JP3827767B2 - Data converter - Google Patents

Description

【００５２】
上述のように、実際の圧縮デジタルデータ中に含まれているフレームは全てＦrameＢであるから、この場合は、圧縮デジタルデータ中に含まれる全フレームの４／７を、ＦrameＢ（パディング有りのフレーム）からＦrameＡ（パディング無しのフレーム）に変換するようにデータを加工することにより、圧縮デジタルデータのデコーダ２への入力速度とデコーダ２での処理速度との整合を図り、デコーダ２が正常に動作するようにすることができる。
【００５３】
なお、このようにパディングを変更したことによる副作用についても同時に対応する必要がある。具体的には、ＭＰＥＧオーディオの場合、圧縮デジタルデータのビットストリームのヘッダ中に含まれているＣＲＣ（Cyclic Redundancy Check ）符号によりデータエラーの判別を行うプロテクションProtectionがイネーブルのときには、ＣＲＣ符号値の変更を行う。また、実際のフレーム長が変わることに対して、不要なデータの削除や空データの追加等を行う。
【００５４】
以上のように、本実施形態によれば、圧縮デジタルデータのビットストリーム中に含まれているパディングの実際の比率と規格上の比率とが異なってしまった場合でも、圧縮デジタルデータのデコーダ２への入力速度と上記デコーダ２の処理速度とが常に一致するようにすることができ、デコード動作に異常が生じないようにすることができる。
【００５５】
上述した実施形態は、データ変換装置１の外部から入力される圧縮デジタルデータの入力速度と原周波数との何れにおいても誤りがない（少なくとも両者の周波数比に誤差がない）場合の例であった。しかしながら、この周波数比に関しても、正常な比率から逸脱することがあり得る。
【００５６】
例えば、圧縮デジタルデータの転送ビットレートＢＲと原周波数ＣＬＫｍとの正常な比率は、
ＢＲ／ＣＬＫｍ＝２２４ＫＨｚ／１３．５ＭＨｚ
であるのに対し、規格と異なるビットレートＢＲ′（例えば２１９Ｋｂｐｓ）で圧縮デジタルデータが転送されたり、規格と異なる原周波数ＣＬＫｍ′（例えば１３．２ＭＨｚ）でクロックが入力されたりする場合は、周波数の比率は、
ＢＲ′／ＣＬＫｍ′＝２１９ＫＨｚ／１３．２ＭＨｚ
となる。
【００５７】
このように、圧縮デジタルデータの転送ビットレート自体にエラーが発生したり、あるいは外部入力のクロックの原周波数自体にエラーが発生した場合でも、上記したパディングの含有率についてエラーが発生した場合と同様に、誤差の蓄積によりデコーダ２が異常な動作をしてしまうことがある。そこで、本実施形態では、このような場合にも対応できるようにするために、図１に示すように、比較手段６を設けている。
【００５８】
比較手段６は、エラーが発生していない場合の周波数比とエラーが発生している場合の周波数比とを比較し、その誤差を検出する。すなわち、比較手段６は、正常なビットレートＢＲと誤りのビットレートＢＲ′との絶対的なずれや、正常な原周波数ＣＬＫｍと誤りの原周波数ＣＬＫｍ′との絶対的なずれは検出できないが、正常な周波数比と誤りの周波数比とのずれは検出することができる。
【００５９】
この場合、制御手段４は、図２に示すように、比較手段６における上述のような比較結果を比率エラー判断手段４５で受けて、周波数比に関するエラーが発生したかどうかを判定する。周波数比エラーの有無の判定方法としては、上述のパディング含有率エラーの有無の判定方法と同様に、一定時間ごとに検出される誤差の和がある一定値を越えた場合にエラーと判定するようにすれば良い。
【００６０】
そして、上述のような周波数比エラーが発生したと判定した場合は、比率エラー判断手段４５は、次のような式(8) に従ってエラー率ErrFreq を算出する。

そして、その算出したエラー率ErrFreq を制御信号生成手段４４に供給する。
【００６１】
制御信号生成手段４４は、供給されたエラー率ErrFreq に基づいて、ＦrameＡとＦrameＢとの配合比を変更するための制御信号を生成し（式(1)(2)に従ってＦrameＢが含まれる比率ｎの値を計算し）、それをデータ加工手段５に供給する。なお、本例のように周波数比エラーが発生した場合は、パディングの含有率自体には誤りがないので、式(1) 中に示した（入力されたデータの配合比によるフレーム長）は、ＦlengthＮである。
【００６２】
したがって、この場合は、データ加工後に圧縮デジタルデータ中にＦrameＢが含まれるべき比率ｎは、次の式(9)(10) のように計算される。

【００６３】
一方、データ加工前の圧縮デジタルデータ中に含まれているＦrameＢの比率は0.42857(＝３／７）であるから、この場合は、圧縮デジタルデータ中に含まれる全フレームの0.07427(＝0.50284-0.42857)を、ＦrameＡ（パディング無しのフレーム）からＦrameＢ（パディング有りのフレーム）に変換するようにデータを加工することにより、上述した不具合を回避することができる。
【００６４】
以上に述べた例の場合は、エラー率ErrFrame，ErrFreq が何れも１．０に近い値であるために不具合を回避することができたが、発生したエラーのエラー率が１．０から大きく離れている場合は、不具合を回避できない場合がある。この場合の対処方法としては、次の２通りが考えられる。１つは、制御可能な範囲を逸脱したので音の再生を停止させるという方法であり、もう１つは、正常な音ではないにしても取り敢えず何らかの音を出しておくという方法である。
【００６５】
前者の音の再生を停止させるという方法は、制御手段４がデコーダ２に対して再生停止信号を出力することで実現できる。または、無音データの入っているＭＰＥＧオーディオのデータフレームを、再生すべきデータフレームの代わりにデコーダ２に送出するようにすることによっても実現できる。
【００６６】
ここで、音の再生を停止させる場合に注意すべき点は、音の再生停止時間をある程度長くする（秒単位の時間）べき点である。その理由は、音の再生停止時間が極めて短いと、音の再生と停止とが短時間で繰り返され、スピーカ（図示せず）から出力される音が“マシンガン”のような音として聞こえるようになってしまう。これでは非常に耳障りであるとともに、スピーカ自身への悪影響も考えられるので、それを抑制する必要があるからである。
【００６７】
一方、後者の方法のように音の再生を停止させない場合には、ＭＰＥＧオーディオでは、あるデータフレームそのものを複製して再生したり、削除したりすることにより、デコーダ２から出力されるデータは本来のものと変わるものの、エラーが発生する場合に比べて音質への影響を抑えることが可能となる。
【００６８】
さらに、本実施形態においては、図１および図２に示すように、デコーダ２の動作基準クロックを生成する元クロックの原周波数を、データ変換装置１の外部制御により（例えば人手により）制御手段４に対して設定するための手段も設けている。
【００６９】
これにより、例えば、ＭＰＥＧオーディオ以外の圧縮符号化方式に準拠した原周波数を設定することにより、様々な圧縮符号化方式に対応してエラー検出を行い、その圧縮符号化方式の下でデコーダ２が誤りなく動作するようにフレーム長制御コード（パディング）を加工するようにすることができる。つまり、１つの制御手段４を複数の目的に適合させて使用することが容易にできる。
【００７０】
