JPH02153639A - System for controlling division and normalization of band - Google Patents
System for controlling division and normalization of bandInfo
- Publication number
- JPH02153639A JPH02153639A JP63306862A JP30686288A JPH02153639A JP H02153639 A JPH02153639 A JP H02153639A JP 63306862 A JP63306862 A JP 63306862A JP 30686288 A JP30686288 A JP 30686288A JP H02153639 A JPH02153639 A JP H02153639A
- Authority
- JP
- Japan
- Prior art keywords
- normalization
- signal
- normalization coefficient
- band
- coefficient
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000010606 normalization Methods 0.000 title claims abstract description 118
- 238000013139 quantization Methods 0.000 claims abstract description 23
- 230000005236 sound signal Effects 0.000 claims abstract description 19
- 238000001514 detection method Methods 0.000 claims abstract description 13
- 238000000034 method Methods 0.000 claims description 8
- 230000006866 deterioration Effects 0.000 abstract description 5
- 238000010586 diagram Methods 0.000 description 9
- 230000005540 biological transmission Effects 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- ZVQOOHYFBIDMTQ-UHFFFAOYSA-N [methyl(oxido){1-[6-(trifluoromethyl)pyridin-3-yl]ethyl}-lambda(6)-sulfanylidene]cyanamide Chemical compound N#CN=S(C)(=O)C(C)C1=CC=C(C(F)(F)F)N=C1 ZVQOOHYFBIDMTQ-UHFFFAOYSA-N 0.000 description 1
- 230000002542 deteriorative effect Effects 0.000 description 1
- 238000005070 sampling Methods 0.000 description 1
- 239000007787 solid Substances 0.000 description 1
Landscapes
- Analogue/Digital Conversion (AREA)
- Compression, Expansion, Code Conversion, And Decoders (AREA)
- Transmission Systems Not Characterized By The Medium Used For Transmission (AREA)
Abstract
Description
【発明の詳細な説明】
〔概要〕
音声の高能率符号化に於ける帯域分割正規化制御方式に
関し、
補助情報を増加させることなく、量子化部への入力のク
リップを防止することを目的とし、音声信号を複数の帯
域に分割し、各帯域分割信号の正規化係数を算出する正
規化係数部とぐ該正規化係数部に於いて算出した正規化
係数の中の最大値を検出する最大値検出部と、前記帯域
分割信号をそれぞれ正規化する正規化部と、該正規化部
により正規化された信号を量子化する量子化部とを備え
、前記最大値検出部により検出された最大値の正規化係
数を代表正規化係数として、前記正規化部により各帯域
分割信号の正規化を行わせるように構成した。[Detailed Description of the Invention] [Summary] This invention aims to prevent clipping of the input to the quantization unit without increasing auxiliary information regarding a band division normalization control method in high-efficiency audio encoding. , a normalization coefficient section that divides the audio signal into multiple bands and calculates the normalization coefficient of each band-divided signal; and a maximum value that detects the maximum value among the normalization coefficients calculated in the normalization coefficient section. A value detection section, a normalization section that normalizes each of the band-divided signals, and a quantization section that quantizes the signal normalized by the normalization section, and the maximum value detected by the maximum value detection section is The normalization unit is configured to normalize each band-divided signal by using the normalization coefficient of the value as a representative normalization coefficient.
本発明は、音声の高能率符号化に於ける帯域分割正規化
制御方式に関するものである。The present invention relates to a band division normalization control method in high-efficiency audio encoding.
音声信号をディジタル信号に変換する場合、例えば、８
ＫＨｚのサンプリング周波数でサンプリングし、８ビツ
トの量子化により、６４ Ｋ ｂ ／ ｓのディジタル
信号とするのが一般的である。When converting an audio signal to a digital signal, for example, 8
It is common to sample at a sampling frequency of KHz and perform 8-bit quantization to obtain a 64 Kb/s digital signal.
このようなビットレートを更に低下させる高能率符号化
方式で用いられる量子化操作にも、各種の方式が提案さ
れている。例えば、音声信号は、高域より低域のパワー
が一般には大きいものであるから、音声信号を複数の帯
域に分割し、低域側にビット数を多く割当てて量子化す
る方式も知られている。Various methods have been proposed for quantization operations used in such high-efficiency encoding methods that further reduce the bit rate. For example, since audio signals generally have greater power in the low range than in the high range, it is also known to divide the audio signal into multiple bands and quantize it by allocating a larger number of bits to the lower range. There is.
