JP2001209396A - Control method and device for coding sound data, and recording medium with recorded processing program for controlling sound data coding - Google Patents

Abstract

PROBLEM TO BE SOLVED: To make it possible to efficiently decode sound data from a target position when reproducing the data from a position designated by a user for rewinding, fast-forwarding, or the like, in a coding method intended for storing the sound data. SOLUTION: When rewinding is requested from a user, this device sets a presumed rewinding position tx1 based on the request for rewinding and the requested amount; searches for 'ESC' existing in the neighborhood of the presumed rewinding position along a axis time from the presumed position; examines the temporal before-and-after relation between the presumed rewinding position and a target rewinding position t1; when the presumed rewinding position is at a position suitable to the target rewinding position, the device performs null-reproduction by decoding the data from the control data position and then starts actually reproducing at the time of having reached the rewinding position; and the presumed rewinding position is not at a position suitable to the target rewinding position, the device determines another presumed rewinding position again to perform similar processing.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a technique for encoding sound data such as voice and music, and enables efficient rewinding and fast-forwarding operations of mainly stored data. The present invention relates to a sound data encoding control method, a sound data encoding control device, and a recording medium on which a sound data encoding control processing program is recorded.

[0002]

2. Description of the Related Art As an encoding method of sound data, an outline of an encoding method used for transmission system data and a conventional encoding method used for storage system data will be described.

[0003] Research and development and practical application of sound data encoding methods used for transmission system data have been advanced, and there are many very excellent encoding techniques. For example, there is a method of performing encoding using a high correlation between temporally close data. This encoding method is a method that utilizes the characteristic that future data can be efficiently estimated from past data history. More specifically, data immediately following the data history up to a certain point in time is predicted, and only the difference from the prediction is encoded.

[0004] Such a method is generally performed, and as a specific example, the International Telecommunications Union (I
TU). 728 standard and related US patents 5,33
9,384 and others.

[0005] This encoding method has a high encoding efficiency and is widely used. However, since it is a method of predicting subsequent data from previous data and encoding by using a difference between the data, the encoding method at a certain point in time may be considered. If data is lost, there is a problem that an appropriate difference cannot be obtained, and the effect extends to data that is long in the future.

In order to cope with this, in transmission system data, in addition to the original data, measures such as transmission of data for error correction are taken. Although this method has a problem that the entire data amount increases, it can be said that this method is an unavoidable method for performing proper data transmission.

[0007]

It is a natural idea to apply the above-described sound data encoding method used for transmission system data to storage system data. There are different situations and requirements.

That is, in the storage system data, the amount of data after encoding is more important. Also, it is considered that the data error rate is much lower than that of the transmission system. Therefore, the importance of the error correction code for increasing the data amount is relatively lower than that of the transmission system.

In the transmission system, it is required that the decoding process of the coded and transmitted data is performed almost in real time, but the requirement of the real-time processing in the storage system is lower than that in the transmission system. . As a result, at least at the peak time, a larger data amount and calculation amount than the transmission system can be allowed.

In addition, the storage system data requires a rewinding operation or a fast-forward operation, which is almost unnecessary in the transmission system data. This is one of the points that the storage system data is greatly different from the transmission system data. It can be said that there is.

In order to reduce the amount of data after encoding, variable-length encoding according to the situation is effective. That is, it is desirable that a method of discriminating silence, unvoiced sound, voiced sound, and the like, and giving an optimum code to each of them is desirable.

However, according to this method, the length of data changes depending on the situation such as the content of the utterance. At the time of rewinding / fast-forwarding, there is a problem that it is difficult to estimate a target reproduction position.

Further, even if the target position can be estimated, as described above, in the method of simply encoding the difference data, in order to accurately decode the data at the target reproduction position, You need accurate historical data. Therefore, information on data before the target reproduction position is required. In order to obtain this data, it is necessary to perform the decoding process sufficiently before the target reproduction position, which requires a large amount of calculation and time.

Therefore, the present invention mainly aims to enable efficient decoding from a target position when reproducing from a position specified by a user, such as rewinding or fast-forwarding data in storage system data. It is another object of the present invention to improve the tolerance against data errors and data loss in transmission system data.

[0015]

In order to achieve the above object, a sound data encoding control method according to the present invention encodes sound data and decodes the encoded sound data. In the control method, when encoding the sound data, the encoding processing side sets a certain time interval for the sound data to be processed, and at least once in the encoded sound data for each time interval, The control data including the identifier is inserted and the content of the control data is set as an internal history. On the decoding side, when the identifier is found during the decoding process of the encoded data, the content of the control data is stored in the internal history. It is set as data.

Also, the sound data encoding control device of the present invention has a sound data encoding unit having an encoding processing unit for encoding sound data and a decoding processing unit for decoding the encoded sound data. In the control device, the encoding processing unit sets a time interval in the sound data, and generates control data including an identifier at least once within the time interval, and control data generating means for generating the encoded control data including the identifier. Encoding processing means for inserting and outputting the control data generated by the control data generation means into the encoded sound data for each time interval, and output by the encoding means. Encoding data output means for outputting encoded data, wherein the decoding processing unit comprises: encoded data input means for reading the encoded data; and Control data detecting means for detecting whether or not control data is present; decoding processing for decoding the input coded data and setting the contents of the control data detected by the control data detecting means in an internal history And a decryption data output means for outputting data decrypted by the decryption processing means.

Further, the recording medium storing the sound data encoding control processing program of the present invention encodes sound data,
A recording medium on which a sound data encoding control processing program for decoding the encoded sound data is recorded, wherein the sound data encoding control processing program is used to encode sound data when encoding sound data. A certain time interval is set in the sound data to be processed, control data including the identifier is inserted at least once into the coded sound data at each time interval, and the content of the control data is stored in the internal history. On the decoding side, when the identifier is found during the decoding process of the encoded data, the decoding side sets the content of the control data as internal history data.