この場合、外部制御により設定された原周波数は、図２に示した正常周波数比率算出手段４６に与えられる。正常周波数比率算出手段４６は、この設定された原周波数と、解析器３により解析された規格上の転送ビットレート（該当する圧縮符号化方式による圧縮デジタルデータのビットレート）とから、その圧縮符号化方式においてエラーが発生していない場合の正常な周波数比を算出する。
【００７１】
そして、比率エラー判断手段４５は、上記正常周波数比率算出手段４６における算出結果に基づいて、上述の式(8) と同様にしてエラー率ErrFreq を算出し、それを制御信号生成手段４４に供給する。制御信号生成手段４４は、供給されたエラー率ErrFreq に基づいて、ＦrameＡ，ＦrameＢの配合比を変更するための制御信号を生成し、それをデータ加工手段５に供給する。
【００７２】
なお、以上に説明した実施形態は、図１に示したように、パディング含有率の誤りを見てフレーム長制御コードの含有率を変更するための手段、圧縮デジタルデータの転送ビットレートと原周波数との比率の誤りを見てフレーム長制御コードの含有率を変更するための手段を全て含んだものであったが、それぞれを独立してあるいは任意の組み合わせで適用することが可能である。
【００７３】
また、圧縮符号化方式としてＭＰＥＧオーディオのレイヤIIを例に挙げたが、本発明はこれに限られない。例えば、画像圧縮のためのＭＰＥＧ１あるいはＭＰＥＧ２、ＡＣ−３等の圧縮符号化方式ように、圧縮デジタルデータのビットストリーム中にフレーム長制御コードが少なくとも含まれているものであれば何れも適用することができる。
【００７４】
【発明の効果】
本発明は上述したように、圧縮デジタルデータ中に含まれているフレーム長制御コードの含有率を解析し、その解析した実際の含有率と規格上の含有率とを比較することによって両者の誤差を検出し、その誤差をなくすように上記フレーム長制御コードの含有率を変更すべくデータを加工するようにしたので、圧縮デジタルデータ中に含まれているフレーム長制御コードの含有率が規格上の含有率と異なってしまった場合でも、含有率が正しくなるように圧縮デジタルデータを加工し、それをデコーダに供給するようにすることができ、圧縮デジタルデータのデコーダへの入力速度と上記デコーダの処理速度とが常に一致するようにすることができる。
【００７５】
したがって、フレーム長制御コードの含有率の誤りにより圧縮デジタルデータのデコーダへの入力ビットレートと出力データのサンプリングレートとの間で相対誤差が発生した場合でも、デコード動作に異常が生じないようにし、出力データの連続性が失われてしまうという不都合を防止することができる。例えば、ＭＰＥＧオーディオの圧縮デジタルデータを対象とした場合、再生音声の音切れをなくすようにすることができる。
【００７６】
また、本発明の他の特徴によれば、圧縮デジタルデータのデコーダへの転送ビットレートと外部入力クロックの原周波数との実際の周波数比率と、規格上の周波数比率との誤差を検出し、その誤差をなくすように上記フレーム長制御コードの含有率を変更すべくデータを加工するようにしたので、転送ビットレートと原周波数との周波数比率が規格上の周波数比率と異なってしまった場合でも、その誤差に合わせて含有率を修正するように圧縮デジタルデータを加工してデコーダに供給するようにすることができ、圧縮デジタルデータのデコーダへの入力速度と上記デコーダの処理速度とが常に一致するようにすることができる。
【００７７】
したがって、上記周波数比率の誤りが発生した場合でも、デコード動作に異常が生じないようにし、出力データの連続性が失われてしまうという不都合を防止することができる。例えば、ＭＰＥＧオーディオの圧縮デジタルデータを対象とした場合、再生音声の音切れをなくすようにすることができる。
【００７８】
さらに、本発明のその他の特徴によれば、デコーダの動作基準周波数を発生する元となる原周波数を制御手段に対して外部制御により任意に設定するための手段を設けたので、例えば、様々な圧縮符号化方式に準拠した原周波数を制御手段に対して外部制御により設定することにより、様々な圧縮符号化方式に対応して上述のような誤差の検出を行い、その圧縮符号化方式の下でデコーダが誤りなく動作するための動作基準周波数を発生するようにすることができ、１つの制御手段を複数の圧縮符号化方式に適合させて使用することが容易にできる。
【図面の簡単な説明】
【図１】本実施形態によるデータ変換装置を含むデコードシステムの全体構成を示す概略ブロック図である。
【図２】図１に示した制御手段４のより詳細な構成を示すブロック図である。
【符号の説明】
１ データ変換装置
２ デコーダ
３ 解析器
４ 制御手段
５ データ加工手段
６ 比較手段
７ 周波数変換器
４１ フレーム長制御コード監視手段
４２ フレーム長制御コード正常含有率算出手段
４３ フレーム長制御コードエラー判断手段
４４ 制御信号生成手段
４５ 比率エラー判断手段
４６ 正常周波数比率算出手段[0001]
BACKGROUND OF THE INVENTION
The present invention relates to a data conversion apparatus, and more particularly to a data conversion apparatus suitable for application to a decoding system in which a transfer bit rate of compressed digital data to a decoder and an operation reference frequency of the decoder are set independently of each other. It is.
[0002]
[Prior art]
Among the various compression encoding methods proposed in the past, when the compressed digital data is decoded and output, the input bit rate to the decoder of the compressed digital data and the sampling rate of the output data (decoder In some cases, the operation reference frequency) can be set independently of each other. In this type of system, the ratio between the input bit rate of compressed digital data and the sampling rate of output data is usually kept constant.
[0003]
[Problems to be solved by the invention]
