CN101989428B - Bit distribution method, coding method, decoding method, coder and decoder - Google Patents

Abstract

The invention provides a bit distribution method comprising the following steps: carrying out perceptual sorting on all subbands of a signal according to the energy of the subband; according to the magnitude of frequency ranges corresponding to the subbands on a frequency domain and the energy difference of the subbands, adjusting the sequence of the subbands in the perceptual sorting; and distributing the bit resource according to the adjusted perceptual sorting sequence. The invention also provides a coding method, a coder, a bit distribution method, a decoding method and a decoder. By using the invention, the complexity of bit distribution can be lowered, and bit can be more reasonably distributed to the subbands with important perception, thereby ensuring that the coder and the decoder can generate better hearing feeling.

Description

Bit allocation method, encoding method, decoding method, encoder and decoder

technical field

The present invention relates to codec technology, in particular to bit allocation technology.

Background technique

The sub-band coding technology is one of the commonly used coding and decoding technologies at present. Audio signals can be divided into multiple frequency bands from low to high in the frequency domain. A frequency band is called a sub-band, and subsequent encoding, decoding, and other work are performed on the basis of sub-bands. In the process of encoding and decoding, limited bit resources need to be allocated between subbands. Generally, the more bits a subband obtains, the smaller the error in quantization and the better the quality of encoding and decoding. On the contrary, if the subband The less bits the band gets, the greater the error in quantization and the worse the quality of the codec.

The bit allocation is generally guided by the perceptual characteristics of the human ear. The physiological characteristics of the human ear determine that the human ear is not equally sensitive to signals of all frequencies. The specific manifestations are as follows: 1. The human ear is more likely to detect the appearance and change of low-frequency signals, compared with the appearance and change of high-frequency signals; 2. When low-frequency signals and high-frequency signals appear at the same time, the human ear may only be able to detect Hearing low-frequency signals but not hearing high-frequency signals, this phenomenon is called "masking effect". Its occurrence is related to the frequency and energy of the signal, and it will occur when certain conditions are met. This characteristic of the human ear can be simulated by establishing a psychoacoustic model. Another important auditory perception characteristic of the human ear is that in a quiet environment, the human ear can only hear sounds whose amplitude is greater than a certain threshold, and this threshold is the amplitude of the sound that the human ear can just hear. This threshold is called It is the absolute hearing threshold of the human ear (LTQ, Listening Threshold in Quiet). The LTQ curve changes with frequency, which can well reflect the perception characteristics of the human ear for different frequencies. When assigning bits, only the amplitude of the part greater than LTQ can be considered, because the human ear cannot perceive the sound below LTQ.

The optimal bit allocation is to apply the psychoacoustic model to the allocation process, so that the limited bit resources can be allocated as much as possible to the sub-bands that are easy to detect and not masked by the human ear, so as to achieve the optimal hearing experience of the human ear. On the other hand, on the premise of not affecting the hearing effect of the human ear, the coding bit rate is reduced as much as possible to achieve high-efficiency coding.

Currently, there are various bit allocation schemes that can be used for bit allocation. Two of the bit allocation schemes are as follows:

1. A perceptual model is used to guide the bit allocation of the subbands. The essence of the perceptual model is the masking threshold, and the estimation of the masking threshold is real-time. The process is as follows: Calculate the sound pressure level (SPL) of each frequency point according to the frequency point range contained in the current input signal, and then calculate the transfer function of each frequency point (spread function), after which the masking threshold is estimated, and the masking threshold of all frequency points finally determines the masking threshold of each subband. After obtaining the perceptual model, bit allocation is performed according to the signal amplitude of each subband and the signal-to-mask ratio (SMR) of the sub-band's masking threshold.

2. Embedded variable rate codec is adopted, and its 4th to 12th layers are time domain aliasing cancellation (TDAC) encoders, and the Modified Discrete Cosine Transform (MDCT, Modified Discrete Cosine Transform) coefficients are divided into 18 sub-bands, and the perceptual importance is sorted according to the quantization energy of each sub-band. The greater the quantization energy, the more important the perception is, and then bit allocation is performed according to the perceptual importance ranking, using the reverse water filling principle The binary search method for optimal bit allocation. Among them, the subbands with high perceptual importance will be preferentially coded and transmitted at lower layers, and will be decoded preferentially at the decoding end; the subbands with low perceptual importance will be coded and transmitted at higher layers, which may be affected by the transmission channel , the higher-level coded information cannot be received at the decoding end, and in this case, it cannot be decoded.