又量子化器の量子化可能の最大入力レベルに各分割帯域
の信号レベルを抑制する為の正規化を行う方式も知られ
ており、量子化器に於いてクリップされることによる音
質低下を防止している。このような帯域分割された信号
の正規化を行う方式に於いて、補助情報の増加がなく、
且つ音質の劣化がないように、量子化できることが要望
されている。There is also a known method of normalization to suppress the signal level of each subband to the maximum input level that can be quantized by the quantizer, thereby preventing deterioration in sound quality due to clipping in the quantizer. are doing. In this method of normalizing band-divided signals, there is no increase in auxiliary information,
In addition, it is desired to be able to perform quantization without deteriorating the sound quality.
第３図は従来例のブロック図であり、音声信号は帯域分
割部２２で、例えば、離散フーリエ変換（Ｄ Ｆ Ｔ
； Ｄｉｓｃｒｅｔｅ Ｆｏｕｒｉｅｒ Ｔｒａｎｓｆ
ｏｒｍ）により周波数成分に展開されて、低域信号Ｘｌ
と中域信号ｘ２と高域信号ｘ３に帯域分割され、正規化
係数部２１−１〜２１−３によりそれぞれ正規化係数ｓ
、−３３が算出され、正規化部２３−１〜２３−３に於
いて正規化処理が行われる。そして、それぞれ量子化部
２４−１〜２４−３に於いて割当てられたビット数に従
って量子化され、多重化部２５に於いて量子化ビットが
多重化され、且つ正規化の補助情報として各正規化係数
Ｓ、〜Ｓ３が多重化されて送出される。FIG. 3 is a block diagram of a conventional example, in which the audio signal is processed by the band division section 22, for example, by discrete Fourier transform (DFT).
; Discrete Fourier Transf
orm) into frequency components, and the low frequency signal Xl
The band is divided into a mid-range signal x2 and a high-frequency signal
, -33 are calculated, and normalization processing is performed in the normalization units 23-1 to 23-3. Then, the quantization units 24-1 to 24-3 quantize each bit according to the allocated bit number, the multiplexing unit 25 multiplexes the quantized bits, and each normalization bit is used as auxiliary information for normalization. The coding coefficients S, -S3 are multiplexed and sent out.
受信側では、分割帯域対応の正規化係数Ｓ、〜Ｓ３に従
って帯域分割された信号を伸張し、逆ＤＦＴにより音声
信号を再生することになる。On the receiving side, the band-divided signal is expanded according to the normalization coefficients S, ~S3 corresponding to the divided bands, and the audio signal is reproduced by inverse DFT.
第４図は正規化動作説明図であり、例えば、低域、中域
、高域の帯域Ｂ、、Ｂ２．Ｂ３に分割された場合の量子
化部２４−１〜２４−３に割当てられたレベル数をｎ１
〜ｎ３とし、量子化部２４−１〜２４−３の入力レベル
をＱＬ （ｎｌ） 〜ＱＬ（ｎ３）、又各分割帯域Ｂ
、 ％Ｂ ３の最大値をＸ１□８〜Ｘ３．□とすると、
各分割帯域の正規化係数５Ｉ−３３は、
Ｓ 、 ＝ＱＬ （ｎ １） ／Ｘ １．、ＸＳ ２
”ＱＬ （ｎ ２） ／ Ｘ ２ｍａｍＳ ３ ＝ＱＬ
（ｎ ３） ／ Ｘ ３ａａｍによって求められる。FIG. 4 is an explanatory diagram of the normalization operation, and includes, for example, low, middle, and high frequency bands B, , B2 . The number of levels assigned to the quantization units 24-1 to 24-3 when divided into B3 is n1.
~n3, the input level of the quantizers 24-1 to 24-3 is QL (nl) ~QL(n3), and each divided band B
, %B 3 maximum value from X1□8 to X3. If □,
The normalization coefficient 5I-33 for each subband is S, =QL (n 1) /X 1. ,XS 2
”QL (n 2) / X 2mamS 3 =QL
It is determined by (n 3) / X 3aam.
この正規化係数ｓ、ｘｓ３により、第４図の曲線ａに示
す信号は、正規化部２３−１〜２３−３に於いて点線曲
線すに示すように正規化される。Using the normalization coefficients s and xs3, the signal shown by the curve a in FIG. 4 is normalized in the normalization sections 23-1 to 23-3 as shown by the dotted curve.