In each of these inventions, the control data includes first and second control data, and it is determined that the first control data may be processed as silence data in the input sound data. It is generated when a silence data section of a possible length or sound data equivalent thereto appears, and can be distinguished from sound data, and as an identifier indicating the start of control data and information indicating the type of the control data. The second control data is composed of information indicating that silence has continued for a certain length or more, information indicating the length of a silent section, and time information that can specify the location of the control data. In addition, it is generated when it is detected that the silence data section or the sound data equivalent thereto does not appear even after a preset time has elapsed, it is possible to distinguish the sound data from the sound data, and An identifier indicating the start of control data, information indicating that no silent data section having a length equal to or greater than the predetermined threshold or data equivalent thereto did not appear as information indicating the type of the control data, And the time information by which the location of the control data can be specified.

The data capable of reproducing the internal data history incorporated in the second control data is the data which can optimally approximate the internal data history of the encoder at that time or the data of the encoder at that time. Is the internal data history itself.

If the bit pattern per unit length of the identifier and the sound data is the same, a bit pattern in which two bit patterns on either side of the identifier or the sound data are formed is formed. By doing so, the two are distinguished.

Further, when the processing performed on the decoding side is a processing for a rewind / fast-forward request issued by a user, a rewind / fast-forward request issued by the user and the amount of the rewind / fast-forward are requested. An estimated fast-forward position is set, control data existing near the estimated rewind / fast-forward position is searched along the time axis from the estimated rewind / fast-forward position, and the time information included in the searched control data is The temporal relationship between the estimated rewind / fast-forward position and the target rewind / fast-forward position is checked, and if the estimated rewind / fast-forward position is an appropriate position with respect to the target rewind / fast-forward position. After performing the decoding process from the control data position and performing idle playback, actual playback is started when the target rewind / fast-forward position is reached. If the estimated rewind / fast-forward position is not an appropriate position with respect to the target rewind / fast-forward position, the estimated rewind position is set again and the same processing as described above is performed.

As described above, according to the present invention, the control data including the identifier is inserted into the encoded sound data at a certain time interval, and the decoding side performs the decoding process of the encoded sound data during the decoding process. When the identifier is found, the content of the control data is set as internal history data, and decoding from the control data position is enabled based on the internal history data. Accordingly, when data is reproduced from a position specified by the user, such as rewinding or fast-forwarding data, decoding processing can be performed from control data having an identifier near the target position, and decoding can be performed efficiently. It becomes possible.

As the control data, two types of control data (first and second control data) are prepared, and the first control data is a silence data section having a length that can be processed as silence or a silence data section therefor. The second control data is forcibly incorporated when the corresponding sound data appears, and the second control data is forcibly incorporated when such a silent section or the corresponding sound data does not appear even after a set time has elapsed.

As a result, any of the control data is always incorporated into the coded data within a predetermined time, and by searching the coded data along the time axis, the control data can be relatively obtained. The control data can always be found in a short time, and the reproduction process from an arbitrary position can be performed efficiently.

The above-mentioned data which can reproduce the internal data history of the encoder may be data which can optimally approximate the internal data history at that time, or the internal data history of the encoder at that time. The data history itself may be used. When the former internal data history on the encoder side is data that can be optimally approximated, the data amount after encoding can be reduced although some errors occur. On the other hand, if the latter is the internal data history on the encoder side at that point in time, the amount of data after encoding is slightly increased, but since the internal data history on the encoder side is itself, the error should be zero. Can be.

By incorporating such control data into the encoded sound data, for example, in the case of performing a fast rewind operation, it is possible to reproduce the control data located near an arbitrary target position specified by the user. Thus, efficient rewind processing and fast-forward processing can be performed with a small amount of calculation.

It is also conceivable that the identifier indicating the start of the control data and the bit pattern of the sound data may be the same. However, in the present invention, the bit pattern of two consecutive bit patterns of the sound data is considered. By forming a pattern, both can be reliably distinguished. For example, if the encoded data is a byte string, and if the bit pattern of one byte on the sound data side is the same as the bit pattern of one byte of the identifier, the bit pattern on the sound data side is one byte. By providing a rule that each is a continuous two-byte bit pattern, the two can be reliably distinguished.

[0028]

Embodiments of the present invention will be described below. The contents described in this embodiment are as follows.
It is a description of the sound data encoding control method and the sound data encoding control device of the present invention, and specific examples of the sound data encoding control processing program in a recording medium on which the sound data encoding control processing program of the present invention is recorded. It also includes the processing contents.

The present invention mainly aims at realizing an optimum encoding method for accumulated data, and in particular, has a main object of realizing an encoding method capable of efficiently performing a rewind / fast-forward operation. And First, the basic contents of the present invention will be described.

A certain time interval is set for the data to be encoded, and the control data is incorporated at least once in the data at each time interval. At that time, the set time interval can be made different depending on the purpose of use. In addition, the control data to be incorporated includes time information (which may include date information). For example, the time information is such that the data to be processed records some kind of sound information (such as voice or music) in a recording unit. If this is the case, it may be the elapsed time from the recording start time or data indicating the current date and time. These can be mutually converted. In other words, if the date and time of the recording start are known, it is possible to obtain the current date and time from the elapsed time from the recording start time. Conversely, the relationship between the current date and time data and the recording start time data can be obtained. It is possible to get the elapsed time from. The control data includes the first and second control data.
And the first and second control data are respectively incorporated into encoded data (referred to as encoded data) in the following two situations.

One is a case where a sufficiently long silent section or a similar data section (a sufficiently long data section close to almost no sound) appears in the input sound data. Is detected, and the first control data (hereinafter, referred to as silent control data) is incorporated.
Another situation is a case where a silent section does not appear in the set time interval. If no silent section appears in the set time interval, the second situation is forcibly performed. Incorporates control data (this will be referred to herein as synchronous control data).

Here, a silence section having a sufficient length is a silence section having a length that can be determined to be processed as silence data. As a criterion, there is a case in which a silent section length continues to such a degree that it can be determined that all of the silent section data as the internal history required for decoding can be processed as silent data. For example,
If a silent section or a data section equivalent thereto continues for about several hundred msec, the next prediction hardly deviates even if it is assumed that the data on which decoding is based is almost zero. Therefore,
A silent section having such a length or a data section corresponding thereto is referred to as a “sufficiently silent section” here.