In the process of implementing the present invention, the inventors of the present invention found that in the bit allocation scheme 1, in order to achieve the purpose of high-efficiency coding, it is necessary to establish a fine perceptual model in real time, which requires a large amount of calculation, and then multiple iterations are required. In order to achieve the purpose of optimal bit allocation, the computational complexity of the whole process is very high; in bit allocation scheme 2, the perceptual importance of each sub-band is only measured according to the sub-band energy, and the sensitivity of the human ear to low-frequency signals is not considered The perceptual characteristics of high sensitivity and low sensitivity of high-frequency signals are not considered, and the masking effect of low-frequency signals covering high-frequency signals is not considered, resulting in a mismatch between bit allocation and perceptual characteristics, which affects the quality of encoding and decoding.

Contents of the invention

Embodiments of the present invention provide a bit allocation method, an encoding method, a decoding method, an encoder and a decoder, which can ensure that the codec can produce better subjective auditory effects while reducing the complexity of the perception model.

A bit allocation method, comprising: perceptually sorting all subbands of a signal according to the energy level of the subbands; Order in sorting; allocates bits in the order of the adjusted perceptual sorting.

A coding method, comprising: perceptually sorting all subbands of a signal according to the energy level of the subbands; The sequence in the sequence; the bit resources are allocated according to the adjusted perceptual sorting sequence; according to the bit allocation result, the signal is quantized and coded.

An encoder, comprising: a sorting unit for perceptually sorting all subbands of a signal according to the energy levels of the subbands; an adjustment unit for perceptually sorting the frequency bands corresponding to the subbands in the frequency domain and the energy of the subbands The difference between the subbands is adjusted in the order of the perceptual ordering; the bit allocation unit is used to allocate the bit resources according to the order of the adjusted perceptual ordering; the quantization coding unit is used to allocate the bit resources in the bit allocation unit according to After the adjusted order of perceptual sorting is allocated, the signal is quantized and coded according to the bit allocation result.

A bit allocation method, including: decoding the energy of each subband of the signal from the code stream; perceptually sorting all the subbands according to the energy level of the subbands; according to the level of the frequency band corresponding to the subband in the frequency domain The energy difference between subbands and subbands adjusts the order of the subbands in the perceptual ordering; the bit resources are allocated according to the order of the adjusted perceptual ordering.

A decoding method, including: decoding the energy of each subband of a signal from a code stream; perceptually sorting all subbands according to the energy level of the subbands; Adjust the order of the subbands in the perceptual order according to the energy difference of the bands; allocate the bit resources according to the adjusted order of the perceptual order; decode the signal from the code stream according to the result of the bit allocation.

A decoder, comprising: a first decoding unit, configured to decode the energy of each subband of a signal from a code stream; a sorting unit, configured to perceptually sort all subbands according to the energy level of the subbands; adjust The unit is used to adjust the order of the subbands in the perceptual sorting according to the difference between the frequency band corresponding to the subband in the frequency domain and the energy of the subbands; the bit allocation unit is used to arrange the bit resources according to the adjusted perceptual sorting Allocating in order; the second decoding unit is configured to, after the bit allocation unit allocates the bit resources according to the adjusted perceptual sorting order, decode the code stream to obtain the signal according to the result of the bit allocation.

In the embodiment of the present invention, after perceptually sorting all the subbands of the signal according to the energy levels of the subbands, the subbands can be adjusted according to the difference between the frequency domain level of the frequency band corresponding to the subbands and the energy of the subbands. Order in perceptual ordering, in this case the subbands are actually ordered by their perceptual importance. In this way, when allocating bits, according to the order in which the subbands are arranged, bits may be preferentially allocated to perceptually important subbands.

When the bit resource is insufficient, the embodiment of the present invention can ensure that perceptually important subbands can obtain bits and be quantized to be transmitted to the decoding end, and the decoding end can restore the information of such subbands, and such subbands have a great influence on the auditory experience. The impact is relatively large. Although other perceptual sub-bands may not get bits, and the decoder may not be able to recover the information of such sub-bands, but since such sub-bands have little impact on auditory experience, they will not cause Large auditory error. Therefore, the auditory effect of the codec can be guaranteed through the embodiments of the present invention.

When there are more bit resources, if there are still remaining bits after allocating bits for all subbands, then the perceptually important subbands can be prioritized to allocate bits again, so that the quantization accuracy of such subbands will increase with the increase of bits. The information of such subbands recovered by the decoder will be more accurate and closer to the original signal.

Description of drawings

In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings that need to be used in the description of the embodiments or the prior art. Obviously, the accompanying drawings in the following description are only These are some embodiments of the present invention, and those skilled in the art can also obtain other implementation modes according to these drawings without any creative effort.

