JP3227945B2 - Encoding device - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an encoding apparatus for reducing a bit rate used in a stereo or so-called multi-surround sound system such as a movie film projection system, a video tape recorder, a video disk player, and the like.

[0002]

2. Description of the Related Art There are various methods and apparatuses for high-efficiency encoding of signals such as audio or voice. For example, a time domain audio signal or the like is divided into blocks for each unit time, and a time axis signal for each block is obtained. Is converted into a signal on the frequency axis (orthogonal transform), divided into a plurality of frequency bands, and a so-called transform coding method, which is a block frequency band division method for encoding each band, an audio signal in the time domain, and the like. Is a non-blocking frequency band division scheme (sub-band coding: S
BC) system and the like. Further, a high-efficiency coding method and apparatus combining the above-mentioned band division coding and transform coding are also considered. In this case, for example, after performing band division by the above-mentioned band division coding method, Then, the signal for each band is orthogonally transformed into a signal in the frequency domain by the above-described transform coding method, and encoding is performed for each band subjected to the orthogonal transformation.

Here, as a band division filter used in the above-mentioned band division coding system, for example, QMF (Q
uadrature Mirror filter).
1976R.E.Crochiere Digital coding of speech in sub
bands Bell Syst.Tech. J.Vol.55, No.8 1976. Also, ICASSP 83, BOSTON PolyphaseQuad
rature filters-A new subband coding technique Jose
ph H. Rothweiler describes an equal bandwidth filter dividing method and apparatus such as a polyphase quadrature filter.

As the above-described orthogonal transform, for example, an input audio signal is divided into blocks in a predetermined unit time (frame), and a fast Fourier transform (FF) is performed for each block.
T), cosine transform (DCT), modified DCT
There is an orthogonal transformation that transforms a time axis into a frequency axis by performing transformation (MDCT) or the like. Regarding the above MDCT, ICASSP 1987 Subband / Transform Coding Using Fil
ter Bank Designs Basedon Time Domain Aliasing Canc
ellation JPPrincen ABBradley Univ. of Surrey R
oyal Melbourne Inst. of Tech.

Further, as a frequency division width when quantizing each frequency component divided into frequency bands, for example, there is a band division in consideration of human auditory characteristics. That is, an audio signal may be divided into a plurality of bands (for example, 25 bands) with a bandwidth generally called a critical band (critical band) such that the bandwidth becomes wider as the band becomes higher. When encoding data for each band at this time, predetermined bits are allocated to each band, or encoding is performed by adaptive bit allocation to each band. For example, when the MDCT coefficient data obtained by the MDCT processing is encoded by the bit allocation, the MDCT coefficient for each band obtained by the MDCT processing for each block is used.
Coding is performed on the coefficient data with an adaptive number of allocated bits.

The following two methods and devices are known as the above-mentioned bit allocation method and a device therefor. IEEE Tra
nsactions of Accoustics, Speech, and Signal Processi
In ng, vol. ASSP-25, No. 4, August 1977, bits are allocated based on the magnitude of the signal in each band. Also IC
ASSP 1980 The critical band coder--digital encodin
g ofthe perceptual requirements of the auditory sy
stem MA Kransner MIT describes a method and apparatus for obtaining a required signal-to-noise ratio for each band and performing fixed bit allocation by using auditory masking.

[0007]

However, these bit allocation techniques are intended to perform reproduction (decoding) at a certain bit rate on the reproduction side (decoding side). When the reproduction is performed at a bit rate lower than the rate, a significant deterioration in sound quality is caused. That is, decoding is performed using a bit rate lower than the bit rate used at the time of encoding. For example, when a part of the bits after the encoding process is diverted for another data transfer on the encoding side, the decoding is performed on the reproducing side. Because playback will be performed at a bit rate lower than the bit rate at the time of encoding,
In the above-mentioned known bit allocation technique, which is expected to reproduce at the encoding bit rate on the reproducing side,
At the time of reproduction (decoding), remarkable deterioration of sound quality is caused.

Further, for example, in a case where a reproducing apparatus for reproducing at a low bit rate is already used, even if an attempt is made to introduce a system of good sound quality using a higher bit rate, the already used low bit Good playback cannot be performed with a playback device that performs playback at a rate.

That is, in the conventional bit allocation technology, there is no backward compatibility.

SUMMARY OF THE INVENTION It is an object of the present invention to provide an encoding apparatus which can minimize such sound quality deterioration and has backward compatibility.

Further, in the case where information obtained by encoding signals such as voice and audio is recorded on a storage medium using a storage device such as a so-called IC card, the storage device is expensive. It is desired that recording be performed for a longer time, and that sound quality degradation be minimized. Therefore, another object of the present invention is to perform recording on a storage medium using an expensive storage device, for example, when the recording time is to be extended from the initial setting for a long time recording, or when recording is performed or during recording. It is an object of the present invention to provide an encoding apparatus capable of appropriately reducing the bit rate of the encoded information to extend the recording time and minimizing the sound quality deterioration at this time.

[0012]

[0013]

SUMMARY OF THE INVENTION The present invention has been proposed to achieve the above-mentioned object, and an encoding apparatus according to the present invention is capable of converting a time domain sample or a frequency domain sample of an input signal into predetermined bits. In an encoding device that quantizes by a number, a first quantization unit that quantizes the time domain sample or the frequency domain sample, and a quantization error calculation unit that calculates a quantization error of the first quantization unit. ,
A second quantization unit that further quantizes the quantization error; and an encoding unit that encodes an output of the first quantization unit and an output of the second quantization unit.