However, there may be a relative error in the ratio between the input bit rate of the compressed digital data and the sampling rate of the output data. When such a relative error occurs, an abnormality occurs in the decoding process operation. In a normal decoder, if an abnormality occurs, the compressed digital data that is continuously input is decoded after resetting, so that the continuity of output data is lost when an abnormality occurs. There was an inconvenience.
[0004]
For example, a decoder compliant with the MPEG1 audio (ISO / IEC 11172-3) compression coding system has a buffer memory for internal processing and the like, and this buffer capacity has a certain margin. Even when a relative error occurs, the decoded speech can be normally output as much as the margin.
[0005]
However, when a certain amount of time relative error occurs, extra data is input beyond the margin of the buffer memory, or the buffer memory becomes empty, and an abnormal buffer overflow or data shortage occurs. As a result, the audio data in the section where such an abnormality has occurred is interrupted. For example, when it is used for music, the music is interrupted in the middle, and music appreciation is hindered.
[0006]
As an example of the cause of the relative error in the ratio between the input bit rate of the compressed digital data and the sampling rate of the output data, for example, in order to adjust the frame length of the compressed digital data to a standard length There is a case where there is an error in the frame length control code itself.
[0007]
For example, in an audio layer II of MPEG audio, one audio frame is composed of 1152 PCM units. Here, when the bit rate of the compressed digital data is 224 Kbps and the sampling frequency at the time of compression is 44.1 KHz, the number of bits per unit time of the audio frame is 224 K × (1152 / 44.1 K) = 5851.42857. (Hereinafter, a normal frame length according to this standard is assumed to be FlengthN).
[0008]
As described above, when the number of bits of the audio frame per unit time is accurately calculated, a number with a decimal point is obtained. However, since the number of bits must be an integer, rounding of the decimal part is performed. In the MPEG audio layer II, the length per frame is standardized so that the number of bits is divisible by 8. In this case, 5848 bits (hereinafter, referred to as the two values closest to the above accurately calculated value) are used. The frame length A is rounded to either FlengthA) or 5856 bits (hereinafter, the frame length B is assumed to be FlengthB).
[0009]
However, if such rounding is simply performed, an error occurs between the original bit rate and inconvenience occurs. Therefore, a frame length control code (dummy data) called padding is used to eliminate the apparent difference between the amount of data represented by the rounded frame length and the amount of compressed digital data actually input. To adjust the frame length.
[0010]
In other words, in the MPEG audio layer II, a frame length A is not padded, and a frame length B is padded, and a plurality of audio frames having both attributes are transmitted at a constant ratio, so that it can be seen in the long term. This is intended to be consistent with the amount of compressed digital data actually sent. In the above example, the ratio of no padding to padding must be 4: 3.