Fig. 1 is the flowchart of a kind of bit allocation method of the embodiment of the present invention;

FIG. 2 is a flowchart of an encoding method according to an embodiment of the present invention;

FIG. 3 is a schematic diagram of a logic structure of an encoder according to an embodiment of the present invention;

FIG. 4 is a flowchart of another bit allocation method according to an embodiment of the present invention;

FIG. 5 is a flowchart of a decoding method according to an embodiment of the present invention;

FIG. 6 is a schematic diagram of a logic structure of a decoder according to an embodiment of the present invention.

Detailed ways

The following will clearly and completely describe the technical solutions in the embodiments of the present invention with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only some, not all, embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by persons of ordinary skill in the art without creative efforts fall within the protection scope of the present invention.

The embodiments of the present invention can be described from the perspectives of an encoder and a decoder respectively. Firstly, the embodiment of the present invention is described from the perspective of an encoder.

Firstly, a bit allocation method in the embodiment of the present invention is described. As shown in Figure 1, including:

S101: Perceptually sort all subbands of the signal according to the energy level of the subbands. In practical applications, all subbands of the signal can be sorted according to the energy of the subbands from large to small, that is, the subbands with high energy are arranged before the subbands with small energy, or the subbands can be sorted according to the energy of the subbands. The energy is sorted from small to large, that is, the subbands with small energy are arranged before the subbands with large energy.

The energy can be represented by a spectral envelope or its derived value (eg quantized value, quantized value after LTQ weighting).

S102: Adjust the order of the subbands in the perceptual sorting according to the difference between the frequency bands corresponding to the subbands in the frequency domain and the energy difference of the subbands.

After the sub-bands are preliminarily sorted according to their energy, the order of the sub-bands in the perceptual sorting can also be adjusted according to two factors: the height of the frequency band corresponding to the sub-bands in the frequency domain and the energy difference of the sub-bands. It should be noted that the frequency band here refers to a frequency range, for example, the frequency band of a certain sub-band is 2kHz˜6kHz.

Taking the preliminary sorting of the subbands according to the order of energy from large to small as an example, if the position of any first subband in the perceptual sorting is ranked after the position of the second subband in the perceptual sorting, the The frequency band corresponding to the second subband is adjacent to the frequency band corresponding to the first subband, and the frequency band corresponding to the first subband is lower than the frequency band corresponding to the second subband in the frequency domain, then comparing the The energy of the first subband and the energy of the second subband, if the difference between the energy of the first subband and the energy of the second subband reaches a preset threshold value, swap the second subband A subband and the position of the second subband in perceptual ordering. For example, assuming that the subbands of the frequency band from low to high are: subband 1, subband 2, subband 3, subband 4, ..., after performing perceptual sorting according to the energy of the subbands, each subband is perceptually sorted The order in is: 4, 3, 7, 6, .... If it is necessary to exchange the order of each sub-band in the perceptual ordering, first determine that the position of sub-band 1 in the perceptual ordering is 4, and the position of sub-band 2 adjacent to sub-band 1 in the perceptual ordering is 3; The position of band 1 in the perceptual ranking is after the position of sub-band 2, so judge the energy gap between sub-band 1 and sub-band 2, and if the energy gap reaches the preset threshold value, swap sub-band 1 and sub-band 2 in The position in the perceptual sorting, that is, the position of sub-band 1 in the perceptual sorting is changed to 3, and the position of sub-band 2 in the perceptual sorting is changed to 4; after that, it is determined that the changed position of sub-band 2 in the perceptual sorting is The position of sub-band 3 is before, so it is not exchanged; after that, it is determined that the position of sub-band 3 in the perceptual ranking is after the position of sub-band 4, and then compare the energy gap between sub-band 3 and sub-band 4, if the energy gap does not reach the pre- If the threshold value is set, the positions of sub-band 3 and sub-band 4 in the perceptual ordering will not be exchanged, and so on, and the final adjusted order is: 3, 4, 7, 6, . . .

Taking the preliminary sorting of the subbands according to the order of energy from small to large as an example, if the position of any first subband in the perceptual sorting is ranked before the position of the second subband in the perceptual sorting, the The frequency band corresponding to the second sub-band is adjacent to the frequency band corresponding to the first sub-band, and the frequency band corresponding to the first sub-band is higher in the frequency domain than the frequency band corresponding to the second sub-band, then comparing the The energy of the first subband and the energy of the second subband, if the difference between the energy of the first subband and the energy of the second subband reaches a preset threshold value, the The positions of the first subband and the second subband in perceptual ordering.