As silence control data (first control data) incorporated into the encoded data, in addition to the identifier "ESC" indicating the beginning of the silence control data,
A special code consisting of three elements: a code "SLT" indicating silence, information "DATA" indicating how long the silent section lasts, and an elapsed time "TIME" from the reference point is inserted.

By incorporating such silence control data into the coded data, the coded data has a variable length. However, since the silence data portion can be represented efficiently, the data amount of the coded data can be reduced. it can.

On the other hand, if the next silence section does not appear even if the elapsed time from the last silence control data or the synchronization control data in the processing up to a certain time point is equal to or longer than the set time interval, the synchronization control data ( In addition to the identifier “ESC” indicating the beginning of the second control data), the following 3
Forcibly incorporate synchronization control data consisting of three elements.
The three elements are a code "SYN" indicating that the data is synchronous data, data "DATA" for reproducing the internal data history of the encoder at that time, and an elapsed time "TIME" from the reference point. . The data for reproducing the internal data history of the encoder at that time may be data that can optimally approximate the internal data history of the encoder at that time, as described above. The internal data history of the encoder side at the time may be used.

FIG. 1A shows an example of encoded data into which the above-described control data (silence control data and synchronization control data) is inserted. In FIG. 1, D1 represents silence control data, and D2 represents D2. Represents synchronization control data.
In the silence control data D1, a code "SLT" indicating silence is written after the identifier "ESC" indicating the beginning of the silence control data D1, and thereafter, indicates how many seconds the silence section lasted. The code “DATA” is written, followed by time information “TIME” indicating the time elapsed from the reference point. Note that “TIME” may be the time at that point instead of the elapsed time from the reference point, as described above.

If no silence period appears even after the set time interval, the synchronization control data D2 indicating that the data is synchronous data is incorporated. The synchronous control data D2 is the synchronous control data D2. After the identifier "ESC" indicating the beginning of the data, a code "SYC" indicating that the data is synchronous control data is written, and thereafter, data for reproducing the internal data history of the encoder at that time as synchronous data. “DATA” is written, and then
Time information “TIME” indicating the elapsed time from the reference point is written.

When the internal data of the encoder itself is used as data for reproducing the internal history data of the encoder at that time written as the synchronous data "DATA", the data amount after encoding is It grows but the error is zero. Further, as data for reproducing the internal history data of the encoder at that time written as the synchronization data “DATA”, data approximating the internal data history of the encoder at that time includes a plurality of previously prepared data. From the data, the data that can optimally approximate the internal state of the encoder at that time is selected. In that case,
Although some errors occur, the amount of data after encoding can be reduced.

When decoding coded data incorporating such control data (silence control data or synchronous control data), as shown in FIG. 1B, decoding processing is sequentially performed along the time axis. , The identifier "ESC"
, The decoding process is temporarily stopped, and the process for the control data is performed. That is, if the code following the identifier “ESC” is “SLT”, time information indicating how many seconds the silent section lasts is obtained from the content of “DATA”, and
In the decoder, a silence section (“SILEN
T "), and set it as internal history data.

Then, the decoding process is restarted. After that, when the next identifier "ESC" is found, the decoding process is temporarily stopped again and the process for the control data is performed. That is,
Assuming that the code following the identifier "ESC" is "SYC" (code indicating that the data is synchronous control data), in this case, nothing is reproduced as decoded data, and only the internal state of the decoder is obtained. Set the internal history data to rewrite only the value. That is, "DATA" following "SYN"
Is rewritten with the internal history data of the encoder at that point or the internal history data approximating the same.

The rewind / fast-forward processing for the encoded data in which such control data is written will be described with reference to FIG.

It is assumed that there is coded data as shown in FIG. 2A and a rewind request is given at time t2 when the coded data is being decoded. It is assumed that the actual rewind request point is time t1. In other words, although the rewind request is given by the user at time t2, the actual rewind request point (reproduction start point by rewind) is time t1.

Thus, the rewind request generation time t
As shown in FIG. 2B, a rewind estimation point tx1 is determined for the time being from FIG. 2 and the requested rewind amount. Then, when control data is searched from the rewind estimation point tx1 in the direction of arrow A, "ESC" will be found soon. After that, "S" indicating that the data is synchronous control data
YN ", and time information is written in the subsequent" TIME ". Based on the time information,
It can be determined whether the estimated rewind point tx1 is located before or after the actual rewind request point t1.

In this example, the rewind estimation point tx1 is located after the actual rewind request point t1, and the rewind estimation point must be brought to a position that is further earlier than that. I understand.

Then, as shown in FIG. 2C, the rewind estimation point is set to the position of tx0. Then, when control data is searched from the rewind estimation point tx0 in the direction of arrow A, "ESC" is found soon. After that, "S" indicating silence control data is indicated.
LT ”and time information is written in the subsequent“ TIME ”, so based on the time information,
It can be determined whether the estimated rewind point tx0 is located before or after the actual rewind request point t1.

In this example, the estimated rewind point tx0 is located before the actual required rewind point t1, and if reproduction is performed from the estimated rewind point tx0, the rewind required point t1 will soon be reached. You can see that. As a result, as shown in FIG. 2D, idle reproduction is performed from "ESC" located near the estimated rewind point tx0, and actual reproduction processing is performed when the rewind request point t1 is reached.

By performing the above-described processing, the idle reproduction until the actual reproduction is started is performed at a set time interval, that is,
It is not necessary to perform the time longer than the time Tmax shown in FIG. 2A, and the amount of calculation can be reduced. That is, in general, the reproduction process requires a large amount of calculation. However, according to the present invention, the length of the idle reproduction is limited to within a set time interval, whereby the calculation during the rewind operation is performed. The amount can be significantly reduced.