従って、量子化部２４−１〜２４−３へ入力される信号
ＱＢ’−Ｑ３は、
Ｑ ｒ −Ｘ ＋ ／ Ｓ Ｌ
Ｑ ｚ ”” Ｘ ２ ／ Ｓ ２
Ｑ ｓ ” Ｘ ｓ ／ Ｓ ｓ
となり、各分割帯域の信号Ｑ、％Ｑ３は、正規化マスク
パターンＣを超えないものとなり、それぞれ量子化部２
４−１〜２４−３に於いて量子化される０例えば、低域
分割帯域Ｂ１の信号Ｑ、は４ビツト、中域分割帯域Ｂ２
の信号Ｑ２は２ビツト、高域分割帯域Ｂ３の信号Ｑ３は
１ビツトにそれぞれ量子化される。Therefore, the signal QB'-Q3 input to the quantization units 24-1 to 24-3 becomes Qr-X+/SLQz""X2/S2Qs"Xs/Ss, The signals Q and %Q3 of each subband do not exceed the normalized mask pattern C, and the quantization unit 2
For example, the signal Q of the low frequency subband B1 is quantized in 4-1 to 24-3, and the signal Q of the low frequency subband B1 is 4 bits.
The signal Q2 of the high frequency subband B3 is quantized to 2 bits, and the signal Q3 of the high frequency division band B3 is quantized to 1 bit.
前述の第３図及び第４図に示す従来例に於いては、分割
帯域対応の正規化係数８１〜Ｓ３を正規化の補助情報と
して伝送するものであるが、この補助情報を削減する為
に、従来は、例えば、第５図又は第６図に示す構成が提
案されている。In the conventional example shown in FIGS. 3 and 4 described above, the normalization coefficients 81 to S3 corresponding to the divided bands are transmitted as auxiliary information for normalization, but in order to reduce this auxiliary information, Conventionally, for example, a configuration shown in FIG. 5 or 6 has been proposed.
第５図に於いては、低域選択部３２により音声信号の低
域分割帯域Ｂ、の信号を選択し、この信号を用いて正規
化係数部３１に於いて正規化係数Ｓ、を算出し、この正
規化係数Ｓ−を代表正規化係数として、正規化部３３に
於いて音声信号の正規化を行って量子化部３４に加える
もので、補助情報としては、１個の低域分割帯域Ｂ、の
正規化係数Ｓ、のみとなり、伝送情報量を削減すること
ができる。In FIG. 5, the low frequency selection section 32 selects the signal of the low frequency subband B of the audio signal, and using this signal, the normalization coefficient section 31 calculates the normalization coefficient S. , using this normalization coefficient S- as a representative normalization coefficient, the normalization section 33 normalizes the audio signal and adds it to the quantization section 34.As auxiliary information, one low frequency subband There is only the normalization coefficient S of B, and the amount of transmitted information can be reduced.
又第６図に於いては、音声信号の各分割帯域の信号電力
を電力検出部４５により検出し、その検出信号に従って
代表帯域選択部４２で分割帯域の信号を選択し、この信
号を基に正規化係数部４１に於いて正規化係数を算出し
、この正規化係数を代表正規化係数として正規化部４３
に於いて各分割帯域の信号を正規化し、量子化部４４に
於いて量子化するもので、この場合も、補助情報として
は、１個の代表正規化係数となる。Further, in FIG. 6, the signal power of each divided band of the audio signal is detected by the power detection section 45, and the signal of the divided band is selected by the representative band selection section 42 according to the detected signal, and based on this signal. A normalization coefficient is calculated in the normalization coefficient section 41, and the normalization section 43 uses this normalization coefficient as a representative normalization coefficient.
The signals of each subband are normalized in the quantization section 44, and then quantized in the quantization section 44. In this case as well, the auxiliary information is one representative normalization coefficient.
前述の第３図に示す従来例に於いては、分割帯域Ｂ、〜
Ｂ３対応の正規化係数Ｓ、Ａ−Ｓ３をそれぞれ正規化の
補助情報として伝送するものであるから、低ビツトレー
トの伝送システムに適用した場合に、量子化した主情報
に割当てられるビット数を少なくしなければならない。In the conventional example shown in FIG. 3 mentioned above, the divided bands B, ~
Since the normalization coefficients S and A-S3 corresponding to B3 are transmitted as auxiliary information for normalization, when applied to a low bit rate transmission system, the number of bits allocated to the quantized main information can be reduced. There must be.