Further, in the encoding apparatus of the present invention, the second
And the output bit rates of the first and second quantization units are set to a constant bit rate in fixed time units. In these cases, the encoding device includes a first block floating unit that blocks the time domain sample or the frequency domain sample for each of a plurality of samples and performs a block floating process, and a block floating unit that outputs an output of the quantization error calculation unit. A second block floating means for processing, wherein the second block floating means obtains a scale factor for the block floating based on the scale factor in the first block floating means, The quantizing means quantizes the output of the first block floating means, and the second quantizing means quantizes the output of the second block floating means.

In the encoding apparatus of the present invention, the second
The block floating means determines the scale factor for the block floating based on the scale factor in the first block floating means and the word length in the first quantization means.

In the above case, with respect to the sample data in the small block subdivided in time and frequency, quantization having the same block floating and word length is performed in the small block. In addition, in order to obtain a sample in a small block subdivided with respect to the time and frequency, after performing a deblocking frequency analysis such as a filter, an output of the deblocking frequency analysis such as the filter is subjected to an orthogonal transformation or the like. To analyze the blocking frequency. At this time, the fact that the frequency bandwidth of the non-blocking frequency analysis is the same in at least the two lowest bands is useful in reducing costs. Further, the frequency bandwidth of the non-blocking frequency analysis is wider at least in the highest band and high band,
It is important in utilizing the effect of hearing based on the critical band. Further, in the block frequency analysis, an optimal process corresponding to the time characteristic of the input signal can be performed by adaptively changing the block size according to the time characteristic of the input signal. Here, the change of the block size is performed independently for each of the output bands of at least two of the non-blocking frequency analysis, so that optimal processing is performed independently for each band component while preventing mutual interference between frequency components. It is effective on.

Further, since the bit allocation amount given to each channel is determined by the scale factor or the sample maximum value of each channel, the calculation is simple, so that it is effective in reducing the number of calculations. In addition to this, changing the bit allocation amount given to each channel according to the temporal change of the amplitude information represented by the scale factor of each channel is also useful for lowering the bit rate.

Further, the encoding apparatus according to the present invention comprises a first quantization means for quantizing the time domain sample or the frequency domain sample, and a quantization error for calculating a quantization error of the first quantization means. Calculation means, second quantization means for further quantizing the quantization error, and an output sample of the first quantization means and an output sample of the second quantization means in one sync block. And recording means for recording separately. Further, the encoding apparatus of the present invention includes a first quantization unit for quantizing the time domain sample or the frequency domain sample, and a quantization error calculation unit for calculating a quantization error of the first quantization unit. , A second quantizing means for further quantizing the quantization error, and an output sample of the first quantizing means and an output sample of the second quantizing means in one sync block. Recording means for recording alternately in order of frequency or frequency.

[0019]

According to the present invention, when a time-domain sample or a frequency-domain sample of an input signal is quantized with a predetermined number of bits, the time-domain sample or the frequency-domain sample is quantized by the first quantization means. The quantization error of the first quantization means calculated by the quantization error calculation means is further quantized by the second quantization means,
The output of the first quantization means and the output of the second quantization means are encoded by the encoding means.

The output bit rate of the second quantization means may be set to a constant bit rate in a fixed time unit, or the output bit rates of the first quantization means and the second quantization means may be set in a fixed time unit. Setting a constant bit rate is effective in simplifying a recording method on a recording medium such as a disk and a tape.

[0021] In these cases, the encoding apparatus outputs the output of the first block floating means for performing block floating processing by blocking the time domain sample or the frequency domain sample for each of a plurality of samples, and the output of the quantization error calculating means. Block floating processing second
Block floating means, wherein the second block floating means obtains a scale factor for the block floating based on the scale factor in the first block floating means, and wherein the first quantization means Quantizing the output of the first block floating means and quantizing the output of the second block floating means by the second quantizing means is effective in increasing the coding efficiency.

Further, in order to obtain a sample in a small block subdivided with respect to time and frequency, after performing a deblocking frequency analysis such as a filter, an output of the deblocking frequency analysis such as the filter is subjected to an orthogonal transformation. By performing a block frequency analysis such as that described above, it is possible to generate quantization noise in consideration of auditory masking in the time domain and the frequency domain, and it is possible to obtain a frequency analysis that is preferable for hearing. At this time, the fact that the frequency bandwidth of the non-blocking frequency analysis is the same in at least the two lowest bands is useful in reducing costs. In addition, by making the frequency bandwidth of the non-blocking frequency analysis wider at least in the highest band and in the higher band, the effect of hearing based on the critical band can be used efficiently. In this block frequency analysis, an optimal process corresponding to the time characteristic of the input signal can be performed by adaptively changing the block size according to the time characteristic of the input signal. In addition, performing the block size change independently for each of the at least two output bands of the deblocking frequency analysis is effective in preventing the mutual interference between the frequency components and performing the optimum processing independently for each band component. It is a target.

Also, a first quantization means for quantizing the time domain sample or the frequency domain sample, a quantization error calculation means for calculating a quantization error of the first quantization means, And the output sample of the first quantization unit and the output sample of the second quantization unit are separately recorded in one sync block. The provision of the recording means is effective in that a bit string portion to be removed when reproducing at a reduced bit rate can be collectively removed.

Also, a first quantization means for quantizing the time domain samples or the frequency domain samples, a quantization error calculation means for calculating a quantization error of the first quantization means, And the output sample of the first quantization means and the output sample of the second quantization means are alternately arranged in time order or frequency order in one sync block. Is effective in that the bit string portion to be removed can be collectively removed by limiting the frequency band when reproducing at a reduced bit rate.

[0025]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS An example of a high-efficiency coding apparatus (encoder) as an embodiment of the coding apparatus according to the present invention and an example of a corresponding high-efficiency code decoding apparatus (decoder) will be described below with reference to the drawings. .

In this embodiment, an input digital signal such as an audio PCM signal is subjected to band division coding (SBC), adaptive conversion coding (ATC), and adaptive bit allocation (AP).
High-efficiency coding is performed using each technique of C-AB). This technique will be described with reference to FIG.