Of course, this is just an example to illustrate how to adjust the perceptual ranking according to the difference between the frequency band corresponding to the sub-band in the frequency domain and the energy of the sub-band. The difference between the height of the corresponding frequency band in the frequency domain and the energy of the sub-bands is used to adjust the perceptual ranking, and no examples are given here.

Perceptual ordering is a very important step in the bit allocation process, which determines the order in which bits are allocated to each subband. Or take the preliminary sorting of the subbands according to the order of energy from large to small as an example. When the bit resources are insufficient, the subbands in the lower order may not be allocated bits, and the subbands that cannot be allocated bits will not be quantized. Encoding, the decoder will not be able to obtain the subband information that has not been quantized and encoded, so the quality of this subband will be lost. If perceptual sorting is performed reasonably, important subbands will be assigned bits first, thereby ensuring that important subbands can be quantized and coded, and thus the decoder can obtain information about important subbands.

S103: Allocate bit resources according to the adjusted order of perceptual ranking.

When allocating bits, the subbands are allocated bits according to their positions in the perceptual ranking. Specifically, if the subbands are preliminarily sorted in descending order of energy, bits are allocated to the subbands in sequence from front to back. Still taking the order of subband arrangement "subband 1, subband 2, subband 3, subband 4, subband 5, ..." as an example, first assign bits to subband 1, and then assign bits to subband 2, By analogy, bits are allocated to the subbands in sequence. Of course, if the subbands are preliminarily sorted in ascending order of energy, bits are allocated to the subbands sequentially in order from back to front. Still taking the order of the subbands "subband 1, subband 2, subband 3, subband 4, subband 5, ..." as an example, first assign bits to the last subband, and then assign bits to the last subband Bits are allocated to the second subband, and so on, bits are allocated to subband 5, bits are allocated to subband 4, bits are allocated to subband 3, bits are allocated to subband 2, and finally bits are allocated to subband 1.

In practical applications, after a bit allocation, there may be bits remaining, that is, there are remaining bits that have not been allocated. In this case, the remaining bits can be allocated to the subbands according to the order of the adjusted perceptual sorting , until all remaining bits are allocated. For example, suppose the bit set table of a subband of a certain signal is {6, 9, 12}. The meaning of this bit set table is that when allocating bits to a certain subband of this signal, only this subband can be allocated 6 bits, 9 bits or 12 bits, of course, the maximum number of bits allocated for a subband of this signal is 12. Taking the preliminary sorting of the subbands according to the order of energy from large to small as an example, assuming that in the first bit allocation, 9 bits are allocated to subband 2 of this signal, if after the first bit allocation, there are still There are unallocated bits, and all subbands before subband 2 have undergone the second bit allocation, then when allocating bits for subband 2, 3 bits can be allocated for subband 2, so that the subband The number of bits obtained with 2 reaches the maximum number of bits, which is 12 bits. Of course, how to allocate the remaining bits can also be designed by those skilled in the art according to actual needs, and will not be illustrated here one by one.

Obviously, the execution subject of S101, S102 and S103 is the encoder.

S101, S102, and S103 in the method shown in FIG. 1 may be applied in an encoding process, and for this, an embodiment of the present invention provides an encoding method. As shown in Figure 2, including:

S201: Perceptually sort all subbands of the signal according to the energy level of the subbands;

S202: Adjust the order of the subbands in the perceptual ranking according to the difference between the frequency band corresponding to the subband in the frequency domain and the energy difference of the subband;

S203: Allocate the bit resources according to the adjusted perceptual order;

S204: Quantize and encode the signal according to the bit allocation result.

For the related description of the method shown in Figure 2, please refer to the related description of the method shown in Figure 1, for example, for the related description of S201, please refer to the related description of S101, for the related description of S202, please refer to the related Refer to the relevant description of S103, which will not be repeated here.

Corresponding to the method shown in FIG. 1 and the method shown in FIG. 2 , an embodiment of the present invention provides an encoder. Please refer to FIG. 3 again, including: a sorting unit 301, configured to perceptually sort all subbands of the signal according to the energy level of the subbands; The energy difference of the bands adjusts the order of the subbands in the perceptual ordering; the bit allocation unit 303 is used to allocate the bit resources according to the order of the adjusted perceptual ordering; the quantization encoding unit 304 is used for the bit allocation unit 303 After the bit resources are allocated according to the adjusted order of perceptual ordering, the signal is quantized and coded according to the bit allocation result.

