KR100351772B1 - Digital encoding device, digital recording signal derivation device and digital signal data derivation method - Google Patents

Abstract

(Configuration) In the bit distribution circuit 304 depending on the signal energy, in terms of the total bits available for bit allocation, when adaptively bit-aligning the spectrum data supplied from the input terminal 301 by orthogonal conversion Is divided into allocations and the allocations in the fixed bit allocation circuit 305 are divided.

(Effect) Even when the signal spectrum is dispersed, even when the tonality of the signal is high, good characteristics can be obtained, respectively.

Description

Digital encoding device, digital record signal derivation device and digital signal data derivation method

The present invention provides a digital encoding device, a digital input signal compression system, and a compressed digital recording signal deriving device, which perform encoding, transmission, recording, reproducing, and decoding according to an input digital according to the present invention by so-called high efficiency encoding to obtain a reproduction signal. And to a method.

There are various methods of high-efficiency encoding of signals such as audio or voice. For example, non-blocking frequency band division in which a plurality of frequency bands are divided and encoded in a plurality of frequency bands without blocking the audio signals on the time axis. Examples of the scheme include band division coding (sub band coding: SBC), and a block that converts a signal on a time axis into a signal on a frequency axis (orthogonal transformation), divides the signal into a plurality of frequency bands, and encodes each band. Frequency band division, so-called transform coding, and the like.

In addition, a method of high efficiency coding combining the above-described band division coding and transform coding can also be considered. In this case, for example, after band division is performed by the band division coding, signals for each band are obtained. Orthogonal transformation is performed on the signal on the frequency axis, and encoding is performed for each of the orthogonally transformed bands.

Here, as the above-described filter for band division, for example, there is a QMF filter, 1976 R. E. Crochiere, Digital coding of speech in subbands, Bell Syst. Tech. J. Vol. 55, No. 8, 1976. In addition, ICASSP 83, BOSTON Polyphase Quadrature filters-A new subband coding technique, Joseph H. Rothweiler, describes an equal bandwidth filter division technique. Next, as the orthogonal transform described above, for example, the input audio signal is blocked at a predetermined unit time (frame), and for each block, a fast Fourier transform (FFT), a discrete cosine transform (DCT), and a modified DCT transform (MDCT). And the like to convert the time axis into a frequency axis. Regarding MDCT, ICASSP 1987 Subband / Transform Coding Using Filter Bank Designs Based On Time Domain Aliasing Cancellation. J. P. Princen, A. B. Bradley, Univ. of Surrey Royal Melbourne Inst. of Tech.

Furthermore, as the frequency division width for quantizing each frequency component divided into frequency bands, for example, band division in consideration of human auditory characteristics is performed. In other words, the higher bandwidth generally referred to as the critical band (critical band) is a bandwidth in which the bandwidth is wider, and there is a case where an audio signal is divided into a plurality of bands (for example, 25 bands). Further, at this time, when encoding according to the present invention for each band, encoding by predetermined bit allocation for each band or adaptive bit allocation for each band is performed. For example, when encoding the coefficient data obtained by the MDCT processing by the bit allocation, the number of adaptive allocation bits for the MDCT coefficient data for each band that can be obtained by the MDCT processing for each block is obtained. Encoding is done.

As the bit allocation method, the following two methods are known.

First, IEEE Transaction of Acoustics, Speech, and Signal Processing, vol. In ASSP-25, No. 4, and August 1977, bit allocation is performed based on the size of each band. In this system, the quantization noise spectrum is flattened and the noise energy is minimized. However, since the masking effect is not used acoustically, the actual noise is not optimal.

Next, in ICASSP 1980 The critical hand coder-digital encoding of the perceptual requirements of the auditory system, MAKransner, MIT, auditory masking is used to obtain the required signal-to-noise ratio for each band and to perform fixed bit allocation. This is described. However, even when the sine wave input characteristic is measured by this method, the characteristic value does not become such a good value because the bit allocation is fixed.

In this way, if the bit allocation is performed by the signal size of each band and the quantization noise energy is minimized, the auditory noise level is not minimized. If the bit allocation is fixed to each band in consideration of the masking effect, It is difficult to obtain good signal-to-noise characteristics.