In the specific high-efficiency coding apparatus of this embodiment shown in FIG. 1, the input digital signal is obtained by dividing the input digital signal into a plurality of frequency bands by using a filter or the like and performing orthogonal transform for each frequency band. The spectrum data on the frequency axis is adaptively bit-distributed and encoded for each so-called critical bandwidth (critical band) in consideration of human auditory characteristics described later. At this time, in the high band, a band obtained by further dividing the critical bandwidth is used. Of course, a non-blocking frequency division width by a filter or the like may be an equal division width.

Further, in the embodiment of the present invention, before the orthogonal transform, the block size (block length) is adaptively changed according to the input signal, and the critical bandwidth (critical band) in the unit of a critical band or in a high band.
Floating processing is performed in blocks that are further subdivided. Note that this critical band is a frequency band divided in consideration of human auditory characteristics, and the narrow band noise of the same strength near the frequency of a certain pure tone causes the noise of the pure tone to be masked. It is the band that you have. The bandwidth of this critical band increases as the frequency band increases. For example, the entire frequency band of 0 to 20 kHz is divided into, for example, 25 critical bands.

That is, in FIG. 1, an audio PCM signal of, for example, 0 to 22 kHz is supplied to the input terminal 10. This input signal is divided into a band of 0 to 11 kHz and a band of 11 to 22 kHz by a band division filter 11 such as a so-called QMF, for example.
The band signals are also divided by a band division filter 12, such as a so-called QMF, into a band of 0 to 5.5 kHz and a band of 5.5 to 11 k.
Hz band.

11k from the band division filter 11
The signal in the 22 kHz band is represented by M which is an example of an orthogonal transformation circuit.
DCT (Modified Discrete Cosine Transform) circuit 1
5.5k from the band division filter 12
The signal in the ~ 11 kHz band is sent to the MDCT circuit 14,
The signals in the 0 to 5.5 kHz band from the band division filter 12 are sent to the MDCT circuit 15 to be subjected to the MDCT processing. Note that each MDCT circuit 13,
In blocks 14 and 15, the block decision circuit 1 provided for each band is used.
MDCT processing is performed based on the block sizes determined by 9, 20, and 21.

Here, the block decision circuits 19, 2
0, 21 each MDCT circuit 13, 14,
Specific examples of the block size at 15 are shown in FIGS. FIG. 2A shows a case where the orthogonal transform block size is long (orthogonal transform block size in long mode).
2B shows the case where the orthogonal transform block size is short (orthogonal transform block size in the short mode).

In the specific example of FIG. 2, each of the three filter outputs has two orthogonal transform block sizes. That is, for a signal in the 0-5.5 kHz band on the low band side and a signal in the 5.5 kHz-11 kHz band in the middle band, in the case of a long block length (A in FIG. 2), the number of samples in one block is reduced. When a short block is selected (B in FIG. 2), the number of samples in one block is set to every 32 samples. On the other hand, for a signal in the 11 kHz to 22 kHz band on the high frequency side, in the case of a long block length (A in FIG. 2), the number of samples in one block is 256, and when a short block is selected ( FIG. 2B) shows that the number of samples in one block is 32.
It is a block for each sample. When a short block is selected in this way, the number of samples of the orthogonal transform block in each band is kept the same, and the higher the frequency, the higher the time resolution and the type of window used for blocking.

The above-mentioned block decision circuits 19, 20,
Information indicating the block size determined in 21 is sent to adaptive bit allocation encoding circuits 16, 17, and 18, which will be described later, and output from output terminals 23, 25, and 27.

Referring again to FIG. 1, each MDCT circuit 13,
The spectral data or the MDCT coefficient data in the frequency domain obtained by performing the MDCT processing in 14 and 15 are grouped in a so-called critical band (critical band) or a band obtained by further dividing the critical band in a high band, and are adapted to an adaptive bit allocation code. To the conversion circuits 16, 17 and 18.

The adaptive bit allocation encoding circuits 16, 17, 1
8, each spectrum data (or MDCT coefficient data) according to the block size information and the critical band (critical band) or the number of bits allocated to each band obtained by further dividing the critical band in the high band.
Is requantized (normalized and quantized).

Each of these adaptive bit allocation encoding circuits 16,
The data encoded by 17 and 18 is output to output terminal 2
It is taken out via 2, 24, 26. Also, in the adaptive bit allocation encoding circuits 16, 17, and 18, a scale factor indicating what kind of signal size has been normalized and a bit length indicating what kind of bit length has been used for quantization. Information is also obtained, and these are also output from the output terminals 22, 24, 26 at the same time.

Each MDCT circuit 13 in FIG.
From the outputs of 14 and 15, calculate the energy for each band obtained by further dividing the critical band in the critical band or the high band, for example, the root mean square of each amplitude value in the band. And so on. Of course, the scale factor itself may be used for subsequent bit allocation. In this case, a new calculation for energy calculation is not required, so that the hardware scale is saved. Further, instead of the energy for each band, a peak value, an average value, or the like of the amplitude values can be used.

Next, a specific bit allocation method in the adaptive bit allocation circuit for performing the above-mentioned bit allocation will be described using a bit allocation strategy shown in FIG.

In the present embodiment, based on the total bit allocation in step ST1, first, a basic bit allocation of 128 kbps per channel (step ST2), and secondly, 64 kbps
and ps additional bit allocation (step ST3).

Of these, the basic bit allocation is further divided into bit allocation (1) (step ST4) and bit allocation (2) (step ST5).

First, from step ST1 to step ST2
A method of allocating the basic bits to the above will be described. Here, bit allocation is performed adaptively by looking at the frequency distribution of the scale factor.