The rewind process has been described above, but the fast-forward operation can be performed according to a procedure similar thereto. This fast-forward operation will be described with reference to FIG.

The same coded data as shown in FIG.
(a)), when the encoded data is being decoded, it is assumed that a fast-forward request is given at time t3. It is assumed that the target fast-forward request point is time t4.
In other words, a fast-forward request is given by the user at time t3, but the actual fast-forward request point (reproduction start point by fast-forward) is time t4.

Thus, based on the fast-forward request occurrence time t3 and the fast-forward request amount given by the user, FIG.
As shown in (1), a fast-forward estimation point tx2 is determined for the time being. When control data is searched in the direction of arrow A from the estimated fast-forward point tx2, "ESC" is found soon in the case of FIG. After that, there is “SYN” indicating that the data is synchronous control data, and time information is written in “TIME” following the “SYN”.
It can be determined whether 2 is located before or after the actual fast-forward request point t4.

In this example, the fast-forward guess point tx2 is located after the actual fast-forward demand point t4 (a point passing through the fast-forward demand point t4), and the fast-forward guess point is located at a position further earlier than that. It turns out that you have to take it.

Next, as shown in FIG. 3C, the estimated fast-forward position is set to the position of tx3. Then, when the identification information is searched for in the direction of arrow A from the estimated fast-forward point tx3, "ESC" is found soon. And
After that, there is “SLT” indicating a silent section, and time information is written in “TIME” following the “SLT”. Based on the time information, the fast forward estimation point tx3 is set to the actual fast forward. It can be determined whether the position is located before or after the request point t4.

In this example, the estimated fast-forward point tx3 is located before the actual required fast-forward point t4, and if the reproduction is performed from the estimated fast-forward point tx3, it will be understood that the fast-forward estimated point t4 will be reached soon. As a result, FIG.
As shown in (2), idle playback is performed from "ESC" existing near the fast-forward estimation point tx3, and actual playback processing is performed when the fast-forward request point t4 is reached.

By performing the above-described processing, the idle playback until the actual playback is started, as in the rewind operation.
It is not necessary to perform the operation for a time longer than the set time interval (Tmax shown in FIG. 3A), and the amount of calculation can be reduced. That is, the fast-forward playback process is similar to the rewind playback,
Although a large amount of calculation is required, according to the present invention, the length of the idle playback is limited within a set time interval, thereby greatly reducing the amount of calculation during the fast-forward operation. it can.

Next, a specific example of realizing the present invention will be described. The encoded data is regarded as a byte string. All bit patterns may appear in the encoded data.

On the other hand, it is necessary to determine three types of control identification codes of one byte as control data (silence control data and synchronization control data). The three types of control identification codes are a code "ESC" indicating the start of each control data described above, "SLT" indicating a silent section, and "SYN" indicating synchronous data. These are mutually different bit patterns so that they can be identified.

As described above, all the bit patterns may appear in the encoded data, and at least the start of the control data among the bit patterns appearing in the encoded data and each control identification code is indicated. Code "ESC"
It is necessary to be able to distinguish between Therefore, if the same bit pattern as “ESC” appears in the encoded data, as a post-process of the encoding process, two bytes of two consecutive same bit patterns as “ESC” appearing in the encoded data are processed. Replace with the bit pattern of That is,
A bit pattern of “ESC / ESC” is formed, which indicates that the bit pattern of “ESC / ESC” is data that is not control data (called sound data).

At the time of decoding on the decoder side, if a 2-byte bit pattern of "ESC / ESC" exists in the encoded data to be processed, it is converted to "ESC".
The decoding process is performed after replacing the bit pattern with the 1-byte bit pattern of C "and recognizing the sound data.

By providing such a rule, even if data of the same bit pattern as the code "ESC" indicating the start of the control data appears in the encoded data, it is replaced with the code "ESC" indicating the start of the control data. And the processing when the code "ESC" indicating the start of the actual control data appears does not hinder the processing.

In the actual encoding process, when a silent section having a predetermined length is detected during the encoding process, encoded data that has been encoded up to that point is output, and the encoding process is performed once. The measurement is stopped, and measurement of the length of the silent section (elapsed time of the silent section) is started. When the end point of the silent section (the point at which the sound data appears) is detected, the control code “ESC” and “SL” indicating the silent section are used.
T ", the duration" DATA "of the silent section, time information" TIME "indicating the elapsed time from the reference point, etc., and after the silent data is set in the internal history on the encoding side, the encoding process is restarted. The duration of the silent section “DAT
A ", time information" TI indicating the elapsed time from the reference point, etc.
ME "is fixed-length data. These data may be variable-length data. If the data is variable-length data, it is necessary to insert a code indicating the end of the data.

On the other hand, when the decoding side finds 1-byte "ESC" in the encoded data sent from the encoding side, it recognizes that the control data has started, and at that time, the decoding is started. Stop the process and follow the “ESC”
Based on “SLT”, “DATA”, and “TIME”, silence section data for a specified time is output, and then silence history data indicating that all data in a predetermined section is silence is set as an internal state. After that, the decoding process is restarted.

If the elapsed time from the last silent control data or the synchronization control data in the processing up to that time has exceeded the set time, the coded data coded up to that point is output, and the coded data is output. "ESC", which indicates that control data is started, by temporarily stopping the
“SYN” indicating that the data is the synchronization control data, data “DATA” for reproducing the internal data history at that time as the synchronization data, and time information “TIME” indicating the elapsed time from the reference point and the like are written.

As described above, the data for reproducing the internal data history of the encoder at that time as the synchronous data may be the internal data of the encoder itself, or the data at that time may be used. The internal data history of the encoder may be optimally approximated. After setting any of these as the internal state of the encoder, the encoding process is restarted.

On the other hand, if the decoder detects 1-byte "ESC" in the encoded data sent from the encoder, it recognizes that the data is synchronous control data, and performs decoding processing up to that point. Stop once and follow the “SSC” following the “ESC”.
After setting the internal history on the decoding side by “YN”, “DATA”, and “TIME”, the decoding process is restarted.