その為に、受信側の再生音声の品質が劣化する欠点があ
る。Therefore, there is a drawback that the quality of the reproduced audio on the receiving side deteriorates.
これに対して、前述の第５図及び第６図に示す従来例は
、１個の代表正規化係数を正規化の補助情報とするもの
であるから補助情報量が少なくなり、低ビツトレートの
伝送システムに適用した場合でも、量子化した主情報の
割当ビット数を特に少なくする必要はない。その場合、
代表正規化係数を３１とすると、各分割帯域８１〜Ｂ３
対応の信号Ｘ ｌ ”−Ｘ ３が正規化されて量子化部
に人力される信号ＱＩ＃Ｑ３は、
Ｑ １ ”’ Ｘ Ｉ／ ３ 。On the other hand, in the conventional example shown in FIGS. 5 and 6 described above, one representative normalization coefficient is used as auxiliary information for normalization, so the amount of auxiliary information is small, and low bit rate transmission is possible. Even when applied to a system, there is no need to particularly reduce the number of bits allocated to the quantized main information. In that case,
Assuming that the representative normalization coefficient is 31, each divided band 81 to B3
The signal QI#Q3 which is normalized from the corresponding signal Xl''-X3 and inputted to the quantization section is Q1'''XI/3.
Ｑ２＝ｘ２／５ｌＩ
Ｑ３＝Ｘ３／Ｓｌｌ
となる。このように、低域分割帯域Ｂ、或いは信号電力
最大の分割帯域の正規化係数を代表正規化係数としても
、他の分割帯域の最大値が大きい場合には、その分割帯
域の正規化係数を大きくしないと、その分割帯域の信号
がクリンプされることになる。例えば、第７図に示すよ
うに、各分割帯域Ｂ１〜Ｂ３の点線曲線ａ°に示す信号
の最大値をＸｌ、、、−Ｘ３．、、とじ、低域分別帯域
Ｂ、の正規化係数Ｓ、を代表正規化係数として正規化を
行うと、実線曲線ｂ“に示す信号となり、この信号が量
子化部に入力される。この場合に、中域と高域の帯域分
割信号の最大値Ｘ２．、、、Ｘ３．□が大きいと、正規
化マスクパターンＣ°以上のレベルの斜線を施した部分
はクリップされることになる。又信号電力最大の分割帯
域の正規化係数を代表正規化係数とした場合に於いても
、同様の問題が生じる。Q2=x2/5lI Q3=X3/Sll. In this way, even if the normalization coefficient of the low frequency subband B or the subband with the maximum signal power is used as the representative normalization coefficient, if the maximum value of the other subbands is large, the normalization coefficient of that subband is Otherwise, the signal in that subband will be crimped. For example, as shown in FIG. 7, the maximum value of the signal indicated by the dotted curve a° for each of the divided bands B1 to B3 is Xl, . . . -X3. When normalization is performed using the normalization coefficient S of the low-frequency classification band B as the representative normalization coefficient, a signal shown by the solid curve b is obtained, and this signal is input to the quantization section. If the maximum value X2.,...,X3.□ of the mid-range and high-range band division signals is large, the hatched portion with a level equal to or higher than the normalized mask pattern C° will be clipped. A similar problem occurs even when the normalization coefficient of the subband with the maximum signal power is used as the representative normalization coefficient.
このように、正規化信号が量子化部の量子化レベル以上
となると、正確に量子化された信号が伝送されないこと
になり、受信側の再生音声の品質が劣化する欠点があっ
た。As described above, when the normalized signal exceeds the quantization level of the quantizer, an accurately quantized signal is not transmitted, resulting in a disadvantage that the quality of reproduced audio on the receiving side deteriorates.
本発明は、補助情報を増加させることなく、量子化部へ
の入力のクリップを防止することを目的とするものであ
る。The present invention aims to prevent clipping of the input to the quantizer without increasing auxiliary information.
本発明の帯域分割正規化制御方式は、各分割帯域の最大
の正規化係数を代表正規化係数とするものであり、第１
図を参照して説明する。The band division normalization control method of the present invention uses the maximum normalization coefficient of each division band as the representative normalization coefficient, and the first
This will be explained with reference to the figures.