For the relevant description of the encoder shown in FIG. 3, please refer to the relevant description of the method shown in FIG. 1. For example, for the relevant description of the sorting unit 301, please refer to the relevant description of S101. For the related description of the bit allocating unit 303, refer to the related description of S103, which will not be repeated here.

It should be noted here that when encoding, the encoder may encode the quantized value of the spectral envelope of each subband, so that the quantized value representing the energy of each subband is encoded into the code stream.

The method shown in Figure 1, the method shown in Figure 2, and the encoder shown in Figure 3 are all described from the perspective of the encoder. As mentioned above, the embodiment of the present invention can also be described from the perspective of the decoder .

First, a bit allocation method in the embodiment of the present invention is described from the perspective of a decoder. As shown in Figure 4, including:

S401: Decode the code stream to obtain the energy of each subband of the signal.

After the encoder encodes the quantized value of the spectral envelope of each subband, the quantized value is encoded into the code stream. Correspondingly, the decoder can decode the code stream to obtain the quantized value of the spectral envelope of each subband, so as to obtain the energy of each subband.

S402: Perceptually sort all the subbands according to the energy level of the subbands.

For related descriptions, refer to related descriptions of S101 in the method shown in FIG. 1 .

S403: Adjust the order of the subbands in the perceptual ranking according to the difference between the frequency bands corresponding to the subbands in the frequency domain and the energy difference of the subbands.

For related descriptions, refer to the related descriptions of S102 in the method shown in FIG. 1 .

S404: Allocate bit resources according to the adjusted order of perceptual ranking.

For related descriptions, refer to related descriptions of S103 in the method shown in FIG. 1 .

S401, S402, S403, and S404 in the method shown in FIG. 4 may be applied in a decoding process, and for this, an embodiment of the present invention further provides a decoding method. As shown in Figure 5, including:

S501: Decode the code stream to obtain the energy of each subband of the signal.

For related descriptions, refer to related descriptions of S401 in the method shown in FIG. 4 .

S502: Perceptually sort all the subbands according to the energy level of the subbands.

For related descriptions, refer to related descriptions of S101 in the method shown in FIG. 1 .

S503: Adjust the order of the subbands in the perceptual sorting according to the difference between the frequency bands corresponding to the subbands in the frequency domain and the energy difference of the subbands.

For related descriptions, refer to the related descriptions of S102 in the method shown in FIG. 1 .

S504: Allocate the bit resources according to the adjusted perceptual order.

For related descriptions, refer to related descriptions of S103 in the method shown in FIG. 1 .

S505: According to the bit allocation result, decode the code stream to obtain a signal.

Corresponding to the method shown in FIG. 4 and the method shown in FIG. 5 , an embodiment of the present invention provides a decoder. Please refer to FIG. 6, including: a first decoding unit 601, which is used to decode the energy of each subband of the signal from the code stream; a sorting unit 602, which is used to perceptually sort all subbands according to the energy level of the subbands The adjustment unit 603 is used to adjust the order of the sub-bands in the perceptual sorting according to the difference between the frequency bands corresponding to the sub-bands in the frequency domain and the energy of the sub-bands; the bit allocation unit 604 is used to allocate the bit resources according to the adjusted The second decoding unit 605 is configured to, after the bit allocation unit 604 allocates the bit resources according to the adjusted perceptual order, decode the bit stream to obtain the signal according to the bit allocation result.

For the relevant description of the decoder shown in Figure 6, please refer to the relevant description of the method shown in Figure 1 and the method shown in Figure 4, for example, for the relevant description of the first decoding unit 601, please refer to S401 in the method shown in Figure 4 For the relevant description of the sorting unit 602, please refer to the relevant description of S101 in the method shown in FIG. 1. For the relevant description of the adjustment unit 603, please refer to the relevant description of S102. description, and will not be repeated here.

In order to make those skilled in the art understand the embodiment of the present invention more clearly, a preferred embodiment is introduced below.

At the encoding end, the obtained set of M-point MDCT coefficients Y={y(j), j=0,...,K} is divided into N subbands band(i), i=0,..., N-1. The number of corresponding MDCT coefficients in each subband is nb_coef(i), $\underset{i}{\mathrm{\Σ}}\mathrm{nb}\_\mathrm{coef}\left(i\right)=m,i=0,...,N-1,$ The spectral envelope of each subband is log_rms(i), and the number of bits available for quantization of MDCT coefficients at the encoding end is nb_bits_max. The method of bit allocation is as follows:

Encoder:

(1) All subband spectral envelopes are quantized and coded first, and put into the code stream.

