S3, the process returns to step S1 to continue the transfer of audio data, and the above operation is repeated until a transfer error is detected. In contrast, when it is determined in step S3 that a transfer error has been detected, an error occurrence time point P and an error end time point Q are detected (step S4). For example, in the example of FIG. 4A, this operation detects that a transfer error has occurred between the time points P and Q during the data transfer as described above, and the time is U. Will be detected.

  Next, a point B where the sign of the data changes is detected from the data immediately before the error occurs (step S5). In the operation of FIG. 4A, this operation detects that the sign changes from negative to positive at time B immediately before error occurrence time P. In the example of FIG. 1, this operation is performed by the data code change point detection unit 14 of the interpolation data creation unit 13 in the main data interpolation processing unit 11.

  Thereafter, an operation is performed to detect a point A1 that changes in the same direction, that is, in the above case, from negative to positive in the forward direction from the point B when the change of the sign is detected (step S6). With this operation, in the example of FIG. 4A, a time point A1 in which the sign changes in the same direction as the direction in which the sign change point changes from negative to positive is detected in front of the sign change time point B. This operation is also performed by the data code change point detector 14 in FIG.

  Thereafter, it is determined whether or not a time point A1 at which the sign changes is detected from the error occurrence time point P to a first predetermined time interval R such as 3 ms (step S7). In the example of FIG. 4A, an example in which the sign change time A1 is detected within the first predetermined time interval R is shown. Here, as will be described later, when the interpolation data S is obtained, such a determination is made because too long interpolation data is often too long as the interpolation data S in the transfer error portion. This operation is performed by the sign change point interval measuring unit 15 in the interpolation data creating unit 13 of FIG.

  In step S7, when a time point A1 at which the sign changes is found within a first predetermined time R such as 3 ms as shown in FIG. 4A, the process proceeds to step S8, and from the sign change time point A1 to the first sign change time point B. Until the interval T is less than a second predetermined time such as 1 ms (step S8). In the example of FIG. 4A, it is shown that the sign change time A1 exists at a time longer than 1 ms. Here, when the interpolation data S is shorter than 1 ms, an acoustic inconvenience can be caused by a large number of repeated audio signals in a short time when the same interpolation data is embedded in a transfer error period or the like as will be described later. A check will be made to prevent it as much as possible. This operation is also performed by the sign change time interval measuring unit 15 as described above.

  When it is detected in step S8 that the interval T from the sign change time A1 to the sign change time B is less than a second predetermined time such as 1 ms, the process proceeds to step S11, and further within the predetermined time R from the error occurrence time P. A point of change A (2-n) in the same direction of the forward code is detected (step S11). This operation can be performed by the cooperation of the data code change point detection unit 14 and the code change point interval measurement unit 15 of FIG. When it is detected in step S8 that the interval T from the code change time point A1 to the code change time point B is 1 ms or more, and in step S11, the detection operation of the forward code same direction change time point A (2-n) is performed. Then, the process proceeds to step S9, where the foremost code change time point A (1 to n) meeting the above condition is set as the interpolation data start time point A, and as a result, the interpolation data S between the time points A and B is determined. (Step S9).

  In step S7, when the time A1 when the sign changes between the error occurrence time P and the first predetermined time R such as 3 ms is not detected, the normal data portions at the error occurrence time P and the error end time Q are straightened. Is interpolated (step S10). Such linear interpolation is conventionally known as an interpolation process when an error occurs, but in the present invention, as described above, the above-mentioned code is added between the error occurrence point P and a first predetermined time R such as 3 ms. When there is no change, that is, this means that the frequency component of the audio data immediately before the error occurs is low, and even if linear interpolation is performed in such a state, there is no abnormal noise as in the past, This technique can be used effectively. In the example of FIG. 1, this operation is executed by performing linear interpolation of missing data in the auxiliary data interpolation processing unit 12 in the data interpolation processing unit 10.

  After performing the linear interpolation as described above and further obtaining the interpolation data S, the main data interpolation process when a transfer error occurs is performed as shown in FIG. 3 (step S14). In the main data interpolation process when a transfer error occurs in FIG. 3, first, data between the sign change time B immediately before the error occurrence time P and the error occurrence time P as shown in FIG. S21). This operation is performed by the unnecessary data deletion unit 16 in the main data interpolation processing unit 11 of FIG.