FIG. 4 shows the coded data and the decoded data according to the above description, which are almost the same as the data in FIG. 1, but as shown in FIG.
For data having the same bit pattern as above, "ESC" of two bytes, that is, "ESC" / "ESC" in which two "ESCs" are consecutive. Therefore, when such a bit pattern appears on the decoder side, the processing is performed by replacing the bit pattern with one "ESC" as shown in FIG.

Next, the configuration of the sound data encoding control device of the present invention will be described. This sound data encoding control device includes an encoding control unit and a decoding control unit. FIG. 5 shows a configuration of the encoding control unit, and FIG. 6 shows a configuration of the decoding control unit.

First, the encoding control unit will be described with reference to FIG. The encoding control unit is roughly divided into audio data input means 1 for inputting audio data to be processed, encoding processing means 2, control data generation means 3, encoded data output means 4 for outputting encoded data, and the like. Is done.

The encoding processing means 2 includes an encoder 21, an internal history data setting unit 22 which can set and output internal history data, and an internal history data encoding unit 23 which encodes the internal history data. , Silence control data (“ESC”, “SLT”, “DATA”, “TIM
E ”) and synchronization control data (“ ESC ”,“ SY
N), "DATA", "TIME") and the like, and a control data addition processing unit 24 for incorporating control data at predetermined positions of encoded data.

The control data generating means 3 includes a silent data detecting section 31 for detecting silent data from the input sound data,
Silence time measuring means 32 for measuring how long the silence data lasts, an elapsed time threshold value setting unit 33 for setting a threshold value for the elapsed time, silence control data for generating and outputting silence control data The output unit 34 generates a synchronous control data output unit 35 for generating and outputting synchronous control data when the next silent period does not appear even after the time set by the elapsed time threshold setting unit 33 has elapsed from the silent period. It has a configuration having.

The silence control data output from the silence control data output means 34 includes the aforementioned "ESC", "S
LT, “DATA”, and “TIME”, and the synchronization control data output from the synchronization control data output means 35 includes the aforementioned “ESC”, “SYN”, “DATA”,
It consists of "TIME".

The silence control data is generated when silence data of a predetermined length appears, and the synchronization control data is generated when the silence control data appearing before the present time or the elapsed time from the synchronization control data exceeds a set time interval. (Equal to or greater than the threshold value set in the elapsed time threshold value setting means 33). Since the data contents of the silence control data and the synchronization control data have already been described, the description thereof is omitted here.

The silence control data output from the silence control data output unit 34 or the synchronization control data output from the synchronization control data output unit 35 is output from the encoder 21 by the control data addition processing unit 24. Embedded in the encoded data. In this way, encoded data as shown in FIG. 4 is created, and the encoded data is output from the encoded data output unit 4 and supplied to the decoding control unit shown in FIG.

If no silence control data or synchronization control data has been generated, the control data addition processing unit 24 controls to pass the sound data encoded by the encoder 21 as it is, Alternatively, when the synchronization control data is generated, a process of discriminating the bit pattern (1 byte) of the identifier “ESC” included in the control data from the bit pattern of each byte of the sound data is also performed. That is, as described above, when the bit pattern of the sound data is the same as the bit pattern of “ESC”, two of the bit patterns are successively set to “ESC · ESC”.
Also, a process of replacing the bit pattern with the bit pattern is performed.

On the other hand, as shown in FIG. 6, when roughly divided, the decoding control section includes a coded data input means 5 for inputting coded data to be processed, a decoding processing means 6, a control data detecting means 7, It comprises a decoded data output means 8 for outputting the decoded data.

The decryption processing means 6 includes a data control unit 61,
It has a decoder 62, an internal history data setting unit 63, an internal history data decoding unit 64, and the like. Data control unit 61
Checks the contents of the coded data input to the coded data input means 5 and controls whether to send the coded data to the decoder 62 or to the control data detecting means 7 according to the contents. That is, if it is normal sound data that is not control data, it is passed to the decoder 62 side, and if it is found that it is control data (if "ESC" is found), it is passed to the control data detecting means 7. When two “ESCs” are consecutive such as “ESC · ESC”, they are combined into one “ESC”.
C ", and is recognized as normal sound data and passed to the decoder 62.

The control data detecting means 7 detects "SLT" indicating silence control data, and if it is determined that the data is silence control data, the control data detecting means 7 performs silence control based on "DATA" contained in the silence control data. The silence control data detecting means 71 which generates and outputs data, detects "SYN" indicating that it is synchronous control data, and if it is determined that the data is synchronous control data, the internal history data included in the synchronous data "DATA" Is output to the internal history data decoding unit 64.

If the control data is synchronous control data, the internal history data included in the synchronous control data (data capable of reproducing the internal history data of the encoder 21 at that time) is decoded by the internal history data decoding. Conversion means 64
After that, it is set as an internal history by the internal history data setting means 63.

The silence data detected by the silence control data detection section 71 is set as an internal history by the internal history data setting means 63 and sent to the decoded data output means 8.

FIG. 7 is a flow chart for explaining the procedure of encoding the encoded data performed on the non-encoding side. Although the outline of the processing procedure has already been described, it will be described again here with reference to the configuration diagram of FIG. 5 and the flowchart of FIG.

In FIG. 7, first, it is determined whether or not a preset time has elapsed (step s1).
It is determined whether or not a silent section has appeared within the time set by the elapsed time threshold setting section 33. If a predetermined silent section has not appeared even after the set time has elapsed, the synchronization data is forcibly transmitted. That is, it performs a process of creating.
Therefore, in step s1, if the set time has elapsed, the process proceeds to step s17, and if not, the process proceeds to step s2.

If it is determined that the set time has not elapsed, the input data is read (step s2), and it is checked whether or not the data is silent data (step s3). If the data is voiced data, the data is encoded by the encoder 21 (step s4), and the bit pattern of the encoded data is "ESC".
It is checked whether the result is the same as (step s5). If the bit pattern of the encoded data is different from "ESC", the encoded data is output (step s6), and step s1
Return to On the other hand, if the bit pattern of the encoded data is “ES
If it is the same as "C", data of the same bit pattern ("ESC") is created (step s7),
It is output (step s6). In other words, in this case, two “ESCs” are output as “ESC · ESC”. Then, the process returns to step s1.