First, the amount of bits to be used for bit allocation (1) is determined. For that purpose, tonality information is used among the spectrum information of the signal information. As the tonality index, a value obtained by dividing the sum of the absolute values of the differences between adjacent values of the signal spectrum by the number of signal spectra is used. As a simpler index, as shown in FIG. 4, an average value of a difference between adjacent scale factor indexes in a scale factor for each block for so-called block floating can be used. This scale factor index roughly corresponds to the logarithmic value of the scale factor.

In the embodiment, the bit amount to be used for the bit allocation (1) is set to a maximum of 8 in correspondence with the value representing the tonality.
0 kbps and a minimum of 10 kbps are set. This tonality calculation is performed as in the following equation.

T = (1 / (WLmax * (N-1)) (ΣABS (SFn-SFn-1)) WLmax: Maximum word length = 16 SFn: Scale factor index corresponding to the logarithm of the approximate peak value N: Block floating band number N: Number of block floating bands

The tonality index T and the amount of bit allocation (1) thus obtained are associated with each other as shown in FIG. The bit allocation (1) here is performed in the frequency and time domains depending on the scale factor.

After the amount of bits used in bit allocation (1) is determined in this way, the allocation of bits not used in bit allocation (1), that is, the bit allocation
Go to (2). Here, various kinds of bit allocation are performed, but two examples are shown below.

First, a uniform distribution is performed for all sample values. FIG. 6 shows the quantization noise spectrum for the bit allocation in this case (the noise spectrum of the uniform allocation of the bit allocation (2)). According to this, the noise level can be reduced uniformly in all frequency bands.

Second, bit allocation is performed to obtain an audible effect having dependency on the frequency spectrum and level of the signal information. FIG. 7 shows an example of a quantization noise spectrum for the bit allocation in this case (a noise spectrum by the bit allocation for obtaining an auditory effect having dependency on the frequency spectrum and the level of the signal information). In this example, bit allocation depending on the spectrum of the information signal is performed. In particular, bit allocation is performed with a weight on the low side of the spectrum of the information signal, and masking on the low side which occurs compared to the high side. It compensates for the reduced effect. This is based on the asymmetricity of a masking curve that emphasizes the lower side of the spectrum in consideration of masking between adjacent critical bands.

FIG. 8 is a diagram showing bit allocation (allocation) at the time of uniform allocation of bit allocation (2), and shows bit allocation corresponding to FIG. FIG. 9 is a diagram showing bit allocation for obtaining an auditory effect having a dependency corresponding to the frequency spectrum and level of signal information, and shows the bit allocation corresponding to FIG. In FIGS. 6 and 7, S denotes a signal spectrum, NL1 denotes a noise level by bit allocation (1), and NL2 denotes a noise level by bit allocation (2). AQ1 in FIGS. 8 and 9 is bit allocation
In the figure, AQ2 indicates the bit amount of bit allocation (2).

Next, another method of basic bit allocation will be described. The operation of the adaptive bit distribution circuit in this case will be described with reference to FIG. 10. The magnitude of the MDCT coefficient is obtained for each block, and the MDCT coefficient is supplied to the input terminal 801. The MDCT coefficient supplied to the input terminal 801 is given to the energy calculation circuit 803 for each band. Energy calculation circuit 803 for each band
Then, in the critical band or the high band, the signal energy for each band obtained by further subdividing the critical band is calculated. Energy calculation circuit 8 for each band
The energy for each band calculated in 03 is
This is supplied to the energy-dependent bit distribution circuit 804.

In the energy-dependent bit distribution circuit 804,
Using the total available bits from the total available bits generation circuit 802, in this embodiment, a certain ratio of 128 Kbps, a bit allocation is performed to generate white quantization noise. At this time, the higher the tonality of the input signal, that is, the greater the unevenness of the spectrum of the input signal, the greater the ratio of the bit amount to the above 128 Kbps. In order to detect the unevenness of the spectrum of the input signal, the sum of the absolute values of the differences between the block floating coefficients of adjacent blocks is used as an index. Then, for the determined available bit amount, bit allocation is performed in proportion to the logarithmic value of the energy of each band.

The bit allocation calculation circuit 805 depending on the permissible noise level of the hearing first calculates the permissible noise amount for each critical band in consideration of the so-called masking effect based on the spectrum data divided for each critical band. Are allocated by subtracting the energy-dependent bits from the total available bits so as to give the permissible noise spectrum. The energy-dependent bits obtained in this way and the bits depending on the permissible noise level are added, and the adaptive bit allocation coding circuits 16 and 1 shown in FIG.
According to 7 and 18, each spectral data (or MDCT coefficient data) is re-quantized in accordance with the number of bits allocated to each critical band or a band obtained by further dividing the critical band into a plurality of bands in the high band. ing. The data encoded in this way is taken out via the output terminals 22, 24, 26 of FIG.

The permissible noise spectrum calculation circuit in the permissible noise spectrum dependent bit allocation circuit 805 will be described in more detail.
The MDCT coefficients obtained in 4 and 15 are given to an allowable noise spectrum calculation circuit in the bit allocation circuit 805.

FIG. 11 is a block circuit diagram showing a schematic configuration of one specific example in which the allowable noise spectrum calculating circuit is collectively described. In FIG. 11, spectrum data in the frequency domain is supplied to the input terminal 521 from the MDCT circuits 13, 14, and 15.

The input data in the frequency domain is sent to the energy calculation circuit 522 for each band, and the energy for each critical band (critical band) is calculated, for example, by summing the square of each amplitude value within the band. Required. Instead of the energy for each band, a peak value or an average value of the amplitude value may be used. As an output from the energy calculation circuit 522, for example, the spectrum of the sum value of each band is generally called a bark spectrum. FIG. 12 shows such a bark spectrum SB for each critical band. However, in FIG. 12, the number of the critical bands is represented by 12 bands (B1 to B12) to simplify the illustration.