SUMMARY OF THE INVENTION The present invention has been made in view of the above circumstances, and provides a method and apparatus for efficiently encoding digital data having a bit allocation method that is acoustically preferable and obtains good characteristics even for an isolated spectral input such as a 1 kHz sine wave input. For the purpose.

The high efficiency coding method of digital data according to the present invention divides an input digital signal into a plurality of frequency bands, adaptively changes the block size for each frequency band, and then obtains spectrum data by performing orthogonal transformation. In the high-efficiency encoding method of digital data in which the spectral data is adaptively assigned bits for each critical band, a predetermined bit allocation pattern for a predetermined time, and a time for all the bits that can be used for bit allocation, The above-mentioned problem is solved by being distributed to allocations which depend on the magnitude of the signal in the small block subdivided with respect to frequency. In addition, by having a plurality of fixed bit allocation patterns, more efficient allocation is possible.

Further, means for dividing the input digital signal into a plurality of frequency bands, means for adaptively changing the block size for each frequency band, and spectral data for performing orthogonal transformation for each frequency band for which the block size is changed And means for adaptively allocating the spectral data to every critical band, wherein the bit allocation means is predetermined fixed bit allocation pattern for any time, for all the bits available for bit allocation. The above-mentioned problem is also solved by a high efficiency encoding apparatus for digital data distributed to an allocation portion depending on the magnitude of the signal in the small block subdivided with respect to the branch and time and frequency.

In this case, the smaller the signal size is, the more advantageous it is to select a pattern having a small high frequency bit allocation ratio after effective distribution. In a high efficiency code using a frequency decomposition technique of further blocking frequency dividing the non-blocking frequency division output, it is also advantageous to select a pattern by the magnitude of the non-blocking frequency division output. In addition, the hardware scale can be reduced by making the number of bits of a plurality of fixed bit allocation patterns the same.

According to the present invention, even when the spectrum is dispersed like a music signal, the noise level can be lowered acoustically due to the masking effect, and the signal-to-noise ratio can be gathered into a larger band even when a sine wave is input. Can be increased.

1 shows a high efficiency encoding of an input digital signal such as an audio PCM signal using respective techniques of band division coding (SBC), adaptive transform coding (ATC), and adaptive bit allocation (APC-AB). It demonstrates with reference to.

In the specific high efficiency encoding apparatus shown in FIG. 1, an input digital signal is divided into a plurality of frequency bands by a filter or the like, orthogonal transformation is performed for each frequency band, and the spectral data of the obtained frequency axis is described below. The encoding is performed by adaptively bit allocation for each so-called critical band (critical band) width in consideration of characteristics. Of course, the frequency division width of non-blocking by a filter or the like may be an equal division width. Furthermore, in the embodiment of the present invention, the block size (block length) is adaptively changed in accordance with the input signal before the orthogonal transform, and the floating processing is performed in the critical band unit or the high frequency block to further refine the threshold bandwidth. . This critical band is a frequency band divided in consideration of human auditory characteristics, and is a band of the noise when the pure tone is masked by narrow-band noise of the same intensity near the frequency of a certain pure tone. This critical band has a wider bandwidth at higher frequencies, and the entire frequency band of 0 to 20 kHz is divided into 25 critical bands, for example.

That is, in Fig. 1, an audio PCM signal of 0 to 20 kHz is supplied to the input terminal 10, for example. This input signal is divided into a band of 0 to 10 kHz and a band of 10 k to 20 kHz by a band dividing filter 11 such as a QMF filter, for example, and a signal of the band of 0 to 10 kHz is similarly a band of a so-called QMF filter. The split filter 12 divides the signal into 0 to 5 kHz bands and 5 to 10 kHz bands. Signals in the 10 k to 20 kHz band of the band split filter 11 are sent to the Modified Discrete Cosine Transform (MDCT) circuit 13, which is an example of an orthogonal transform circuit, and has a band of 5 k to 10 kHz in the band split filter 12. The signal is sent to the MDCT circuit 14, and the signals of the 0 to 5 kHz band in the band division filter 12 are sent to the MDCT circuit 15 to be MDCT processed, respectively.