Incidentally, in the above-mentioned step s3,
If it is determined that the data is silent data, the silent time measuring unit 32
Then, it is determined whether or not the silence data has passed a predetermined time (step s8).

The determination as to whether or not the silence data has passed for a predetermined period of time refers to the extent to which the silence data has continued (the length of the silence data), and may be processed as silence data. Performed to determine whether or not. As described above, as a criterion, silence data is treated as a silence data when the length of the silence section continues to such a degree that it can be determined that all data in the silence section may be processed as silence data as the internal history. Therefore, a short silence section existing between speech sections and speech sections,
That is, a silent section for a short time in which the influence of the previous sound section remains remains is not regarded as the silent data. In the case of such a short silent section, encoding processing is performed in step s4 as ordinary sound data.

On the other hand, if it is determined that the processing can be performed as silence data, the encoding process of the encoder 21 is temporarily stopped (step 9), and a code indicating the start of silence data ("" SLT ") and time information (including date information) from a reference point or the like are output (step s1).
0), it is determined whether or not subsequent data is silent data (steps s11 and s12). When data that is not silent data (voiced data) appears, the elapsed time of the silent data up to that time is output (step 13), and the internal history of the encoding processing means 2 is set to the silent data (step s14). , The encoder 21 is restarted (step s1)
4).

Then, the data which has become a sound after the end of the silent data section is sent to the encoder 21 and the process of step 4 is started.

By the way, if it is determined in step s1 that the preset time has elapsed, that is, if a predetermined silent section has not appeared even after the preset time has elapsed, the synchronization control data is forcibly transmitted. The process for inserting is started. In this case, first, the encoding process of the encoder 21 is temporarily interrupted (step 17), and a synchronization code (“SYN”) indicating synchronization control data and time information (including date information) from a reference point or the like are included. ) Is output. Then, the internal history data set in the internal history data setting unit 22 is output (step s18), and the internal history data is set as “DATA” of the synchronization control data (step s1).
9) Restart the encoder 21 (step s21),
It returns to step s1.

The above is the procedure of the encoding process performed on the encoding processing unit side. Next, a procedure for decoding encoded data performed by the decoding processing unit will be described with reference to the flowchart of FIG. 8 and the configuration diagram of FIG.

First, the encoded data input means 5 inputs the encoded data sent from the encoding processing section, and passes it to the data control section 61. The data control unit 61 sequentially reads the encoded data (step s31), and checks whether or not the currently read encoded data includes “ESC” indicating control data (step s3).
2) If there is no "ESC", it is determined that the data is sound data, and the encoded data is passed to the decoder 62. As a result, the decoder 62 performs a decoding process on the encoded data, outputs the encoded data to the decoded data output means 8 (step s33), and returns to step s31. Read.

When the data control unit 61 determines that the encoded data read at that time is "ESC", the subsequent data is read (step s34), and the read data is "ESC". Is checked (step s35). If the data is “ESC”, two “ESCs” will follow. In that case, it is determined that the sound data is general sound data, and 1
Is replaced with two "ESCs" and passed to the decoder 62, and the same decoding process as described above is performed and output (step s).
33) Then, the process returns to step s31.

On the other hand, in step s35, “ES
If it is not "C", that is, if two "ESCs" are not consecutive, it is determined that the data is control data, and the subsequent data is checked to determine whether it is silent control data (step s36). If the data indicates silence control data ("SLT"), it is determined that the data is silence control data, and the data control unit 61 temporarily stops the decoding process of the decoder 62 (step s37). The silence control data is passed to the silence control data detection unit 71. The silence control data detection unit 71 outputs the silence data to the decoded data output unit 8 for the time specified by the content of "DATA" (step s38). The silence data is also given to the internal history data setting unit 63, and the internal history data setting unit 63 sets the internal history to silence data (step s39). Te, then, restart the decoder 62 (step s40), the flow returns to step s31.

If it is determined in step s36 that the data is not silent control data, it is checked whether the data is synchronous control data (step s41), and data indicating that the data is synchronous control data ("SYN"). If ")" is found, it is determined that the data is synchronous control data, and the
Is temporarily stopped (step s42), the internal history data specified by the content of "DATA" included in the synchronization control data is decoded by the internal history data decoding unit 64, and then decoded. Set as an internal history by the setting means 63 (step s4)
3). After that, the decoder 62 is restarted (step s40), and the process returns to step s31. If it is determined in step s41 that the data is not synchronous control data, it is processed as an error (step s44).

FIG. 9 is a flowchart for explaining the processing procedure of the rewinding process performed by the decoding processing unit. This has already been described, but will be described again with reference to this flowchart.

First, when a rewind request is issued by the user, the user moves to an estimated rewind position based on the requested amount (step s51). Then, the encoded data at that position is read (step s52), and it is checked whether the read encoded data is "ESC" (step s52).
53). If it is not “ESC”, reading of the encoded data is repeated until “ESC” appears. This is shown in FIG.
This is the processing described in (b). When the data reaches the "ESC", the data following the "ESC" is read (step s54), and the read data is stored in the "ESC".
It is checked whether it is "C" (step s55).
In the case of, two "ESCs" are consecutive and are not control data, so the flow returns to step s52 again to perform the same processing. Then, the 1-byte "ES
If "C" is found, it is determined that the data is control data, and time information (elapsed time from a reference point or the like) is read from the content of "DATA" (step s56). It is checked whether or not it is located behind the rewind position (step s57). If it is located behind the target rewind position (rewind required point shown in FIG. 2) (FIG. 2 (b) State)), the rewind position is estimated again (step s58), and step s5
2, the same process is performed.