Here, in order to consider the influence of the bark spectrum SB on so-called masking, a convolution process is performed such that the bark spectrum SB is multiplied by a predetermined weighting function and added. Therefore, the output of the energy calculation circuit 522 for each band, that is, each value of the bark spectrum SB is sent to the convolution filter circuit 523. The convolution filter circuit 523
Includes, for example, a plurality of delay elements for sequentially delaying input data, a plurality of multipliers (for example, 25 multipliers corresponding to each band) for multiplying an output from these delay elements by a filter coefficient (weighting function). , And a sum adder for summing the outputs of the multipliers.

The above-mentioned masking is a phenomenon in which a certain signal masks another signal and becomes inaudible due to human auditory characteristics. The masking effect includes a time domain of an audio signal in a time domain. There are an axis masking effect and a simultaneous masking effect by a frequency domain signal. Due to these masking effects, even if noise is present in the masked portion, this noise will not be heard. For this reason, in an actual audio signal, the noise within the masked range is regarded as acceptable noise.

A specific example of the multiplication coefficient (filter coefficient) of each multiplier of the convolution filter circuit 523 is as follows. When the coefficient of the multiplier M corresponding to an arbitrary band is set to 1, the multiplier M −1, the coefficient 0.15, the multiplier M-2, the coefficient 0.0019, and the multiplier M-3, the coefficient 0.00000.
086, a coefficient 0.4 by a multiplier M + 1, and a multiplier M + 2
By multiplying the output of each delay element by the coefficient 0.06 by the multiplier M + 3 and the coefficient 0.007 by the multiplier M + 3, the convolution process of the bark spectrum SB is performed. However, M is 1
It is an arbitrary integer of 25.

Next, the output of the convolution filter circuit 523 is sent to a subtractor 524. The subtracter 524 calculates a level α corresponding to an allowable noise level described later in the convolved area. The level α corresponding to the permissible noise level (permissible noise level) is, as described later, a level at which the permissible noise level of each critical band is obtained by performing inverse convolution processing. is there.

Here, an allowance function (a function expressing a masking level) for obtaining the level α is supplied to the subtractor 524. The level α is controlled by increasing or decreasing the allowable function. The permissible function is a (n-ai) function generation circuit 52 described below.
5.

That is, the level α corresponding to the allowable noise level can be obtained by the following equation, where i is a number sequentially given from the lower band of the critical band. α = S− (n−ai) In this equation, n and a are constants, a> 0, and S is the intensity of the convolution-processed Bark spectrum, where (n-ai) is an allowable function. As an example, n = 38 and a = -0.5 can be used.

Thus, the level α is obtained, and this data is transmitted to the divider 526. In the divider 526, the level α in the convolved area is inversely convolved. Therefore, by performing the inverse convolution processing, a masking threshold can be obtained from the level α. That is, this masking threshold becomes the allowable noise spectrum. Note that the above inverse convolution process requires a complicated operation, but in this embodiment, inverse convolution is performed using a simplified divider 526.

Next, the masking threshold is transmitted to the subtractor 528 via the synthesis circuit 527. Here, the output from the energy detection circuit 522 for each band, that is, the above-described bark spectrum SB is supplied to the subtracter 528 via the delay circuit 529. Accordingly, the subtraction operation of the masking threshold and the bark spectrum SB is performed by the subtractor 528, so that the bark spectrum SB is converted to the masking threshold M as shown in FIG.
The level below the level indicated by the level S is masked. Note that the delay circuit 529 is connected to the synthesis circuit 52.
7 is provided to delay the bark spectrum SB from the energy detection circuit 522 in consideration of the amount of delay in each circuit before 7.

The output from the subtracter 528 is taken out via an allowable noise correction circuit 530 and an output terminal 531. For example, a ROM in which distribution bit number information is stored in advance is provided.
Etc. (not shown). This ROM or the like distributes each band in accordance with the output (the level of the difference between the energy of each band and the output of the noise level setting means) obtained from the subtraction circuit 528 via the allowable noise correction circuit 530. Output bit number information.

In this way, the energy-dependent bits and the bits depending on the permissible noise level are added, and the information on the number of allocated bits is added to the adaptive bit allocation coding circuits 16 and 1.
7 and 18 so that the MDCT circuit 13
Each spectrum data in the frequency domain from 14 and 15 is quantized by the number of bits allocated to each band.

That is, in summary, in the adaptive bit allocation encoding circuits 16, 17, and 18, the critical band is further divided into a plurality of bands in each of the critical bands (for each critical band) or in the high band. The spectrum data for each band is quantized by the number of bits allocated according to the level of the difference between the energy or peak value of the band and the output of the noise level setting means.

By the way, at the time of synthesizing by the synthesizing circuit 527, data indicating a so-called minimum audible curve RC which is a human auditory characteristic as shown in FIG. The above-mentioned masking threshold MS can be synthesized. At this minimum audible curve, if the absolute noise level is below this minimum audible curve, the noise will not be heard. This minimum audible curve is different even if the coding is the same, for example, due to the difference in playback volume during playback,
In a realistic digital system, for example, there is not much difference in entering music into a 16-bit dynamic range, so if quantization noise in the most audible frequency band, for example, around 4 kHz is not heard, then other It is considered that quantization noise below the level of the minimum audible curve is not heard in the frequency band. Therefore, for example, the dynamic range of the system is 4 kHz.
If it is assumed that the usage is such that nearby noise is not heard, and an allowable noise level is obtained by combining the minimum audible curve RC and the masking threshold MS, the allowable noise level in this case is as shown in FIG. Up to the portion indicated by the oblique lines. In addition,
In this embodiment, the 4 kHz level of the minimum audible curve is adjusted to the lowest level corresponding to, for example, 20 bits. FIG. 14 also shows the signal spectrum SS.