[0011]
However, if the ratio of the presence / absence of padding is incorrect, the input speed of the compressed digital data and the decoding processing speed differ in the actual decoding process. For example, if there is no padding, the compressed digital data enters at a speed faster than the decoding speed, resulting in buffer overflow. As a result, as described above, there is a problem in that the output audio data is interrupted by the reset operation and the music appreciation is hindered.
[0012]
The present invention has been made to solve such a problem, and even when a relative error occurs between the input bit rate of the compressed digital data to the decoder and the sampling rate of the output data from the decoder. An object is to prevent an abnormality from occurring in the decoding operation (for example, to prevent output audio data from being interrupted).
[0013]
[Means for Solving the Problems]
In the data conversion apparatus of the present invention, the transfer bit rate to the decoder of the compressed digital data in which at least the frame length control code for adjusting the frame length is included in the bit stream and the operation reference frequency of the decoder are mutually In an independently set system, an analysis means for analyzing the presence or absence of a frame length control code included in the compressed digital data for each frame unit, and the frame length control based on an analysis result by the analysis means By comparing the actual content ratio of the code with the content ratio determined from the standard of the above compressed digital data, an error between the two is detected, and a control signal indicating the content ratio is generated to correct the error. According to the control signal generated by the control means and the control means. And a data processing means for processing the data so as to change the content of the frame length control codes.
[0014]
Another feature of the present invention is that the data processing means is a combination of a frame including a frame length control code and a frame not including the frame length control code in the compressed digital data in accordance with a control signal generated by the control means. The content rate of the frame length control code is changed by changing the ratio.
[0015]
Another feature of the present invention is that the control means performs control so as to obtain an actual content rate of a frame length control code included in the compressed digital data by statistical processing every predetermined time. Features.
[0016]
Another feature of the present invention is that the compressed digital data includes a frame length code for representing a standard frame length in addition to the frame length control code in the bit stream. The means is characterized in that a standard content rate of the frame length control code is obtained based on the frame length code analyzed by the analyzing means.
[0017]
Another feature of the present invention is that the frequency ratio between the actual transfer bit rate of the compressed digital data to the decoder and the original frequency of the clock input from the outside in order to generate the operation reference frequency of the decoder. A comparison means for detecting an error between the two by comparing with a frequency ratio determined from the standard; a control means for generating a control signal for correcting the error detected by the comparison means; and the control means And data processing means for processing the data so as to change the content rate of the frame length control code in the compressed digital data in accordance with the control signal generated by.
[0018]
Another data conversion apparatus according to the present invention provides an actual data transfer to the decoder of the compressed digital data in a system in which the transfer bit rate of the compressed digital data to the decoder and the operation reference frequency of the decoder are set independently of each other. Comparison that detects the error between the transfer bit rate and the original frequency of the clock input from the outside to generate the operation reference frequency of the decoder by comparing it with the frequency ratio determined from the standard Means, a control means for generating a control signal for correcting the error detected by the comparison means, and a frame length control code in the compressed digital data according to the control signal generated by the control means. Data processing means for processing data so as to change the content rate.
[0020]
Since the present invention comprises the above technical means, even if the content rate of the frame length control code included in the compressed digital data is different from the standard content rate, the actual content rate and the standard content rate. The compressed digital data is processed by the control signal corresponding to the error and the content rate of the frame length control code is corrected by the control signal corresponding to the error. The input speed of data to the decoder is controlled so as to always match the processing speed of the decoder.
[0021]
According to another aspect of the present invention, the frequency ratio between the actual transfer bit rate of the compressed digital data to the decoder and the original frequency from which the operation reference frequency of the decoder is generated is different from the standard frequency ratio. Even if the error has occurred, the error in the frequency ratio between the two is detected by the comparison means, the compressed digital data is processed by the control signal corresponding to the error, and the content rate of the frame length control code is corrected. The input speed of the compressed digital data to the decoder is controlled so as to always match the processing speed of the decoder.
[0022]
Furthermore, according to another feature of the present invention, for example, by setting an original frequency compliant with various compression encoding schemes to the control means by external control, the above-described configuration is compatible with various compression encoding schemes. Thus, it is possible to generate an operation reference frequency for the decoder to operate without error under the compression encoding method by the frequency conversion means.
[0023]
DETAILED DESCRIPTION OF THE INVENTION
Hereinafter, an embodiment of the present invention will be described with reference to the drawings.
FIG. 1 is a schematic block diagram showing an overall configuration of a decoding system including a data conversion apparatus according to the present embodiment. In the following description, it is assumed that the decoding system shown in FIG. 1 conforms to MPEG audio layer II.
[0024]
In FIG. 1, compressed digital data is supplied to a decoder 2 via a data converter 1, where decoding processing is performed. This decoding process is performed according to a predetermined operation reference clock, and the operation reference clock is given by multiplying the original frequency of the clock input from the outside of the data converter 1 by the frequency converter 7.
[0025]
The frequency converter 7 includes, for example, two frequency dividers having different frequency division ratios, a phase comparator that obtains a phase difference between both outputs of the frequency dividers, and an oscillation frequency based on the obtained phase difference. And a voltage-controlled oscillator (VCO) that feeds it back to one of the frequency dividers.