Then, the processing in step s57 is
If it is determined that it is not behind the target rewind position, that is, it is determined that it is located before the target rewind position (the state shown in FIG. The idle reproduction is started from "ESC" (step s59).
Then, it is determined whether or not the idle reproduction has reached the target rewind position (step s60). If the target has been reached, the reproduction is started from the target rewind position. (Step s61).

Although the above is a flowchart for explaining the rewinding processing procedure, the fast-forward operation can be performed in accordance with this, so that the flowchart showing the processing procedure in the fast-forward operation and its description are omitted here.

The present invention is not limited to the embodiment described above, but can be variously modified without departing from the gist of the present invention. For example, in the above-described embodiment, a case has been described in which the present invention is applied to accumulated data, but the present invention is also applicable to transmission data.

Further, the sound data encoding control processing program for performing the processing of the present invention described above can be recorded on a recording medium such as a floppy disk, an optical disk, or a hard disk. Including. Further, the processing program may be obtained from a network.

[0098]

As described above, according to the present invention, coded control data including an identifier is inserted into coded sound data at a certain time interval, and the coded sound data is inserted on the decoding side. When the identifier is found during the decoding process of the sound data, the content of the control data is set as internal history data, and decoding can be performed based on the internal history data.

Thus, when data is reproduced from a position specified by the user, such as rewinding or fast-forwarding data, decoding processing can be performed from control data having an identifier existing near the target position. It is possible to perform efficient rewinding processing and fast-forwarding processing with an amount of calculation. Also,
When the present invention is applied to a transmission system, even if part of data is lost during transmission, normal reproduction can be performed from the point where the next control data is found, so that damage to the entire transmitted data is minimized. Can be minimized.

[Brief description of the drawings]

FIG. 1 is a diagram showing an example in which control data according to the present invention is incorporated into encoded data.

FIG. 2 is a diagram illustrating an operation when a rewind operation is performed using control data according to the present invention.

FIG. 3 is a diagram illustrating an operation when a fast-forward operation is performed using control data according to the present invention.

FIG. 4 is a diagram showing an example in which control data according to the present invention is incorporated into encoded data and a process for discriminating between control data and sound data is performed.

FIG. 5 is a diagram showing a configuration of an encoding processing unit side in the sound data encoding control device of the present invention.

FIG. 6 is a diagram showing a configuration of a decoding processing unit side in the sound data encoding control device of the present invention.

FIG. 7 is a flowchart illustrating an encoding processing procedure on the encoding processing unit side in the sound data encoding control device of the present invention.

FIG. 8 is a flowchart illustrating a decoding processing procedure on the decoding processing unit side in the sound data encoding control device of the present invention.

FIG. 9 is a flowchart illustrating a rewinding procedure on the decoding control device side in the sound data encoding control device of the present invention.

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 1 Sound data input means 2 Encoding processing means 3 Control data generation means 4 Encoded data output means 5 Encoded data input means 6 Decoding processing means 7 Control data detection means 8 Decoded data output means 21 Encoder 22 Internal history Data setting unit 23 Internal history data encoding unit 24 Control data addition processing unit 31 Silence data detection unit 32 Silence time measurement unit 33 Elapsed time threshold setting unit 34 Silence control data output unit 35 Synchronization control data output unit 61 Data control unit 62 decoder 63 internal history data setting unit 64 internal history data decoding unit 71 silence control data detection unit 72 synchronization control data detection unit D1 silence control data D2 synchronization control data

Claims (15)