(2) Sort all subbands according to the order of spectral envelope log_rms(i) from large to small, the sorting result is ord[i], ord[i]=0,..., N-1, spectral envelope The larger the value, the smaller the ord[i] and the more important the perception.

(3) According to the order from low frequency to high frequency, judge whether the perceptual sequence of each sub-band conforms to the perceptual characteristics in turn, and the judgment process is as follows:

If the perceptual order ord[i] of the i-th sub-band is greater than the perceptual order ord[i+1] of the i+1-th sub-band

If the difference between the spectral envelope log_rms(i) of the ith subband and the spectral envelope log_rms(i) of the i+1th subband>Thre

Swap the perceptual order of the ith subband and the i+1th subband

end

end

It should be noted that: a) The sub-bands are arranged in order from low frequency to high frequency, therefore, the i-th sub-band is the low-frequency band relative to the i+1-th sub-band, and the i+1-th sub-band is relative to the i+2-th sub-band for the low frequency band, and so on. b) Thre is the threshold of the energy difference, which can be determined by using empirical values or optimization algorithms, Thre∈R, in this embodiment, Thre=0.5.

(4) Assign bits to each subband sequentially according to the adjusted perceptual ranking. The theoretical number of bits that should be allocated to each MDCT coefficient in a subband is directly determined by the spectral envelope of the subband. The theoretical number of bits that should be allocated to each subband is:

bit_rms(i)=log_rms(i)×nb_coef(i)

According to the maximum number of bits nb_bits_max, the optimal bit allocation for each subband is:

$\mathrm{<span\; class="notranslate">\; bat</span>}<span\; class="notranslate">\; \_</span>\mathrm{<span\; class="notranslate">\; opt</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span><span\; class="notranslate">\; [</span>\frac{\mathrm{<span\; class="notranslate">\; nb</span>}<span\; class="notranslate">\; \_</span>\mathrm{<span\; class="notranslate">\; bits</span>}<span\; class="notranslate">\; \_</span>\mathrm{<span\; class="notranslate">\; max</span>}}{\mathrm{<span\; class="notranslate">\; m</span>}}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; log</span>}<span\; class="notranslate">\; \_</span>\mathrm{<span\; class="notranslate">\; rms</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; -</span>\frac{\mathrm{<span\; class="notranslate">\; \&Sigma;\; bit</span>}<span\; class="notranslate">\; \_</span>\mathrm{<span\; class="notranslate">\; rms</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; )</span>}{\mathrm{<span\; class="notranslate">\; m</span>}}<span\; class="notranslate">\; ]</span><span\; class="notranslate">\; \&times;</span>\mathrm{<span\; class="notranslate">\; nb</span>}<span\; class="notranslate">\; \_</span>\mathrm{<span\; class="notranslate">\; coef</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; )</span>$

(5) If there are still bit resources remaining, the process of reallocating according to the perceptual importance of subbands is as follows:

a) If the number of remaining bit resources allocatedbits<nb_bits_max and 0≤i<N, repeat steps b to d.

b) Select the subbands band(ord[i]) in order of perceived importance, from important to unimportant.

c) Known possible bit allocation set R={q_bit ₀ , q_bit ₁ , ...}, where q_bit ₀ <q_bit ₁ <... is the number of bits that may be allocated to a subband. If the bit resource allocated to the subband is bat_bit(ord[i])=q_bit _k , if the allocated bits are added to the subband, considering the known possible allocation to the bit set R, then the subband is allocated to The bits shall be q_bit _k+1 .

d) If allocatedbits+q_bits _k+1 -q_bits _k ≤ nb_bits_max, then bat_bit(ord[i])=q_bit _k+1 , allocatedbits=allocatedbits+q_bits _k+1 -q_bits _k , otherwise bat_bit(ord[i])= q_bit _k .

(6) Quantize and encode the MDCT coefficients of each subband according to the final bit allocation result bat_bit(i), and encode them into the code stream.

Decoder:

(1) Decode the received code stream to obtain the spectral envelope log_rms(i).

(2) Determine the number of bits bat_bit(i) allocated to each subband and the order of importance ord[i] of the subbands by the same means as those at the encoding end (2)-(5).

(3) sort ord[i] according to the number of bits bat_bit(i) assigned to each sub-band and the importance of the sub-bands, and sequentially decode the received code stream to obtain the MDCT coefficients of each sub-band.

In summary, the embodiment of the present invention combines the influence of the energy and frequency of the subbands to determine the perceptual ordering of the subbands, ensuring that the perceptually important subbands are given priority to bit resources, and the subbands that are sensitive to the human ear Quantization errors are reduced, greatly improving listening quality. At the same time, the complex calculation of the psychoacoustic model is avoided, the calculation complexity is effectively reduced, the delay is small, and the resource consumption is small.