音声信号を複数の帯域に帯域分割部５により分割し、各
帯域分割信号の正規化係数を算出する正規化係数部１−
１〜１−ｎと、それらの正規化係数部１−１〜１−ｎに
於いて算出した正規化係数の中の最大値を検出する最大
値検出部２と、帯域分割信号をそれぞれ正規化する正規
化部３と、この正規化部３で正規された信号を量子化す
る量子化部４とを備え、最大値検出部２で検出した最大
値の正規化係数を代表正規化係数として、正規化部３に
於ける各分割帯域信号の正規化を行わせるものである。A normalization coefficient unit 1- which divides the audio signal into a plurality of bands by a band division unit 5 and calculates a normalization coefficient for each band division signal.
1 to 1-n, a maximum value detection unit 2 that detects the maximum value among the normalization coefficients calculated in the normalization coefficient units 1-1 to 1-n, and normalizes the band division signals, respectively. and a quantization unit 4 that quantizes the signal normalized by the normalization unit 3, and a normalization coefficient of the maximum value detected by the maximum value detection unit 2 as a representative normalization coefficient, This is to cause the normalization section 3 to normalize each divided band signal.
音声信号は帯域分割部５等に於いて、ＤＦＴ等により周
波数成分に展開され、複数の帯域に分割される。そして
、各分割帯域信号は正規化係数部１−１〜１−ｎに加え
られて正規化係数が算出される。最大値検出部２は、各
正規化係数部１−１〜ｌ−ｎに於いて算出された正規化
係数の最大のものを選択して代表正規化係数とし、正規
化部３に於いては、この代表正規化係数を用いて正規化
処理を行うものである。従って、代表正規化係数に対応
する分割帯域以外の分割帯域信号の最大値がクリップさ
れることなく、量子化部４に入力されて量子化されるこ
とになる。The audio signal is expanded into frequency components by DFT or the like in the band dividing section 5 or the like, and is divided into a plurality of bands. Then, each divided band signal is added to normalization coefficient sections 1-1 to 1-n to calculate normalization coefficients. The maximum value detection unit 2 selects the largest one of the normalization coefficients calculated in each normalization coefficient unit 1-1 to l-n as a representative normalization coefficient, and the normalization unit 3 , normalization processing is performed using this representative normalization coefficient. Therefore, the maximum value of the subband signals other than the subband corresponding to the representative normalization coefficient is input to the quantization unit 4 and quantized without being clipped.
以下図面を参照して本発明の実施例について詳細に説明
する。Embodiments of the present invention will be described in detail below with reference to the drawings.
第２図は本発明の実施例のブロック図であり、低域Ｂ
ｌ ＋ 中域Ｂ２．高域Ｂ３の帯域に分割した場合を示
し、１１−１〜１１−３は正規化係数部、１２は最大値
検出部、１３は正規化部、１４は量子化部、１５は帯域
分割部、１６は多重化部である。FIG. 2 is a block diagram of an embodiment of the present invention.
l + midrange B2. This shows the case where the high frequency band is divided into B3 bands, 11-1 to 11-3 are normalization coefficient sections, 12 is a maximum value detection section, 13 is a normalization section, 14 is a quantization section, 15 is a band division section, 16 is a multiplexing section.
帯域分割部１５は、前述のようにＤＦＴ等により音声信
号を周波数成分に展開し、低域Ｂ ｌ ＋中域Ｂ２．高
域Ｂ３の帯域に分割するものであり、又正規化係数部１
１−１〜１１−３は、分割帯域Ｂ、＃Ｂ３の信号からそ
れぞれ正規化係数８１〜Ｓ３を算出して最大値検出部１
２に加える。最大値検出部１２は、正規化係数Ｓ、％Ｓ
３の中の最′大値を検出して、代表正規化係数Ｓ、とし
て正規化部１３及び多重化部１６に加える。The band dividing unit 15 develops the audio signal into frequency components by DFT or the like as described above, and divides the audio signal into frequency components such as low band B l + middle band B2 . It is divided into high frequency band B3, and normalization coefficient section 1
1-1 to 11-3 calculate normalization coefficients 81 to S3 from the signals of divided bands B and #B3, respectively, and calculate the maximum value detection unit 1.
Add to 2. The maximum value detection unit 12 has a normalization coefficient S, %S
3 is detected and added to the normalization section 13 and the multiplexing section 16 as a representative normalization coefficient S.