[0026]
The decoding system shown in FIG. 1 is configured such that the transfer bit rate of compressed digital data to the decoder 2 and the operation reference frequency of the decoder 2 are set independently of each other. In this embodiment, in order to prevent inconvenience such as sound interruption as described in the conventional example, the ratio between the transfer bit rate and the operation reference frequency is always constant, that is, a decoder for compressed digital data. 2 and the processing speed of the decoder 2 always match.
[0027]
The compressed digital data includes a frame length control code (padding) for adjusting the frame length of the compressed digital data and a frame for representing a standard frame length based on MPEG audio layer II in the bit stream. Contains long code (bit rate and sampling frequency information during compression).
[0028]
In the case of layer II of MPEG audio, the frame length code is included in the bit stream of the compressed digital data. However, in order to apply this embodiment, at least the frame length control code is included. Well, the frame length code need not necessarily be included. However, it is assumed that the frame length is set in some form according to the standard of the compression encoding method to be applied (for example, a certain value is set by the standard).
[0029]
Further, the frequency information of the operation reference clock of the decoder 2 may not be included in the bit stream of the compressed digital data. However, like the frame length code, it is set in some form according to the compression coding standard. It shall be. That is, as the frequency of the operation reference clock, it is necessary to select a certain multiple of the sampling frequency at the time of compression for the convenience of the decoding system, but the value is specified in the data converter 1 or set from the outside. And
[0030]
Further, the original frequency of the clock input from the outside of the data converter 1 to generate the operation reference clock of the decoder 2 is also defined in the data converter 1 or set from the outside.
[0031]
In the following, the sampling frequency (hereinafter referred to as Fs) included in the bit stream of the compressed digital data compliant with MPEG audio layer II is 44.1 KHz, and the bit rate (hereinafter referred to as BR) is 224 Kbps. And Further, it is assumed that the magnification (hereinafter referred to as Dfs) of the operation reference clock required by the decoder 2 with respect to the sampling frequency Fs is 384 times and the original frequency (hereinafter referred to as CLKm) is 13.5 MHz.
[0032]
As described above, normally, the bit rate of the compressed digital data is adjusted to the bit rate according to the standard by the frame length control code (padding), and the input speed of the compressed digital data to the decoder 2 and the processing speed of the decoder 2 are Has always been made to match. That is, the audio frame including padding and the audio frame not including padding are transmitted at a constant ratio, so that the total bit rate is adjusted to the standard.
[0033]
However, an error may occur in the content rate of the frame length control code itself, which may cause an abnormality in the decoding operation. In the present embodiment, in order to cope with such a case, the analyzer 3, the control unit 4, and the data processing unit 5 perform processing as described below.
[0034]
The analyzer 3 analyzes three pieces of information (frame length control code, sampling frequency Fs, bit rate BR) included in the bit stream of the compressed digital data. That is, the analyzer 3 checks the presence / absence of a frame length control code for each audio frame unit and analyzes the contents of the sampling frequency Fs and the bit rate BR included in the header of the bit stream.
[0035]
Based on the analysis result by the analyzer 3, the control unit 4 calculates the ratio (content ratio) in which the frame length control code is actually included in the compressed digital data, the sampling frequency Fs, and the bit rate BR. By comparing the calculated content rate with the standard, an error between the two is detected, and a control signal is generated to correct the error.
[0036]
Then, the data processing means 5 processes the data so as to change the content rate of the frame length control code in the compressed digital data in accordance with the control signal generated by the control means 4. Specifically, in accordance with the control signal, an audio frame including a frame length control code (hereinafter, a frame having the frame length B is referred to as Frame B) and an audio frame not including the frame (hereinafter, the frame having the frame length A is referred to as Frame A). The content ratio of the frame length control code in the compressed digital data is changed.
[0037]
That is, by appropriately changing the blending ratio of Frame A and Frame B according to the error rate (described later) of the error that has occurred, the input speed of the compressed digital data input to the decoder 2 and the input speed to the decoder 2 are input. The decoder 2 operates normally by matching with the data processing speed. It should be noted that how to change the blending ratio of Frame A and Frame B is determined by the control means 4 obtaining the value of n from the following equations (1) and (2).
[0038]
(Error rate) × {(Frame length based on blend ratio after change) / (Frame length based on blend ratio of input data)} = 1.0 (1)
However, the frame length based on the blending ratio after the change is n (0 ≦ n ≦ 1), where the ratio of Frame B is included,
FlengthA * (1-n) + FlengthB * n = 5848 * (1-n) + 5856 * n = 5848 + 8 * n (2)
[0039]
FIG. 2 shown next is a block diagram showing a more detailed configuration of the control means 4. Hereinafter, the operations of the control means 4 and the data processing means 5 described above will be described in more detail with reference to FIG.
[0040]
The result of analyzing the presence or absence of the frame length control code by the analyzer 3 in FIG. 1 is given to the frame length control code monitoring means 41 in FIG. Based on the analysis result, the frame length control code monitoring means 41 obtains the ratio of the frame length control code included in the actual bit stream by statistical processing every certain time (for example, 2 seconds).
[0041]
On the other hand, the result of analyzing the frame length code (sampling frequency Fs and bit rate BR) by the analyzer 3 in FIG. 1 is given to the frame length control code normal content calculating means 42 in FIG. The frame length control code normal content rate calculating means 42 calculates the ratio of the normal frame length control code determined from the MPEG audio layer II standard based on the analysis result of the frame length code given from the analyzer 3. calculate.