Those of ordinary skill in the art can understand that all or part of the processes in the methods of the above embodiments can be implemented through computer programs to instruct related hardware, and the programs can be stored in a computer-readable storage medium. During execution, it may include the processes of the embodiments of the above-mentioned methods. Wherein, the storage medium may be a magnetic disk, an optical disk, a read-only memory (Read-Only Memory, ROM) or a random access memory (Random Access Memory, RAM), etc.

The above is only a preferred embodiment of the present invention, it should be pointed out that for those of ordinary skill in the art, without departing from the principle of the present invention, some improvements and modifications can also be made, and these improvements and modifications should also be It is regarded as the protection scope of the present invention.

Claims (22)

1. a Bit distribution method is characterized in that, comprising:

All subbands of signal are carried out the perception ordering according to the height of the energy of subband;

According to the difference of the energy of the corresponding height of frequency range on frequency domain of subband and subband, the order of adjustment subband in perception is sorted;

The bit resource is distributed according to the order of adjusted perception ordering.

2. the method for claim 1 is characterized in that, all subbands of signal are carried out the perception ordering according to the height of the energy of subband be specially: all subbands to signal carry out the perception ordering according to the descending order of the energy of subband.

3. method as claimed in claim 2 is characterized in that, according to the difference of the energy of the corresponding height of frequency range on frequency domain of subband and subband, the order of adjustment subband in perception is sorted specifically comprises:

If the position of any one first subband in said perception ordering comes after the position of second subband in said perception ordering; Frequency range and the corresponding adjacent frequency bands of said first subband that said second subband is corresponding; And the frequency range that said first subband is corresponding is lower than the corresponding frequency range of said second subband on frequency domain; The energy of the energy of then more said first subband and said second subband; If the gap of the energy of the energy of said first subband and said second subband has reached the threshold value that is provided with in advance, then change said first subband and said second subband position in the perception ordering.

4. like claim 2 or 3 described methods, it is characterized in that, the bit resource is distributed according to the order of adjusted perception ordering comprise: according to from front to back order allocation bit resource successively in the perception ordering.

5. method as claimed in claim 4; It is characterized in that; The bit resource distributed also according to the order of adjusted perception ordering comprise: if all distributed bit for each subband but also have unappropriated remaining bits; Then, all be assigned with until all remaining bits again according to the said unappropriated remaining bits of order assignment of said adjusted perception ordering.

6. the method for claim 1 is characterized in that, all subbands of signal are carried out the perception ordering according to the height of the energy of subband be specially: all subbands to signal carry out the perception ordering according to the ascending order of the energy of subband.

7. method as claimed in claim 6 is characterized in that, according to the difference of the energy of the corresponding height of frequency range on frequency domain of subband and subband, the order of adjustment subband in perception is sorted specifically comprises:

If the position of any one first subband in said perception ordering comes before the position of second subband in said perception ordering; Frequency range and the corresponding adjacent frequency bands of said first subband that said second subband is corresponding; And the frequency range that said first subband is corresponding is higher than the corresponding frequency range of said second subband on frequency domain; The energy of the energy of then more said first subband and said second subband; If the gap of the energy of the energy of said first subband and said second subband has reached the threshold value that is provided with in advance, then change said first subband and said second subband position in the perception ordering.

8. like claim 6 or 7 described methods, it is characterized in that, the bit resource distributed according to the order of adjusted perception ordering comprise: according in the perception ordering from the back to preceding order allocation bit resource successively.

9. method as claimed in claim 8; It is characterized in that; The bit resource distributed also according to the order of adjusted perception ordering comprise: if all distributed bit for each subband but also have unappropriated remaining bits; Then, all be assigned with until all remaining bits again according to the said unappropriated remaining bits of order assignment of said adjusted perception ordering.

10. a coding method is characterized in that, comprising:

All subbands of signal are carried out the perception ordering according to the height of the energy of subband;

According to the difference of the energy of the corresponding height of frequency range on frequency domain of subband and subband, the order of adjustment subband in perception is sorted;

The bit resource is distributed according to the order of adjusted perception ordering;

According to the result of Bit Allocation in Discrete, signal is carried out quantization encoding.

11. a scrambler is characterized in that, comprising:

Sequencing unit is used for all subbands of signal are carried out the perception ordering according to the height of the energy of subband;

Adjustment unit is used for the difference according to the energy of the corresponding height of frequency range on frequency domain of subband and subband, the order of adjustment subband in perception is sorted;

The Bit Allocation in Discrete unit is used for the bit resource is distributed according to the order of adjusted perception ordering;

The quantization encoding unit is used for said Bit Allocation in Discrete unit distributes the bit resource according to the order of adjusted perception ordering after, according to the result of Bit Allocation in Discrete, signal being carried out quantization encoding.