In the allowable noise correction circuit 530,
The permissible noise level in the output from the subtractor 528 is corrected based on, for example, information on the equal loudness curve sent from the correction information output circuit 533. Here, the equal loudness curve is a characteristic curve relating to human auditory characteristics. For example, the loudness curve is obtained by calculating the sound pressure of sound at each frequency that sounds as loud as a pure tone of 1 kHz, and is connected by a curve. Also called a sensitivity curve. This equal loudness curve is the minimum audible curve R shown in FIG.
It draws substantially the same curve as C. In this equal loudness curve, for example, at around 4 kHz, even if the sound pressure falls by 8 to 10 dB below 1 kHz, it sounds as large as 1 kHz. It doesn't sound the same size. For this reason, it can be seen that noise exceeding the level of the minimum audible curve (allowable noise level) preferably has a frequency characteristic given by a curve corresponding to the equal loudness curve. From this, it can be seen that correcting the allowable noise level in consideration of the equal loudness curve is suitable for human auditory characteristics.

The above-described spectral shape depending on the permissible noise level is created by bit allocation using a certain percentage of the total available bits of 128 Kbps. This ratio decreases as the tonality of the input signal increases.

Next, a description will be given of a bit amount division method between the two bit distribution methods. Returning to FIG.
The signal from the input terminal 801 to which the output of the T circuit is supplied is
It is also provided to a spectrum smoothness calculation circuit 808, where the spectrum smoothness is calculated. In this embodiment, a value obtained by dividing the sum of the absolute values of the differences between the adjacent values of the absolute value of the signal spectrum by the sum of the absolute values of the signal spectrum is calculated as the smoothness of the spectrum.

The above-mentioned spectrum smoothness calculating circuit 808
Is supplied to a bit division ratio determination circuit 809, where the energy-dependent bit allocation and the bit division ratio between the bit allocations based on the permissible noise spectrum are determined. The bit division ratio is calculated by the spectrum smoothness calculation circuit 80.
It is considered that the larger the output value of 8 is, the less smooth the spectrum is, so that bit allocation is performed with more emphasis on bit allocation based on the permissible noise spectrum than on energy-dependent bit allocation. The bit division ratio determination circuit 809 sends a control output to multipliers 811 and 812 that control the size of the bit allocation based on the energy-dependent bit allocation and the permissible noise spectrum, respectively. Here, the output of the bit division ratio determination circuit 809 to the multiplier 811 is set to 0.8 so that the spectrum is smooth and the weight of the energy-dependent bit allocation is emphasized.
, The output of the bit division ratio determination circuit 809 to the multiplier 812 is 1-0.8 = 0.2. The output of these two multipliers is the adder 806
And output as the final bit allocation information from the output terminal 807.

FIGS. 15 and 16 show how bits are allocated at this time. FIGS. 17 and 18 show the corresponding quantization noise. FIG. 15 shows a case where the signal spectrum is relatively flat, and FIG. 16 shows a case where the signal spectrum exhibits high tonality. Also,
QS in FIGS. 15 and 16 indicates the bit amount corresponding to the signal level, and QN in the drawings indicates the bit amount corresponding to the bit allocation depending on the permissible noise level. 17 and 18, L indicates the signal level, NS in the figures indicates the noise reduction due to the signal level dependency, and NN in the figures indicates the noise reduction due to the bit allocation depending on the permissible noise level. .

First, in FIG. 15 showing the case where the spectrum of the signal is relatively flat, the bit allocation depending on the permissible noise level is useful for obtaining a large signal-to-noise ratio over the entire band. However, relatively low bit allocation is used in the low and high bands. This is because the sensitivity to noise in this band is small acoustically. Although the amount of the bit allocation depending on the signal energy level is small in amount, in this case, the bit allocation is focused on the high-frequency region where the signal level in the middle and low frequencies is high so as to generate a white noise spectrum.

On the other hand, as shown in FIG. 16, when the signal spectrum shows a high tonality, the bit allocation amount depending on the signal energy level increases, and the quantization noise decreases with a very narrow band of noise. Used to reduce The concentration of the bit allocation depending on the permissible noise level is less tight.

As shown in FIG. 16, the improvement of the characteristics in the isolated spectrum input signal is achieved by the sum of the bit allocation of these two.

The additional bit distribution (step ST) is added to the basic bit distribution obtained as described above as follows.
3) Add parts.

Next, the separation of the basic bit allocation portion and the additional bit allocation portion and the coupling at the time of reproduction will be described with reference to FIG.

First, it is assumed that the MDCT coefficients which are the outputs of the MDCT circuits 13, 14, 15 of FIG. 1 are supplied to the input terminal 900 having the configuration of FIG. That is, FIG. 19 is included in the adaptive bit allocation encoding circuit of FIG.

In FIG. 19, the input terminal 900
Are normalized by a normalization circuit 905 for each of a plurality of samples, that is, block floating is performed on the MDCT coefficients (MDCT samples). At this time, a scale factor is obtained as a coefficient indicating how much block floating has been performed.

Next-stage first quantizer 901
Performs quantization at each sample word length given by the basic bit allocation. At this time, in order to reduce quantization noise, quantization by rounding is performed.

Next, the output of the normalizing circuit 905 and the output of the quantizer 901 are sent to a differentiator 902. That is, the difference unit 902 obtains the difference between the input and output of the quantizer 901 (quantization error). This differentiator 902
Is further sent to a second quantizer 903 via a normalization circuit 906.

In the second quantizer 903, for example, 2
Bits are used for each sample. The floating coefficient at this time is automatically determined from the floating coefficient used in the first quantizer 901 and the word length.

That is, on the encoder side having the configuration in FIG. 19, when the word length used in the first quantizer 901 is, for example, N bits, the second quantizer 903 is set to (2 ** N). Is obtained.