[Claims] 1. A sound data encoding control method for encoding sound data and decoding the encoded sound data, wherein the encoding processing side encodes the sound data to be processed when encoding the sound data. A certain time interval is set, control data including an identifier is inserted at least once into the encoded sound data at each time interval, and the content of the control data is set as an internal history, and the decoding process is performed. On the side, when the identifier is found during the decoding process of the encoded data,
A sound data encoding control method, wherein the content of the control data is set as internal history data. 2. The control data includes first and second control data, and the first control data has such a length that it can be determined that the input sound data can be processed as silent data. Is generated when non-speech data section or non-speech data corresponding to the non-speech data section appears, and can be distinguished from sound data, and silence is constant as information indicating the start of control data and information indicating the type of the control data. The second control data includes information indicating that the process has continued as described above, information indicating the length of the silent section, and time information that can specify the location of the control data. Generated when it is detected that the silence data section or the sound data equivalent thereto does not appear even after the set time has elapsed, it is possible to distinguish the sound data from the sound data, and to start control data. And an identifier indicating that no silence data section having a length equal to or greater than the predetermined threshold or data corresponding thereto has not appeared as information indicating the type of the control data, and the internal history data at that time is reproduced. 2. The sound data encoding control method according to claim 1, wherein the sound data encoding control method comprises possible data and time information which can specify a location of the control data. 3. The data capable of reproducing the internal data history incorporated in the second control data is data that can optimally approximate the internal data history of the encoder at that time or the data of the encoder at that time. The sound data encoding control method according to claim 2, wherein the internal data history itself is used. 4. When the bit pattern per unit length of the identifier and the sound data is the same,
4. The sound data encoding according to claim 1, wherein a bit pattern in which two bit patterns on either side of the identifier or the sound data are formed in succession is distinguished from each other. Control method. 5. When the process performed on the decoding side is a process for a rewind / fast-forward request issued by a user, a rewind / fast-forward request issued by the user and the amount of the rewind / fast-forward are requested. Set the fast forward estimation position,
Control data existing near the estimated rewind / fast-forward position is searched along the time axis from the estimated rewind / fast-forward position, and the estimated rewind / fast-forward position is determined based on the time information included in the searched control data. The temporal sequence relationship between the target rewind / fast-forward position and the target rewind / fast-forward position is examined. If the estimated rewind / fast-forward position is appropriate to the target rewind / fast-forward position, decoding is performed from the control data position. After performing the process and performing idle playback, actual playback is started when the target rewind / fast-forward position is reached, and the estimated rewind / fast-forward position is appropriate for the target rewind / fast-forward position. 5. The sound data encoding control method according to claim 1, wherein if the position is not the position, the rewind estimation position is set again and the same processing is performed. . 6. An encoding processing unit for encoding sound data,
A sound data encoding control device having a decoding processing unit for decoding the encoded sound data, wherein the encoding processing unit sets a time interval in the sound data, and sets at least one time interval within the time interval. Times, control data generating means for generating encoded control data including the identifier, and sound data, and the control data generated by the control data generating means, the control data generated by the control data generating means, Encoding processing means for inserting and outputting the data in the data, and encoded data output means for outputting the encoded data output by the encoding means, the decoding processing unit, the decoding processing unit, Encoded data input means for reading, control data detecting means for detecting whether or not the control data is present in the input encoded data, and decoding of the input encoded data. Decoding processing means for setting the contents of the control data detected by the control data detection means in the internal history, and decoding data output means for outputting the data decoded by the decoding processing means. A sound data encoding control device comprising: 7. The control data includes first and second control data, and the first control data has such a length that it can be determined that the input sound data can be processed as silent data. Is generated when non-speech data section or non-speech data corresponding to the non-speech data section appears, and can be distinguished from sound data, and silence is constant as information indicating the start of control data and information indicating the type of the control data. The second control data includes information indicating that the process has continued as described above, information indicating the length of the silent section, and time information that can specify the location of the control data. Generated when it is detected that the silence data section or the sound data equivalent thereto does not appear even after the set time has elapsed, it is possible to distinguish the sound data from the sound data, and to start control data. Information indicating that no silence data section having a length equal to or greater than the predetermined threshold or data equivalent thereto has not appeared as information indicating the type of the control data, and the internal history of the encoder at that time. 7. The sound data encoding control device according to claim 6, wherein the sound data encoding control device is constituted by data capable of reproducing data and time information capable of specifying a position of the control data. 8. The data which can reproduce the internal data history incorporated in the second control data is the data which can optimally approximate the internal data history of the encoder at that time or the data of the encoder at that time. The sound data encoding control device according to claim 7, wherein the internal data history itself is used. 9. When the bit pattern per unit length of the identifier and the sound data is the same,
9. The sound data encoding according to claim 6, wherein a bit pattern in which two bit patterns on either side of the identifier or the sound data are made continuous is distinguished from each other. Control device. 10. When the process performed on the decoding side is a process for a rewind / fast-forward request issued from a user, a rewind / fast-forward request issued from the user and the amount of the rewind / fast-forward are requested. Set the fast forward estimation position,
Control data existing near the estimated rewind / fast-forward position is searched from the estimated rewind / fast-forward position along the time axis, and the estimated rewind / fast-forward position is determined based on the time information included in the searched control data. The temporal sequence relationship between the target rewind / fast-forward position and the target rewind / fast-forward position is examined. If the estimated rewind / fast-forward position is appropriate to the target rewind / fast-forward position, decoding is performed from the control data position. After performing the process and performing the idle playback, the actual playback is started when the target rewind / fast-forward position is reached, and the estimated rewind / fast-forward position is appropriate for the target rewind / fast-forward position. 10. The sound data encoding control device according to claim 6, wherein if it is not the position, the rewind estimation position is set again and the same processing is performed. . 11. A recording medium recording a sound data encoding control processing program for encoding sound data and decoding the encoded sound data, wherein the sound data encoding control processing program comprises: On the processing side, when encoding the sound data, a certain time interval is set for the sound data to be processed, and the control data including the identifier is included at least once in the encoded sound data for each time interval. A procedure for inserting and setting the content of the control data as internal history, and a procedure for setting the content of the control data as internal history data when the decoding side finds the identifier during the decoding process of the encoded data. A recording medium having recorded thereon a sound data encoding control processing program. 12. The control data includes first and second control data, and the first control data has such a length that it can be determined that the input sound data can be processed as silent data. Is generated when non-speech data section or non-speech data corresponding to the non-speech data section appears, and can be distinguished from sound data, and silence is constant as information indicating the start of control data and information indicating the type of the control data. The second control data includes information indicating that the process has continued as described above, information indicating the length of the silent section, and time information that can specify the location of the control data. Generated when it is detected that the silence data section or the sound data equivalent thereto has not appeared even after the set time has elapsed, the sound data can be distinguished from the sound data, and the start of the control data is started. An identifier to be tasted, information indicating that no silence data section having a length equal to or greater than the predetermined threshold or data equivalent thereto has not appeared as information indicating the type of the control data, and the inside of the encoder at that time. 12. The recording medium according to claim 11, wherein the recording medium is composed of data capable of reproducing history data and time information for specifying the position of the control data. 13. The data which can reproduce the internal data history incorporated in the second control data is the data which can optimally approximate the internal data history of the encoder at that time or the data of the encoder at that time. 13. The recording medium storing the sound data encoding control processing program according to claim 12, wherein the internal data history itself is used. 14. When the bit pattern per unit length of the identifier and the sound data is the same, a bit pattern is formed by continuing two bit patterns on either side of the identifier or the sound data. Wherein the two are distinguished from each other.
3. A recording medium storing the sound data encoding control processing program according to any one of 3. 15. When the processing performed on the decoding side is a processing for a rewind / fast-forward request issued by a user, a rewind / fast-forward request issued by the user and a requested amount of rewind / fast-forward are performed. Set the fast forward estimation position,
Control data existing near the estimated rewind / fast-forward position is searched from the estimated rewind / fast-forward position along the time axis, and the estimated rewind / fast-forward position is determined based on the time information included in the searched control data. The temporal sequence relationship between the target rewind / fast-forward position and the target rewind / fast-forward position is examined. If the estimated rewind / fast-forward position is appropriate to the target rewind / fast-forward position, decoding is performed from the control data position. After performing the process and performing the idle playback, the actual playback is started when the target rewind / fast-forward position is reached, and the estimated rewind / fast-forward position is appropriate for the target rewind / fast-forward position. 15. The sound data encoding control according to claim 11, wherein if the position is not the position, the rewind estimation position is set again and the same processing is performed. A recording medium on which a processing program is recorded.