[0042]
The frame length control code error determining means 43 includes an actual ratio of frame length control codes included in the compressed digital data obtained by the frame length control code monitoring means 41, and a frame length control code normal content rate calculating means 42. The ratio of the frame length control code included in the calculated normal data is compared, and the presence / absence of an abnormality (error) is determined from the error.
[0043]
Note that in the case of a compression encoding method in which the sampling frequency Fs and the bit rate BR are not included in the bit stream of the compressed digital data, the frame length control code normal content rate calculating means 42 is unnecessary, and the compression encoding is not necessary. What is necessary is just to perform the above comparison using the content rate determined on the specification of a system.
[0044]
However, the content rate is slightly different from the normal one in the short term because of the relationship between the time intervals in the statistical processing, but it may be normal when considering the time intervals before and after. In such a case, it is not preferable to determine an error, and it is necessary to suppress it.
[0045]
In order to realize this, in the present embodiment, accumulation of errors in the frame length control code content rate is calculated, and an error is determined when the accumulated value exceeds a certain threshold. That is, the content ratios obtained by statistical processing at regular intervals are sequentially FIP0, FIP1, FIP2, ..., the normal content ratio is FIPnorm, the threshold for determining an error is ErrLimit, and the following (3) ( An error may be determined when any of the conditional expressions in 4) is satisfied.
[0046]
Overflow judgment:
sum (FIP0-FIPnorm, FIP1-FIPnorm, FIP2-FIPnorm,…)> ErrLimit (3)
Underflow judgment:
sum (FIP0-FIPnorm, FIP1-FIPnorm, FIP2-FIPnorm,…) <-ErrLimit (4)
[0047]
Here, as a specific example when an error occurs in the content rate of the frame length control code (padding), a case will be described in which padding is included in all audio frames during a certain period of time when statistical processing is performed. When all the padding is present, the number of frames included in one second in the compressed digital data is different from the normal case (no padding: present = 4: 3) as follows.
[0048]
The number of frames per second of a normal bitstream:
224K / 585.429 [frame / sec]
Number of frames included per second of the bitstream including the error:
224K / 5856 [frame / sec]
[0049]
When such an error is detected, the frame length control code error determination means 43 calculates an error rate ErrFrame according to the following equation (5).
ErrFrame = FlengthB / FlengthN = 5856 / 5851.429 (5)
Then, the calculated error rate ErrFrame is supplied to the control signal generation unit 44. Based on the supplied error rate ErrFrame, the control signal generation unit 44 generates a control signal for changing the blending ratio of Frame A and Frame B (the content ratio of Frame A and Frame B in the compressed digital data) (specifically, Calculates the value of n shown in the above equations (1) and (2)) and supplies it to the data processing means 5.
[0050]
When an error that all the padding is present as described above, the length (frame length according to the blend ratio of the input data) shown in the equation (1) is FlengthB. Therefore, in this case, the ratio n at which Frame B should be included in the compressed digital data after data processing is calculated as the following equations (6) and (7) based on the above equations (1) and (2). .
[0051]

[0052]
As described above, since all the frames included in the actual compressed digital data are Frame B, in this case, 4/7 of all the frames included in the compressed digital data are determined as Frame B (frame with padding). By processing the data so that it is converted to Frame A (frame without padding), the input speed of the compressed digital data to the decoder 2 and the processing speed of the decoder 2 are matched, and the decoder 2 operates normally. Can be.
[0053]
In addition, it is necessary to deal with a side effect caused by changing the padding at the same time. Specifically, in the case of MPEG audio, when protection protection for enabling data error discrimination based on a CRC (Cyclic Redundancy Check) code included in the header of a bit stream of compressed digital data is enabled, the CRC code value is changed. I do. In addition, unnecessary data is deleted or empty data is added in response to the actual frame length changing.
[0054]
As described above, according to the present embodiment, even when the actual padding ratio included in the bit stream of the compressed digital data is different from the standard ratio, the compressed digital data to the decoder 2 is obtained. And the processing speed of the decoder 2 can always coincide with each other, so that no abnormality occurs in the decoding operation.
[0055]
The above-described embodiment is an example in the case where there is no error in both the input speed and the original frequency of the compressed digital data input from the outside of the data converter 1 (at least there is no error in the frequency ratio between the two). . However, this frequency ratio can also deviate from the normal ratio.
[0056]
For example, the normal ratio between the transfer bit rate BR of compressed digital data and the original frequency CLKm is
BR / CLKm = 224KHz / 13.5MHz
On the other hand, when compressed digital data is transferred at a bit rate BR ′ (for example, 219 Kbps) different from the standard or a clock is input at an original frequency CLKm ′ (for example 13.2 MHz) different from the standard, the frequency The ratio of
BR ′ / CLKm ′ = 219 KHz / 13.2 MHz
It becomes.
[0057]
In this way, even if an error occurs in the compressed digital data transfer bit rate itself, or an error occurs in the original frequency of the external input clock itself, it is the same as in the case where an error occurs in the padding content rate described above. In addition, the decoder 2 may operate abnormally due to error accumulation. Therefore, in the present embodiment, in order to cope with such a case, the comparison means 6 is provided as shown in FIG.