Here, the specific example of the block size of each MDCT circuit 13, 14, 15 is shown in FIG. In the specific example of FIG. 2, the higher frequency side makes the frequency band wider and the time resolution higher (shorter block length). That is, for the signals in the low band 0-5 kHz band and the signals in the mid-range 5k-10 kHz band, the number of samples in one block b _L and b _M is 256 samples, for example, and the signals in the high band 10k-20 kHz band are used. Regarding, b _H was blocked by the lengths BL _L / 2 and BL _M / 2 of 1/2 of the blocks b _L and b _M on the low and mid-range sides, respectively. In this way, the number of orthogonal transform block samples in each band is made the same. In addition, each band is capable of adaptive block division, such as 1/2, 1/4, again, assuming a large temporal change in the signal.

Further, in FIG. 1, spectrum data or MDCT coefficient data on the frequency axis obtained by MDCT processing in each MDCT circuit 13, 14, 15 is further subdivided into a so-called critical band (critical band) or in a high band. Each block is combined (grouped) and sent to the adaptive bit allocation coding circuits 16, 17, and 18.

Each spectral data (or MDCT coefficient data) is re-coded by the adaptive bit allocation coding circuits 16, 17, and 18 according to the number of bits allocated for each threshold band or for each block in which the threshold bandwidth is further subdivided in the high range. I'm trying to quantize. The data encoded in this manner is taken out through the output terminals 22, 24, and 26. In this case, it indicates which signal has been normalized. Floating information and bit length information indicating which bit length has been quantized are sent simultaneously.

FIG. 3 is a functional block diagram showing a specific example of internal functions of the adaptive bit allocation coding circuits 16, 17, and 18. The output of each of the MDCT circuits 13, 14, and 15 in FIG. It is sent to the energy calculation circuit 303 for each band through the input terminal 301 of the adaptive bit allocation function unit 300 of 3 degrees, and the energy for each said critical band (critical band) is for example in the said band. It is calculated | required by calculating the square root of the square root mean of each amplitude value of, and the like. Instead of the energy for each of these bands, a peak value, an average value, or the like of an amplitude value may be used. 4 shows an example of the spectrum of the sum value in the block as an output from the energy calculation circuit 303, for example, within a critical band (critical band) or in a more detailed subdivision of the critical band (critical band) width. However, in FIG. 4, for the sake of simplicity, the number of the critical bands (critical bands) or the number of block bands in which the critical bands (critical bands) are further subdivided in a high band is represented by 12 bands (B1 to B2). have.

The adaptive bit allocation operation will be described in more detail with reference to FIG. Now, the total number of bits (bit rate) used for transmission or recording by expressing the MDCT coefficients is 100 Kbps. This is illustrated in Figure 3 as 100 Kbps total bits in block 302 representing the total available bits. In this embodiment, 60 Kbps is used for the fixed bit allocation in the fixed bit distribution circuit 305, and the remaining 40 Kbps is used for the bit allocation in the bit distribution circuit 304 depending on the buck spectrum for each band. do. A plurality of bit allocation patterns for fixed bit allocation are prepared, and various selections can be made depending on the nature of the signal. In the embodiment, there are various patterns in which a short amount of block bits corresponding to 100 Kbps is distributed to each frequency.

In particular, in the present embodiment, a plurality of patterns having different bit distribution ratios in the mid-low range and the high range are prepared. And, the smaller the signal size is, the more the pattern of the allocation to the high range is selected. In this way, a small signal time results in a loudness effect in which the sensitivity of the high range is lowered. Although the magnitude | size of the signal of the full band can also be used as a magnitude | size of the signal at this time, the output of a non-blocking frequency division circuit in which a filter etc. are used on it, or MDCT output is used.

In the fixed bit allocation and the critical band (critical band) or high frequency, the sum of the bit allocation values depending on the signal value of the block further subdivided in the critical band (critical band) width is obtained by the sum calculating circuit 306, and the output terminal It is taken out through 307 and used for quantization. Next, FIG. 5 shows a state of bit allocation when the signal spectrum is flat, and FIG. 6 shows a state of quantization noise (noise spectrum) corresponding thereto. In addition, Fig. 7 shows the state of bit allocation when the tonality of the signal spectrum is high, i.e., when the signal is pitched acoustically and the frequency is deflected, and corresponding quantization noise (noise) is shown. The state of the spectrum) is shown in FIG. Here, in Figs. 5 and 7, the white portion represents the bit amount of the fixed bit allocation, and the diagonal portion represents the bit amount of the signal level dependency. 6 and 8, curve a represents a signal level, curve b represents a noise level due to fixed bit allocation, and an oblique section c represents noise reduction due to signal level dependency.