Further, the second quantizer 903 for allocating additional bits has the first quantizer 903 for allocating basic bits.
In the same manner as the quantizer 901, bit allocation including round-off processing is performed. Thus, by the two quantizations, 2
Bit allocation.

Here, even if the word length for the additional bit allocation is not fixed, the size of the additional bit allocation component is determined from the scale factor of the basic bit allocation and the word length as described above. Can be calculated, so only the word length is needed for the decoder. In this embodiment, the word length of the additional bit allocation is fixed at 2 bits, so that even the word length for the additional bit allocation is not necessary. In this way, highly efficient quantization in which the outputs of the quantizers 901 and 903 are rounded off is realized.

When the output bit rates of the quantizers 901 and 903 are both fixed, a disc, a tape, and a
The system can be simplified when recording on media such as media. Further, it is also possible to make the total variable while making both variable. Of course, only the output bit rate of some quantizers may be fixed.

Incidentally, in the configuration (decoder) corresponding to the configuration (encoder) in FIG.
Inverse normalization circuit 90 for performing inverse normalization processing corresponding to 06
8, 907 are provided, and the outputs of the circuits 908, 907 are added by the adder 904. The added output is output terminal 91
It will be taken from 0.

FIG. 20 shows a high-efficiency code decoding apparatus serving as a basic example for decoding again a signal which has been highly-encoded in this way.

In FIG. 20, the quantized MDCT coefficients of each band are input to decoding device input terminals 122 and 124,
The block size information provided to 126 and used is provided to input terminals 123, 125, and 127. The decoding circuits 116, 117 and 118 release the bit allocation using the adaptive bit allocation information.

Next, the IMDCT circuits 113, 114, 1
At 15, the signal in the frequency domain is converted to a signal in the time domain. The time domain signals of these partial bands are decoded by the IQMF circuits 112 and 111 into the entire area signals.

That is, the basic bit allocation of 128 k
64 kbp of the bit allocation of bps and the additional bit allocation
s are the decoding circuits 116, 117, and 118, respectively.
Is decrypted. After these two decoded portions are decoded, the respective samples on the time axis are added to each other to obtain samples with high accuracy.

Of course, in FIG. 20, it is also possible to calculate the basic bit allocation output and the additional bit allocation for each of the outputs of the IMDCT circuits 113, 114 and 115, combine them, and send them to the IQMF circuits 112 and 111.

Further, the decoding circuits 116, 117, 11
8, the basic bit allocation and the additional bit allocation are added after the normalization processing is solved, and the sum is added to the IMDCT, IQMF
It can be processed to get the final output.

Next, the medium used in the embodiment of the present invention is for recording or transmitting a signal coded by the high-efficiency coding apparatus of the embodiment of the present invention as described above. For example, a disk-shaped recording medium such as an optical disk, a magneto-optical disk, or a magnetic disk in which the above-mentioned encoded signal is recorded, a tape-like recording medium such as a magnetic tape in which the above-mentioned encoded signal is recorded, Semiconductor memory, an IC card, and the like in which the conversion signal is recorded. In addition, examples of the transmission medium include an electric wire, an optical cable, and a radio wave.

FIG. 21 shows the arrangement of data on the media used in the embodiment of the present invention. That is, one sync block includes sync information and sub information (scale factor, word length).
, Basic bit allocation, and additional bit allocation.

In this case, after quantizing the time domain sample or the frequency domain sample, at least two quantizing functions are performed by one sample by one quantizing function for further quantizing the quantization error of the preceding stage. Decomposing into words, separately recording or transmitting each quantized output in one sync block, and then decoding and reproducing,
This is effective in that a bit string portion to be removed when reproducing at a reduced bit rate can be collectively removed.

As another method, after quantizing a time-domain sample or a frequency-domain sample, at least one quantization function for further quantizing a preceding-stage quantization error by one sample alone is used. Decomposing into at least two words, recording or transmitting each quantized output alternately in frequency or time order in one sync block, and decoding and reproducing from time domain samples or frequency domain samples lowers the bit rate. This is effective in that a bit string portion to be removed can be collectively removed in a form in which the frequency band is limited when the data is reproduced.

The bit arrangement as described above can be applied to, for example, a so-called mini disk (Mini Disc) using a magneto-optical disk or an optical disk, a magnetic tape medium, a communication medium, and the like.

[0099]

As is clear from the above description, the following effects can be obtained in the present invention. That is, (1) when decoding is performed using a bit rate lower than the bit rate used at the time of encoding, for example, when a part of the bits after the encoding process is diverted for another data transfer on the encoding side, the sound quality is deteriorated. Keep it to a minimum. (2) When a playback device that plays back at a low bit rate is already used, a playback device that plays back at a low bit rate already used and a back A system having word compatibility can be provided. (3) When recording is to be performed on an expensive storage device, for example, a storage medium using an IC card, and when it is desired to extend the recording time from the initial setting, recording is performed by appropriately reducing the bit rate of the encoded information being recorded or being recorded. It is possible to extend the time and minimize the sound quality deterioration at this time. (4) A high-quality sound decoder can be created by using a plurality of inexpensive decoders that perform bit allocation at a lower bit rate than those usually used, thereby making it unnecessary to create a new decoder LSI. The purpose can be achieved at low cost.

[Brief description of the drawings]

FIG. 1 is a block circuit diagram illustrating a configuration example of a high-efficiency encoding device according to an embodiment of the present invention.

FIG. 2 is a diagram illustrating frequency and time division of a signal in the apparatus according to the embodiment.

FIG. 3 is a diagram illustrating a bit allocation strategy according to the present embodiment.

FIG. 4 is a diagram for explaining a method of calculating tonality from a scale factor.

FIG. 5 is a diagram for explaining a method of obtaining a bit allocation amount of bit allocation (1) from tonality.