12. a Bit distribution method is characterized in that, comprising:

Decoding obtains the energy of each subband of signal from code stream;

All subbands are carried out the perception ordering according to the height of the energy of subband;

According to the difference of the energy of the corresponding height of frequency range on frequency domain of subband and subband, the order of adjustment subband in perception is sorted;

The bit resource is distributed according to the order of adjusted perception ordering.

13. method as claimed in claim 12 is characterized in that, all subbands of signal are carried out the perception ordering according to the height of the energy of subband be specially: all subbands to signal carry out the perception ordering according to the descending order of the energy of subband.

14. method as claimed in claim 13 is characterized in that, according to the difference of the energy of the corresponding height of frequency range on frequency domain of subband and subband, the order of adjustment subband in perception is sorted specifically comprises:

If the position of any one first subband in said perception ordering comes after the position of second subband in said perception ordering; Frequency range and the corresponding adjacent frequency bands of said first subband that said second subband is corresponding; And the frequency range that said first subband is corresponding is lower than the corresponding frequency range of said second subband on frequency domain; The energy of the energy of then more said first subband and said second subband; If the gap of the energy of the energy of said first subband and said second subband has reached the threshold value that is provided with in advance, then change said first subband and said second subband position in the perception ordering.

15. like claim 13 or 14 described methods, it is characterized in that, the bit resource distributed according to the order of adjusted perception ordering comprise: according to from front to back order allocation bit resource successively in the perception ordering.

16. method as claimed in claim 15; It is characterized in that; The bit resource distributed according to the order of adjusted perception ordering comprise: if all distributed bit for each subband but also have unappropriated remaining bits; Then, all be assigned with until all remaining bits again according to the said unappropriated remaining bits of order assignment of said adjusted perception ordering.

17. method as claimed in claim 12 is characterized in that, all subbands of signal are carried out the perception ordering according to the height of the energy of subband be specially: all subbands to signal carry out the perception ordering according to the ascending order of the energy of subband.

18. method as claimed in claim 17 is characterized in that, according to the difference of the energy of the corresponding height of frequency range on frequency domain of subband and subband, the order of adjustment subband in perception is sorted specifically comprises:

If the position of any one first subband in said perception ordering comes before the position of second subband in said perception ordering; Frequency range and the corresponding adjacent frequency bands of said first subband that said second subband is corresponding; And the frequency range that said first subband is corresponding is higher than the corresponding frequency range of said second subband on frequency domain; The energy of the energy of then more said first subband and said second subband; If the gap of the energy of the energy of said first subband and said second subband has reached the threshold value that is provided with in advance, then change said first subband and said second subband position in the perception ordering.

19. like claim 17 or 18 described methods, it is characterized in that, the bit resource distributed according to the order of adjusted perception ordering comprise: according in the perception ordering from the back to preceding order allocation bit resource successively.

20. method as claimed in claim 19; It is characterized in that; The bit resource distributed also according to the order of adjusted perception ordering comprise: if all distributed bit for each subband but also have unappropriated remaining bits; Then, all be assigned with until all remaining bits again according to the said unappropriated remaining bits of order assignment of said adjusted perception ordering.

21. a coding/decoding method is characterized in that, comprising:

Decoding obtains the energy of each subband of signal from code stream;

All subbands are carried out the perception ordering according to the height of the energy of subband;

According to the difference of the energy of the corresponding height of frequency range on frequency domain of subband and subband, the order of adjustment subband in perception is sorted;

The bit resource is distributed according to the order of adjusted perception ordering;

According to the result of Bit Allocation in Discrete, decoding obtains signal from code stream.

22. a demoder is characterized in that, comprising:

First decoding unit is used for obtaining from code stream decoding the energy of each subband of signal;

Sequencing unit is used for all subbands are carried out the perception ordering according to the height of the energy of subband;

Adjustment unit is used for the difference according to the energy of the corresponding height of frequency range on frequency domain of subband and subband, the order of adjustment subband in perception is sorted;

The Bit Allocation in Discrete unit is used for the bit resource is distributed according to the order of adjusted perception ordering;

Second decoding unit is used for said Bit Allocation in Discrete unit distributes the bit resource according to the order of adjusted perception ordering after, and according to the result of Bit Allocation in Discrete, decoding obtains signal from code stream.

Images