That is, FIGS. 5 and 6 show the case where the signal spectrum is ratio flat, and the noise level by the fixed bit allocation helps to take some signal noise ratio over the entire band. However, relatively low bit allocations are used in the low and high frequencies. This is due to the low importance of this band acoustically. At the same time, the degree of bit allocation dependent on the signal level can selectively lower the noise level of the band having a large signal size. However, when the spectrum of the signal is ratio flat, this selectivity ratio acts over a wide band.

On the other hand, as shown in Figs. 7 and 8, when the signal spectrum exhibits high tonality, the reduction of the quantization noise due to the degree of bit allocation depending on the signal level can reduce the noise of the extremely narrow band. Used for This achieves an improvement in the characteristics in the isolated spectral input signal.

9 shows a decoding circuit for decoding the high efficiency coded signal in this manner and then decoding it again after transmission or recording and reproduction. The quantized MDCT coefficients of the respective bands are provided to the decoding circuit input terminals 122, 124, and 126, and the used block size information is given to the input terminals 123, 125, and 127. The decoding circuits 116, 117, and 118 release bit allocation using the adaptive bit allocation information. Next, in the IMDCT (inverse MDCT) circuits 113, 114, and 115, the signal on the frequency axis is converted into a signal on the time axis. The signals on the time axis of these partial bands are decoded into full band signals by the IQMF (inverse QMF) circuits 112 and 111 and taken out from the output terminal 110.

As is clear from the above description, according to the high-efficiency encoding method of digital data according to the present invention, after dividing an input digital signal into a plurality of frequency bands, performing orthogonal transformation, and adaptively bitting the obtained spectrum data for each critical band. A high-efficiency encoding method of digital data that allocates a number of bits to a predetermined fixed bit allocation pattern for any time for all bits that can be used for bit allocation, and to a signal size in a small block subdivided with respect to time and frequency. It is distributed among allocating dependencies. This bit allocation method is audibly preferable, and can realize the bit allocation obtained by only one operation without repeatedly adjusting the bit quantity many times with good characteristics for an isolated spectral input such as a 1 kHz sine wave input. . In other words, even when the spectrum is dispersed like a music signal, the masking effect can lower the noise level aurally, and even when the sine wave input collects bits in a large band, the signal-to-noise ratio can be increased. Can be.

The same effect can also be obtained by the high efficiency encoding of the digital data which encodes the digital data and transmits or records the digital data by the high efficiency encoding method.

1 is a block circuit diagram showing an example of a configuration of an encoding apparatus according to an embodiment of the present invention.

2 is a diagram showing a specific example of the frequency and time division of a signal of the example apparatus.

3 is a functional block diagram for explaining an example of a bit allocation algorithm of an adaptive bit allocation coding circuit used in the embodiment apparatus.

4 shows a Berke spectrum.

5 is a diagram showing an example of bit allocation at the time of signal input of the substantially flat spectrum of the embodiment.

6 is a diagram showing an example of a quantization noise spectrum at the time of input of a signal of approximately flat spectrum of the embodiment.

7 is a diagram showing an example of bit allocation at the time of signal input having the high tonality of the embodiment.

8 is a diagram showing an example of a quantization noise spectrum at the time of signal input having the high tonality of the embodiment.

9 is a block circuit diagram showing an example of the configuration of a decoding apparatus for the encoding apparatus of the embodiment.

* Explanation of symbols for the main parts of the drawings *

10: high efficiency coding circuit input terminal

11, 12: QMF circuit

13, 14, 15: MDCT circuit

16, 17, 18: adaptive bit allocation coding circuit

19, 20, 21: block determination circuit

22, 24, 26: coded output terminal

23, 25, 27: block size information output terminal

122, 124, 126: coding input terminal

123, 125, 127: block size information input terminal

116, 117, 118: adaptive bit allocation decoding circuit

113, 114, 115: IMDCT circuit

111, 112: IQMF Circuit

110: high efficiency decoding circuit output terminal

300: adaptive bit allocation function

302: Block indicating total number of bits available

303: energy output circuit for each band

304: energy dependent bit distribution circuit

305: fixed bit distribution circuit

306: sum operation circuit of bits

Claims (18)