正規化部１３は、代表正規化係数Ｓ、により各分割帯域
信号を正規化して量子化部１４に加え、それぞれの分割
帯域に割当てられたビット数により量子化し、多重化部
１６に於いて分割帯域対応の量子化信号を多重化し、又
代表正規化係数Ｓ。The normalization unit 13 normalizes each subband signal using a representative normalization coefficient S, and adds the normalized signal to the quantization unit 14, where it is quantized according to the number of bits assigned to each subband. Quantized signals corresponding to bands are multiplexed, and a representative normalization coefficient S is used.
を正規化の補助情報として送出する。is sent as auxiliary information for normalization.
最大値の正規化係数を代表正規化係数Ｓ、とじて正規化
を行うものであるから、第７図に示すような他の分割帯
域の信号の最大値が、正規化後に於いても正規化マスク
パターンを超えるようなことはなくなる。即ち、クリッ
プを確実に防止することができる。従って、受信側の再
生音声の品質劣化を防止することができる。Since normalization is performed by dividing the normalization coefficient of the maximum value into the representative normalization coefficient S, even if the maximum value of the signal of other divided bands as shown in Fig. 7 is normalized even after normalization. It will no longer exceed the mask pattern. That is, clipping can be reliably prevented. Therefore, it is possible to prevent quality deterioration of the reproduced audio on the receiving side.
前述の実施例は、音声信号を３個の帯域に分割した場合
を示すが、更に多数の帯域に分割することも勿論可能で
ある。Although the above embodiment shows the case where the audio signal is divided into three bands, it is of course possible to divide the audio signal into even more bands.
以上説明したように、本発明は、音声信号を複数の帯域
に分割し、且つ正規化係数部１−１〜１−ｎに於いて帯
域分割信号対応の正規化係数を算出し、その中の最大値
を代表正規化係数として正規化部３に於いて正規化する
ものであり、従って、量子化部４へ入力される帯域分割
信号の正規化信号がクリップされることはなくなり、且
つ正規化の補助情報としても、代表正規化係数のみであ
るから、補助情報を増加させることなく、クリップを防
止して、再生音声の品質劣化を阻止することが可能とな
る利点がある。As explained above, the present invention divides an audio signal into a plurality of bands, calculates normalization coefficients corresponding to the band-divided signals in the normalization coefficient sections 1-1 to 1-n, and calculates normalization coefficients corresponding to the band-divided signals. Normalization is performed in the normalization unit 3 using the maximum value as a representative normalization coefficient. Therefore, the normalized signal of the band division signal input to the quantization unit 4 will not be clipped, and the normalization Since the auxiliary information is only the representative normalization coefficient, there is an advantage that clipping can be prevented and quality deterioration of the reproduced audio can be prevented without increasing the auxiliary information.
第１図は本発明の原理説明図、第２図は本発明の実施例
のブロック図、第３図は従来例のブロック図、第４図は
正規化動作の説明図、第５図及び第６図は従来例の要部
ブロック図、第７図は従来例の代表正規化係数による正
規化動作説明図である。
１−１〜ｌ−ｎは正規化係数部、２は最大値検出部、３
は正規化部、４は量子化部、５は帯域分割部である。Fig. 1 is an explanatory diagram of the principle of the present invention, Fig. 2 is a block diagram of an embodiment of the invention, Fig. 3 is a block diagram of a conventional example, Fig. 4 is an explanatory diagram of normalization operation, Figs. FIG. 6 is a block diagram of the main part of the conventional example, and FIG. 7 is an explanatory diagram of normalization operation using representative normalization coefficients of the conventional example. 1-1 to l-n are normalization coefficient sections, 2 is maximum value detection section, 3
4 is a normalization section, 4 is a quantization section, and 5 is a band division section.