[0058]
The comparison means 6 compares the frequency ratio when no error has occurred with the frequency ratio when an error has occurred, and detects the error. That is, the comparison means 6 cannot detect an absolute deviation between the normal bit rate BR and the error bit rate BR ′, or an absolute deviation between the normal original frequency CLKm and the error original frequency CLKm ′. A deviation between the normal frequency ratio and the error frequency ratio can be detected.
[0059]
In this case, as shown in FIG. 2, the control means 4 receives the comparison result as described above in the comparison means 6 by the ratio error determination means 45 and determines whether or not an error relating to the frequency ratio has occurred. As a method for determining the presence / absence of a frequency ratio error, as in the method for determining the presence / absence of a padding content rate error, an error is determined when the sum of errors detected at a certain time exceeds a certain value. You can do it.
[0060]
If it is determined that a frequency ratio error as described above has occurred, the ratio error determination means 45 calculates an error rate ErrFreq according to the following equation (8).

Then, the calculated error rate ErrFreq is supplied to the control signal generator 44.
[0061]
Based on the supplied error rate ErrFreq, the control signal generating means 44 generates a control signal for changing the blending ratio of Frame A and Frame B (the ratio n including Frame B according to the equations (1) and (2)). A value is calculated) and supplied to the data processing means 5. If a frequency ratio error occurs as in this example, there is no error in the padding content itself, so the equation (1) (the frame length based on the blend ratio of the input data) is FlengthN.
[0062]
Therefore, in this case, the ratio n at which Frame B should be included in the compressed digital data after data processing is calculated as in the following equations (9) and (10).

[0063]
On the other hand, since the ratio of Frame B included in the compressed digital data before data processing is 0.42857 (= 3/7), in this case, 0.07427 (= 0.050284-0.42857) of all the frames included in the compressed digital data. ) Is converted from Frame A (frame without padding) to Frame B (frame with padding), so that the above-described problems can be avoided.
[0064]
In the case of the example described above, the error rates ErrFrame and ErrFreq are both values close to 1.0, so that the problem can be avoided. However, the error rate of the error that has occurred is far from 1.0. If this is the case, the problem may not be avoided. In this case, the following two methods are conceivable. One is a method of stopping the reproduction of the sound because it deviates from the controllable range, and the other is a method of generating a certain sound for the time being even if it is not a normal sound.
[0065]
The former method of stopping the reproduction of sound can be realized by the control means 4 outputting a reproduction stop signal to the decoder 2. Alternatively, it can be realized by sending an MPEG audio data frame containing silent data to the decoder 2 instead of the data frame to be reproduced.
[0066]
Here, a point to be noted when sound reproduction is stopped is that a sound reproduction stop time should be lengthened to some extent (time in seconds). The reason is that if the sound playback stop time is extremely short, the sound playback and stop are repeated in a short time so that the sound output from the speaker (not shown) can be heard as a “machine gun” sound. turn into. This is very harsh and also has an adverse effect on the speaker itself, which needs to be suppressed.
[0067]
On the other hand, in the case where the sound reproduction is not stopped as in the latter method, in MPEG audio, data output from the decoder 2 is originally reproduced by reproducing and deleting a certain data frame itself. Although it is different from the above, it is possible to suppress the influence on the sound quality as compared with the case where an error occurs.
[0068]
Further, in the present embodiment, as shown in FIGS. 1 and 2, the control means 4 controls the original frequency of the original clock for generating the operation reference clock of the decoder 2 by external control of the data converter 1 (for example, manually). Means for setting are also provided.
[0069]
Thereby, for example, by setting an original frequency that complies with a compression encoding scheme other than MPEG audio, error detection is performed in accordance with various compression encoding schemes, and the decoder 2 operates under the compression encoding scheme. The frame length control code (padding) can be processed so as to operate without error. That is, it is possible to easily use one control means 4 in conformity with a plurality of purposes.
[0070]
In this case, the original frequency set by the external control is given to the normal frequency ratio calculating means 46 shown in FIG. The normal frequency ratio calculation means 46 calculates the compression code from the set original frequency and the standard transfer bit rate analyzed by the analyzer 3 (bit rate of the compressed digital data by the corresponding compression encoding method). The normal frequency ratio when no error has occurred in the conversion method is calculated.
[0071]
Then, the ratio error determination unit 45 calculates the error rate ErrFreq based on the calculation result in the normal frequency ratio calculation unit 46 in the same manner as the above equation (8), and supplies it to the control signal generation unit 44. . Based on the supplied error rate ErrFreq, the control signal generating unit 44 generates a control signal for changing the mixing ratio of Frame A and Frame B and supplies it to the data processing unit 5.
[0072]
In the embodiment described above, as shown in FIG. 1, means for changing the content rate of the frame length control code by looking at the error of the padding content rate, the transfer bit rate and the original frequency of the compressed digital data Although all the means for changing the content rate of the frame length control code in view of the error in the ratio are included, they can be applied independently or in any combination.
[0073]
Further, although the MPEG audio layer II has been exemplified as the compression encoding method, the present invention is not limited to this. For example, as long as the frame length control code is included in the bit stream of the compressed digital data, such as a compression encoding method such as MPEG1, MPEG2, AC-3, etc. for compressing an image, any of them may be applied. Can do.