A digital encoding device for compressing a digital input signal to provide a compressed output signal, Frequency dividing means for receiving the digital input signal and dividing the digital input signal into a plurality of frequency ranges; Time division means for time division of at least one of the frequency ranges of the digital input signal into a plurality of blocks; Orthogonal transform means for orthogonally transforming each block to provide a plurality of spectral coefficients; Means for grouping the plurality of spectral coefficients from the orthogonal transforming means into threshold bands each having a band energy and a band frequency by frequency; Bit allocation pattern selecting means for selecting one predetermined bit allocation pattern from the plurality of predetermined bit allocation patterns; A first bit allocation means for allocating a total number of fixed bits between the threshold bands to assign a plurality of fixed bits to the threshold bands for quantizing each spectral coefficient within each threshold band, each threshold The first bit allocation means, the number of fixed bits allocated to a band determined according to the selected predetermined bit allocation pattern; A second bit allocation means for allocating a total number of variable bits between the threshold bands for quantizing each spectral coefficient in each threshold band, for allocating a plurality of variable bits to the threshold band, each threshold The second bit allocation means, the number of variable bits allocated to a band determined according to the band energy of the threshold band; Each using a plurality of bits equal to the sum of the number of fixed bits allocated to quantize each spectral coefficient in the threshold band and the number of variable bits allocated to quantize each spectral coefficient in the threshold band, Means for quantizing each spectral coefficient in a band, wherein the spectral coefficients in at least one of the threshold bands are quantized using more than zero fixed bits and more than zero variable bits, And a quantization means. The method of claim 1, And a total number of the fixed bits and a total number of the variable bits are each kept constant over time. The method of claim 2, Further comprising energy determining means for determining energy of the digital input signal, The bit pattern selection means is arranged in threshold bands with band frequencies on higher frequencies, rather than threshold bands with band frequencies on intermediate frequencies when the energy determining means determines that the energy of the digital input signal is low. And selecting a predetermined bit allocation pattern to which fewer bits are allocated. The method of claim 2, Energy determining means for determining a band energy of one of the threshold bands, The bit pattern selecting means, when the energy determining means determines that the band energy in the critical band of one of the critical bands is low, is a band toward higher frequencies than threshold bands with band frequencies toward intermediate frequencies. Selecting one bit allocation pattern of the plurality of predetermined bit allocation patterns to which fewer bits are allocated to the threshold bands having the frequency. The method of claim 2, Further comprising energy determining means for deriving a derived signal by frequency dividing one of the frequency ranges of the digital input signal, and for determining the energy of the derived signal, The bit allocation pattern selecting means is that, when the energy determining means determines that the energy of the derived signal is low, fewer bits are allocated to the critical bands on the higher frequencies than to the critical bands on the intermediate frequencies. And a bit allocation pattern of one of the plurality of predetermined bit allocation patterns. The method of claim 2, Each block has a duration of time, And said time division means for changing the duration of time of blocks in at least one of said frequency ranges in response to said digital input signal. The method of claim 1, Each block has a duration of time, And said time division means for changing the duration of time of blocks in at least one of said frequency ranges in response to said digital input signal. An apparatus for deriving a compressed digital recording signal from a digital input signal for recording on a recording medium, the apparatus comprising: Means for dividing the digital input signal into a plurality of frequency ranges; Means for time-dividing each of the frequency ranges of the digital input signal into a plurality of blocks; Orthogonal transform means for orthogonally transforming each block to provide a plurality of spectral coefficients; Means for grouping the plurality of spectral coefficients from the orthogonal transforming means into threshold bands each having a band energy and a band frequency by frequency; Bit allocation pattern selecting means for selecting a predetermined bit allocation pattern from the plurality of predetermined bit allocation patterns; First bit allocation means for allocating a total number of fixed bits between the threshold bands for quantizing each spectral coefficient within each threshold band, for allocating a plurality of fixed bits in the threshold band, each threshold The first bit allocation means, the number of fixed bits allocated to a band determined according to the selected predetermined bit allocation pattern; Second bit allocation means for allocating a total number of variable bits between the threshold bands to quantize each spectral coefficient within each threshold band, to assign a plurality of variable bits to the threshold band, The second bit allocation means, wherein the total number of the variable bits allocated to a critical band is determined according to the band