  Thereafter, the same direction code change time C after the error end time Q is detected (step S22). Thereafter, the data between the error end point Q and the sign change point C is deleted (step S23). As a result, in the example of FIG. 4, as shown in FIG. 4B, the data between the transfer error end time Q and the code change time C is deleted, and the code change time B and the transfer error occurrence time P As a result, the interpolated portion V between the sign change points B and C is formed.

  Thereafter, it is determined whether or not the length T of the interpolation data S is longer than the interpolation part length V between the code change time B and the code change time C (step S24). As a result of the determination, when the length T of the interpolation data S is equal to or shorter than the interpolation part length V as shown in FIG. 4B, for example, T ≦ V, the interpolation data S is encoded. After the change detection time B, copying is continuously performed within a range in which all of the interpolation data Y enters until the sign change time C. In the example of FIG. 4, as shown in FIG. 4C, it is shown that all two pieces of interpolation data Y using the interpolation data S are copied to the interpolation part length V.

  After performing such interpolation processing, in the present invention, the data after the sign change time C is moved and connected to the last time D of the interpolation data Y (step S26). This operation is performed in the data connection processing unit 17 in the main data interpolation processing unit 11 in the example of FIG. By performing a series of operations as described above, as shown in FIG. 4D, all data are connected at the point where the sign changes in the same direction, and as a result, a smooth data connection is made. When this audio data is reproduced, no noise is generated. As a result of the above-described interpolation processing, the time of X in FIG. 4 (c) is shortened, but this time is extremely short, for example, 1 ms or less, and the user feels uncomfortable during playback. I will not let you.

  On the other hand, when it is detected in step S24 that the length T of the interpolation data S is not equal to or shorter than the interpolation portion length V between the code change time point B and the code change time point C, that is, T> V. The length T of the interpolation data S is compressed by the ratio (V / T) of the interpolation part length V and the interpolation data length T. This operation is shown in FIG. 5. A transfer error occurs in the period U between time points P and Q as shown in FIG. 5A, and a code before time point P as shown in FIG. 5B. When the change point B is detected and the first sign change point C after the time point Q is detected, the data between the time points B and P and the data between the time points Q and C are deleted, and as a result, the interpolation part is When the interpolation data length is longer, the interpolation portion length V is compared with the time T of the interpolation data S created in the same manner as described above. In this way, the interpolation data S is compressed by the equation shown to obtain the interpolation data Y. This operation is performed in the data compression processing unit 18 in the main data interpolation processing unit 11 in FIG.

  The interpolation data Y, which is the compressed data obtained in this way, is inserted into the error part and copied (step S28). As a result, in the example of FIG. 5, the interpolated data Y conforming to the interpolated portion V is inserted and copied, and the entire data is smoothly connected. As a result, no noise is generated when the audio data is reproduced. When the interpolation data S is obtained by compressing the interpolation data S performed at this time, the amplitude is not changed but only the length in the time direction is compressed, in addition to compressing the entire signal in a similar manner. Also good. After performing the above process, as shown as step S15 in FIG. 2, the process returns to step S1 and the above operation is repeated during data transfer.

  The main data interpolation processing unit 11 in the data interpolation processing unit 10 in FIG. 1 performs the processing as shown in FIGS. 4 and 5. In the auxiliary data interpolation processing unit 12, step S10 in the operation flow of FIG. This processing is performed as shown in FIG. That is, in FIG. 6A, after the occurrence of a transfer error is detected in step S4 in FIG. 2 in the period U between time points P and Q, the data sign changes from the data immediately before the error occurs in step S5. Time B is detected, and at step S6, an operation is performed to detect a time A1 at which the sign changes in the same direction forward from the detection time B. In step S7, an error occurs from the error occurrence time P to a first predetermined time R such as 3 ms. When it is determined that a point where the sign changes in between is not detected, the process proceeds to step S10, and the normal data G at the error start time and the normal data H at the error end time are connected by a straight line J as shown in FIG. Interpolate with.