FIG. 6 is a diagram showing a noise spectrum at the time of uniform distribution in bit distribution (2).

FIG. 7 is a diagram showing an example of a noise spectrum by bit allocation for obtaining an audible effect having dependency on a frequency spectrum and a level of an information signal in bit allocation (2).

FIG. 8 is a diagram showing uniform distribution in bit distribution (2).

FIG. 9 is a diagram showing a bit allocation method using bit allocation for obtaining an audible effect having dependency on the frequency spectrum and level of an information signal in bit allocation (2).

FIG. 10 is a block circuit diagram showing a configuration example of a basic bit allocation function according to the embodiment of the present invention.

FIG. 11 is a block circuit diagram showing a configuration example of an auditory masking threshold calculation function according to the embodiment of the present invention.

FIG. 12 is a diagram illustrating masking by each critical band signal.

FIG. 13 is a diagram illustrating a masking threshold according to each critical band signal.

FIG. 14 is a diagram showing an information spectrum, a masking threshold, and a minimum audible limit.

FIG. 15 is a diagram illustrating bit allocation depending on the signal level and the permissible noise level for an information signal having a flat signal spectrum.

FIG. 16 is a diagram illustrating a signal level-dependent and permissible noise level-dependent bit allocation for an information signal having a high tonality of a signal spectrum.

FIG. 17 is a diagram showing a quantization noise level for an information signal having a flat signal spectrum.

FIG. 18 is a diagram showing a quantization noise level for an information signal having a high tonality.

FIG. 19 is a block circuit diagram showing a specific configuration for dividing a basic bit allocation and an additional bit allocation.

FIG. 20 is a block circuit diagram illustrating a configuration example of a high-efficiency code decoding device used in an embodiment of the present invention.

FIG. 21 is a diagram illustrating a configuration example of a bit arrangement in a medium used in an embodiment of the present invention.

[Explanation of symbols]

10 high efficiency coding circuit input terminal 11, 12 band division filter 13, 14, 15 MDCT circuit 16, 17, 18 ... adaptive bit allocation coding circuit 19, 20, 21 ..Block size determination circuits 22, 24, 26 ... Encoding output terminals 23, 25, 27 ... Block size information output terminals 122, 124, 126 ... Encoding input terminals 123, 125, 127 ... -Block size information input terminals 116, 117, 118 ... adaptive bit allocation decoding circuits 113, 114, 115 ... IMDCT circuits 112, 111 ... IQMF circuits 110 ... high efficiency decoding circuit output terminals 520 ... Allowable noise calculation circuit 521 ... Allowable noise calculation circuit input terminal 522 ... Energy detection circuit for each band 523 ... Convolution filter Data circuit 524 ··· subtracter 525 ··· n-ai function generation circuit 526 ··· divider 527 ··· synthesis circuit 528 ··· subtractor 530 ··· allowable noise correction circuit 532 ··· Minimum audible curve generation circuit 533: Correction information output circuit 801: MDCT circuit output input terminal 802: Available total bit generation circuit 803: Energy calculation circuit for each band 804: Energy-dependent bit Allocation circuit 805: A bit allocation circuit depending on the permissible noise level 806: Adder 807: A bit allocation output terminal for each band 808: Smoothness calculation circuit for spectrum 809: Bit division rate determination Circuits 811, 812 ... Multiplier

──────────────────────────────────────────────────続 き Continued on the front page (58) Field surveyed (Int.Cl. ⁷ , DB name) H03M 7/30 G10L 13/00 G11B 20/10 301

Claims (7)

(57) [Claims] 1. An encoding device for quantizing a time-domain sample or a frequency-domain sample of an input signal with a predetermined number of bits, wherein: a first quantization means for quantizing the time-domain sample or the frequency-domain sample; A quantization error calculation unit for calculating a quantization error of the first quantization unit; a second quantization unit for further quantizing the quantization error; an output of the first quantization unit and the second Encoding means for encoding the output of the quantization means. 2. The encoding apparatus according to claim 1, wherein an output bit rate of said second quantization means is set to a constant bit rate in a fixed time unit. 3. The encoding apparatus according to claim 2, wherein the output bit rates of said first quantization means and said second quantization means are set to a fixed bit rate in fixed time units. 4. A first block floating means for blocking the time domain samples or frequency domain samples for each of a plurality of samples and performing a block floating process, and a second block floating process for performing an output of the quantization error calculating means in a block floating process. And a second block floating means, wherein the second block floating means sets a scale factor for the block floating to the first.
The first quantizing means quantizes the output of the first block floating means, and the second quantizing means quantizes the output of the first block floating means. 2. The encoding apparatus according to claim 1, wherein an output of said encoding is quantized. 5. The second block floating means obtains a scale factor for the block floating based on a scale factor in the first block floating means and a word length in the first quantization means. The encoding device according to claim 4, wherein: 6. An encoding apparatus which quantizes and encodes a time-domain sample or a frequency-domain sample of an input signal by a predetermined number of bits and records the encoded signal on a recording medium, wherein the time-domain sample or the frequency-domain sample is A first quantization unit for quantizing the following, a quantization error calculation unit for calculating a quantization error of the first quantization unit, a second quantization unit for further quantizing the quantization error, An encoding apparatus comprising, in one sync block, recording means for separately recording an output sample of the first quantizing means and an output sample of the second quantizing means. . 7. An encoding apparatus which quantizes and encodes a time-domain sample or a frequency-domain sample of an input signal with a predetermined number of bits and records the encoded signal on a recording medium, wherein the time-domain sample or the frequency-domain sample is A first quantization unit for quantizing the following, a quantization error calculation unit for calculating a quantization error of the first quantization unit, a second quantization unit for further quantizing the quantization error, In one sync block, there is provided recording means for alternately recording the output samples of the first quantization means and the output samples of the second quantization means in time order or frequency order. Encoding device.