Claims (1)
化係数を算出する正規化係数部（１−１〜１−ｎ）と、 該正規化係数部（１−１〜１−ｎ）に於いて算出した正
規化係数の中の最大値を検出する最大値検出部（２）と
、 前記帯域分割信号をそれぞれ正規化する正規化部（３）
と、 該正規化部（３）により正規化された信号を量子化する
量子化部（４）とを備え、 前記最大値検出部（２）により検出された最大値の正規
化係数を代表正規化係数として、前記正規化部（３）に
より各帯域分割信号の正規化を行わせる ことを特徴とする帯域分割正規化制御方式。[Scope of Claims] A normalization coefficient unit (1-1 to 1-n) that divides an audio signal into a plurality of bands and calculates a normalization coefficient for each band-divided signal; a maximum value detection unit (2) that detects the maximum value among the normalization coefficients calculated in steps 1 to 1-n); and a normalization unit (3) that normalizes each of the band-divided signals.
and a quantization unit (4) that quantizes the signal normalized by the normalization unit (3), and converts the normalization coefficient of the maximum value detected by the maximum value detection unit (2) into a representative normalization coefficient. A band division normalization control method characterized in that each band division signal is normalized by the normalization unit (3) as a normalization coefficient.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| JP63306862A JPH02153639A (en) | 1988-12-06 | 1988-12-06 | System for controlling division and normalization of band |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| JP63306862A JPH02153639A (en) | 1988-12-06 | 1988-12-06 | System for controlling division and normalization of band |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| JPH02153639A true JPH02153639A (en) | 1990-06-13 |
Family
ID=17962150
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| JP63306862A Pending JPH02153639A (en) | 1988-12-06 | 1988-12-06 | System for controlling division and normalization of band |
Country Status (1)
| Country | Link |
|---|---|
| JP (1) | JPH02153639A (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JPH03139923A (en) * | 1989-10-25 | 1991-06-14 | Sony Corp | Highly efficient encoder for digital data |
| JP2008309875A (en) * | 2007-06-12 | 2008-12-25 | Casio Comput Co Ltd | Speech encoding device, speech decoding device, speech encoding method, speech decoding method, and program |
| US7933417B2 (en) | 2000-12-22 | 2011-04-26 | Sony Corporation | Encoding device and decoding device |
-
1988
- 1988-12-06 JP JP63306862A patent/JPH02153639A/en active Pending
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JPH03139923A (en) * | 1989-10-25 | 1991-06-14 | Sony Corp | Highly efficient encoder for digital data |
| US7933417B2 (en) | 2000-12-22 | 2011-04-26 | Sony Corporation | Encoding device and decoding device |
| JP2008309875A (en) * | 2007-06-12 | 2008-12-25 | Casio Comput Co Ltd | Speech encoding device, speech decoding device, speech encoding method, speech decoding method, and program |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| KR101026597B1 (en) | Audio encoding device, audio decoding device, audio encoding method, audio decoding method, and information recording medium | |
| CN1032102C (en) | Coding method for digital audio signal | |
| RU2670797C9 (en) | Method and apparatus for generating from a coefficient domain representation of hoa signals a mixed spatial/coefficient domain representation of said hoa signals | |
| JP2841765B2 (en) | Adaptive bit allocation method and apparatus | |
| JPH0591061A (en) | Audio signal processing meth0d | |
| JPH04304031A (en) | Digital signal coder | |
| JPH03256411A (en) | High efficient encoder for digital data | |
| JPH05248972A (en) | Audio signal processing method | |
| US5946652A (en) | Methods for non-linearly quantizing and non-linearly dequantizing an information signal using off-center decision levels | |
| JP2004309921A (en) | Device, method, and program for encoding | |
| JP2000151413A (en) | Method for allocating adaptive dynamic variable bit in audio encoding | |
| JPH02153639A (en) | System for controlling division and normalization of band | |
| JPH0969781A (en) | Audio data encoding device | |
| JPH08307281A (en) | Nonlinear quantization method and nonlinear inverse quantization method | |
| JP3336619B2 (en) | Signal processing device | |
| JPH04215320A (en) | Digital signal coder | |
| JPH08123488A (en) | High-efficiency encoding method, high-efficiency code recording method, high-efficiency code transmitting method, high-efficiency encoding device, and high-efficiency code decoding method | |
| JP3089692B2 (en) | Highly efficient digital data encoding method. | |
| US5875424A (en) | Encoding system and decoding system for audio signals including pulse quantization | |
| JP3108259B2 (en) | Band division coding device | |
| EP1176743B1 (en) | Methods for non-linearly quantizing and dequantizing an information signal | |
| JPH08167851A (en) | Highly efficient encoding method, device therefor and transmitting medium | |
| JPH0750589A (en) | Sub-band coding device | |
| JPH0519797A (en) | Quantizing method | |
| JP3033157B2 (en) | Digital signal coding device |