[0074]
【The invention's effect】
As described above, the present invention analyzes the content rate of the frame length control code included in the compressed digital data and compares the analyzed actual content rate with the standard content rate to determine the error between the two. Since the data is processed to change the content rate of the frame length control code so as to eliminate the error, the content rate of the frame length control code included in the compressed digital data is in accordance with the standard. Even if it differs from the content rate, the compressed digital data can be processed so that the content rate is correct and supplied to the decoder. The input speed of the compressed digital data to the decoder and the decoder It is possible to always match the processing speed.
[0075]
Therefore, even if a relative error occurs between the input bit rate to the decoder of the compressed digital data and the sampling rate of the output data due to an error in the content rate of the frame length control code, an abnormality does not occur in the decoding operation. The inconvenience that the continuity of the output data is lost can be prevented. For example, when compressed digital data of MPEG audio is targeted, it is possible to eliminate interruptions in reproduced sound.
[0076]
According to another feature of the present invention, an error between the actual frequency ratio between the bit rate of the compressed digital data transferred to the decoder and the original frequency of the external input clock and the standard frequency ratio is detected. Since the data was processed to change the content rate of the frame length control code so as to eliminate the error, even if the frequency ratio between the transfer bit rate and the original frequency is different from the standard frequency ratio, The compressed digital data can be processed and supplied to the decoder so that the content rate is corrected in accordance with the error, and the input speed of the compressed digital data to the decoder and the processing speed of the decoder always match. Can be.
[0077]
Therefore, even when the frequency ratio error occurs, it is possible to prevent the decoding operation from becoming abnormal and prevent the inconvenience that the continuity of the output data is lost. For example, when compressed digital data of MPEG audio is targeted, it is possible to eliminate interruptions in reproduced sound.
[0078]
Furthermore, according to the other feature of the present invention, since means for arbitrarily setting the original frequency for generating the operation reference frequency of the decoder to the control means by external control is provided, for example, various By setting the original frequency conforming to the compression coding method to the control means by external control, the error detection as described above is performed in accordance with various compression coding methods, and under the compression coding method, Therefore, it is possible to generate an operation reference frequency for the decoder to operate without error, and it is easy to use one control unit adapted to a plurality of compression coding systems.
[Brief description of the drawings]
FIG. 1 is a schematic block diagram showing an overall configuration of a decoding system including a data conversion apparatus according to an embodiment.
FIG. 2 is a block diagram showing a more detailed configuration of the control means 4 shown in FIG.
[Explanation of symbols]
1 Data converter
2 Decoder
3 analyzer
4 Control means
5 Data processing means
6 comparison means
7 Frequency converter
41 Frame length control code monitoring means
42 Means for calculating normal content rate of frame length control code
43 Frame length control code error judging means
44 Control signal generating means
45 Ratio error judgment means
46 Normal frequency ratio calculation means

Claims (6)

In a system in which a transfer bit rate to a decoder of compressed digital data including at least a frame length control code for adjusting the frame length and the operation reference frequency of the decoder are set independently of each other ,
Analysis means for analyzing the presence or absence of a frame length control code included in the compressed digital data for each frame unit;
Based on the analysis result by the analysis means, the error of both is detected by comparing the actual content rate of the frame length control code with the content rate determined from the standard of the compressed digital data. Control means for generating a control signal indicating the content rate to correct
A data conversion device comprising: data processing means for processing data so as to change a content rate of a frame length control code in the compressed digital data in accordance with a control signal generated by the control means.   The data processing means changes the blending ratio of the frame including the frame length control code and the frame not including the frame length control code in the compressed digital data in accordance with the control signal generated by the control means. The data conversion device according to claim 1, wherein the content rate of the content is changed.   2. The data conversion according to claim 1, wherein the control means performs control so as to obtain an actual content rate of a frame length control code included in the compressed digital data by statistical processing at predetermined time intervals. apparatus. The compressed digital data includes, in the bitstream, a frame length code for representing the standard frame length in addition to the frame length control code,
2. The data conversion apparatus according to claim 1, wherein the control means obtains a standard content rate of the frame length control code based on the frame length code analyzed by the analysis means. Compare the frequency ratio between the actual transfer bit rate of the compressed digital data to the decoder and the original frequency of the clock input from the outside to generate the operation reference frequency of the decoder with the frequency ratio determined from the standard Comparing means for detecting the error between the two,
Control means for generating a control signal for correcting the error detected by the comparison means;
2. Data processing means for processing data so as to change a content rate of a frame length control code in the compressed digital data in accordance with a control signal generated by the control means. The data converter of any one of -4. In a system in which the transfer bit rate of the compressed digital data to the decoder and the operation reference frequency of the decoder are set independently of each other,
Compare the frequency ratio between the actual transfer bit rate of the compressed digital data to the decoder and the original frequency of the clock input from the outside to generate the operation reference frequency of the decoder with the frequency ratio determined from the standard Comparing means for detecting the error between the two,
Control means for generating a control signal for correcting the error detected by the comparison means;
A data conversion device comprising: data processing means for processing data so as to change a content rate of a frame length control code in the compressed digital data in accordance with a control signal generated by the control means.

Images