  In the example of FIG. 6A, an example in which the time point B at which the data sign changes from the data immediately before the occurrence of the error in step S5 of FIG. 2 can be detected is shown, for example, in FIG. 6B. Thus, when no sign change time is detected between the data immediately before the occurrence of the error and a predetermined time R such as 3 ms, the time when the sign changes in the same direction forward from the detection time B is detected in step S6. Since it is natural that the time point A1 cannot be detected, the time point A1 is not detected in step S7, and the process proceeds to step S10 to perform linear interpolation processing as shown in FIG.

  The auxiliary data interpolation processing for linearly interpolating the missing data as described above is a conventionally proposed method, but in the present invention, it is performed when the main data interpolation processing as described above is not performed. . That is, in the present invention, the linear interpolation is performed when the sign change point as described above is not found within a predetermined period of 3 ms or the like, and for example, as shown in FIGS. 6 (a) and 6 (b). This means that the frequency component just before is low, so even if linear interpolation is performed during such a period, noise like when performing linear interpolation processing when the frequency component is high is not generated. The method can be used effectively.

  In the above embodiment, the present invention shows a processing example when a portable audio device and a general audio device such as a vehicle audio device are connected by a USB cable and digital data is transferred by an isochronous method. Data transfer between audio devices in a broad sense that can handle audio data including PCs, PDFs, and even mobile phones is not limited to the above, and data is not retransmitted when an error occurs. The present invention can be applied in the same manner when a defect occurs due to the above.

It is a functional block diagram of the Example of this invention. It is an operation | movement flowchart of the data interpolation process in the Example. It is an operation | movement flowchart of the main data interpolation process in the Example. It is a figure which shows the process example of the main data interpolation process of the Example. It is a figure which shows the other process example of the main data interpolation process of the Example. It is a figure which shows the process example of the auxiliary data interpolation process of the Example.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 Portable audio apparatus 2 USB cable 3 Audio apparatus 4 Speaker 5 USB controller 6 Transfer error detection part 7 Central processing unit 8 Memory 9 Memory access part 10 Data interpolation process part 11 Main data interpolation process part 12 Auxiliary data interpolation process part 13 Interpolation data creation unit 14 Data code change point detection unit 15 Code change point interval measurement unit 16 Unnecessary data deletion unit 17 Data connection processing unit 19 Audio output signal creation processing unit

Claims (4)

In audio devices that perform data transfer with other audio devices in an isochronous manner,
A memory for storing the transferred data;
A transfer error detector for detecting errors in the transferred data;
When a transfer error is detected by the transfer error detection unit, in data before the transfer error occurrence time, when the data is positive and negative, the sign changes, and in the data after the transfer error end time, the sign changes. A data code change time point detection unit for detecting a first code change time point changing in the same direction;
Interpolation data creation that uses data between the time when the sign of the data first changes and the time when the sign changes in the same direction in the data before the transfer error occurs as interpolation data And
An unnecessary data deletion unit that creates an interpolation part by deleting data between a first time when a sign changes in data before the error occurrence time and the data sign change time after the error end time;
When the interpolation data is equal to or shorter than the interpolation portion, the interpolation data created by the interpolation data creation unit is continuously connected within a range in which all the interpolation data enters after the transfer error occurs, An audio apparatus comprising: a data connection processing unit that connects data after the first same direction code change time in data after the transfer error end time to the interpolation data connected with the data.   The interpolating data creation unit is configured such that, in the data before the transfer error occurrence time, at least the time when the code changes in the same direction in the data before the time when the data sign changes first from the transfer error occurrence time. The audio apparatus according to claim 1, wherein the interpolation data is created within a first predetermined time. The interpolating data creation unit has a time interval between a time when the data sign first changes in data before the transfer error occurrence time and a time point when the code changes in the same direction immediately before the first predetermined data When it is less than the second predetermined time shorter than the time, the time when the code changes in the same direction from the previous data is sequentially detected, and the code is finally detected within the first predetermined time from the time when the transfer error occurs. 3. The audio apparatus according to claim 2 , wherein the data between the time when the signal changes and the first code change time is used as interpolation data.
  When the length of the interpolation data created by the interpolation data creation unit is longer than the length of the interpolation part created by the unnecessary data deletion unit, the interpolation data is compressed and interpolated to fit the interpolation part. The audio apparatus according to claim 1, wherein:

Images