energy; Each using a plurality of bits equal to the sum of the number of fixed bits allocated to quantize each spectral coefficient in the threshold band and the number of variable bits allocated to quantize each spectral coefficient in the threshold band, Means for quantizing each spectral coefficient in a band, the spectral coefficients in at least one of the critical bands being quantized using more than zero fixed bits and more than zero variable bits The quantization means for generating word length data indicative of the sum for each threshold band; Means for multiplexing the quantized spectral coefficients and the word length data to provide the compressed digital signal; And recording means for recording the compressed digital signal on the recording medium. The method of claim 8, And the recording medium is an optical disk. The method of claim 8, And the recording medium comprises a semiconductor memory. The method of claim 8, And the total number of fixed bits and the total number of variable bits are each kept constant over time. The method of claim 11, Each block has a duration of time, And means for time-dividing each frequency range comprises means for changing a duration of time of blocks in at least one of the frequency ranges in response to the digital input signal. A method for deriving compressed digital signal data from a digital input signal, the method comprising: Dividing the digital input signal into a plurality of frequency ranges; Time dividing each of the frequency ranges of the uncompressed digital input signal into a plurality of blocks; Orthogonally transforming each block to provide a plurality of spectral coefficients; Grouping the plurality of spectral coefficients provided by the orthogonal transformation by frequency into threshold bands each having a band energy and a band frequency; Providing a total number of variable bits and a total number of fixed bits, wherein the total number of fixed bits is arranged in a plurality of predetermined bit allocation patterns; Selecting one of the plurality of predetermined bit allocation patterns as the selected predetermined bit allocation pattern; Allocating the total number of fixed bits between the threshold bands to assign a plurality of fixed bits to the threshold band to quantize each spectral coefficient within each threshold band, the allocation to each threshold band The fixed bit allocation step being determined according to the selected predetermined bit allocation pattern; Allocating a total number of variable bits between the threshold bands to assign a plurality of variable bits to the threshold band, to quantize each spectral coefficient within each threshold band, The variable bit allocation step, wherein the total number of the variable bits is determined according to the band energy of the threshold band; Each using a number and bits equal to the sum of the number of fixed bits allocated to quantize each spectral coefficient in the threshold band and the number of variable bits allocated to quantize each spectral coefficient in the threshold band, Quantizing each spectral coefficient in a critical band, wherein the spectral coefficients in at least one of the critical bands are quantized using more than zero fixed bits and more than zero variable bits The quantization step; Generating word length data indicative of the sum for each threshold band; Multiplexing the quantized spectral coefficients and the word length data to provide a compressed digital signal. The method of claim 13, In the providing of the total number of fixed bits and the total number of variable bits, the total number of provided fixed bits and the total number of provided variable bits are each kept constant over time. The method of claim 14, In the time division step, each block has a duration of time, And wherein said time division comprises changing a duration of time of said blocks within at least one of said frequency ranges in response to said digital input signal. The method of claim 13, Determining the energy of the digital input signal; Selecting a bit allocation pattern of one of the plurality of predetermined bit allocation patterns, when the energy determining step determines that the energy of the digital input signal is low, rather than threshold bands having band frequencies toward intermediate frequencies And a bit allocation pattern of one of a plurality of predetermined bit allocation patterns to which fewer bits are allocated to threshold bands having band frequencies toward higher frequencies is selected. The method of claim 13, Determining the band energy of one of the threshold bands; Selecting a bit allocation pattern of one of the plurality of predetermined bit allocation patterns, when the energy determining step determines that the band energy of one of the critical bands is low, the band frequency toward the intermediate frequencies Digital signal data in which one bit allocation pattern of one of the plurality of predetermined bit allocation patterns to which fewer bits are allocated is selected in the threshold bands having the band frequencies toward higher frequencies, rather than the threshold bands with Derivation method. The method of claim 13, Deriving a derived signal by frequency dividing one of the frequency ranges of the digital input signal; Determining the energy of the derived signal; In the step of selecting one bit allocation pattern of the plurality of predetermined bit allocation patterns, when the energy determination step determines that the energy of the derived signal is low, rather than threshold bands having band frequencies toward intermediate frequencies, And in the critical bands with band frequencies towards higher frequencies, one bit allocation pattern of the plurality of predetermined bit allocation patterns to which fewer bits are allocated is selected.