JPWO2012002348A1 - Bit allocation apparatus, method, program, and recording medium thereof - Google Patents

Abstract

The process of adding 1 to each bit counter corresponding to the indices k (i),..., K (i) − (L−1) is performed for each of i = 0, 1,. For each sample i having an index greater than or equal to a value Thres satisfying the relationship of Σt = Thres−1 Mbitcount (t)> Maxbit ≧ Σt = ThresMbitcount (t), the smaller number of k (i) −Thres + 1 and L Assign bits. One bit is assigned to each Maxbit−Σt = ThresMbitcount (t) samples among samples having an index equal to or greater than the value Thres−1 and not yet assigned L bits.

Description

  The present invention relates to a technique for encoding sound signals such as voice and music.

  When encoding a sound signal such as speech or music, there is a technique for calculating the number of bits necessary for each piece of information to be encoded and encoding the information with the calculated number of bits. For example, the ITU-T standard G.I. A bit allocation technique used in the low band extension coding of 711.1 is known. In Non-Patent Document 1, L is the maximum number of bits that can be assigned to one sample, and Maxbit bits are assigned to N samples in the frame. With these assigned bits, the low-frequency input signal and G. The difference signal with the 711 decoded signal is encoded.

  Hereinafter, an overview of the allocation method of Non-Patent Document 1 will be described. First, k (n) that is an integer greater than or equal to 0 is assigned to each sample n (n = 0, 1,..., N−1). Here, the sample value of the sample n is x (n), and int (·) is a function for truncating the decimal part of •, k (n) = int (log | x (n) |).

  Next, a table of a two-dimensional array composed of (M + L) × L elements is created based on information on a set of sample numbers n and k (n). M is the maximum value of k (n). Specifically, the leftmost column to which the sample number of k (n) +2 rows is not assigned, the leftmost column to which the sample number of k (n) +1 rows is not assigned, and k (n) rows The process of assigning the sample number n to each of the leftmost columns to which no sample number is assigned is performed for each of n = 0, 1,..., N−1.

  For example, as shown in FIG. 12, it is assumed that a sample number and its k (n) value are given. In this case, a table as shown in FIG. 13 is created. In the examples of FIGS. 12 and 13, it is assumed that L = 3, N = 10, and Maxbit = 10.

  Next, the elements are confirmed in order from the 0th column to the N−1th column from the (M + L−1) th row to the 0th row. When the sample number is assigned to the element, 1 is added to the assigned bit of the sample of the sample number. This operation is repeated until the total number of assigned bits reaches 10 (= Maxbit).

  In the example of FIG. 13, 1 bit is assigned to each sample of the sample number surrounded by a thick line. As a result, bits are assigned as shown in the bottom column of FIG.

Rec. ITU-T G.711.1 (03/2008) “7.3.4.1 Generation of bits allocation table”

  In order to assign bits by the method described in Non-Patent Document 1, it is necessary to prepare a (M + L) × N two-dimensional array. When L, M, and N are increased, there is a problem that the arrangement is increased and the amount of memory consumption is increased.

  An object of the present invention is to provide a bit allocation device, a method, a program, and a recording medium thereof, which consume less memory than before.

According to the bit allocating device of one aspect of the present invention, the maximum number of bits that can be allocated to one sample is L, and in order to allocate Maxbit bits to N samples, the sample number of N samples is set to 0. , 1,..., I,..., N−1, an index representing the importance given to the sample of sample number i is k (i), and the number of possible values of index k (i) + L−1 , Bitcount (k (i)),..., Bitcount (k (i) corresponding to the indices k (i),..., K (i)-(L-1). )-(L-1)) is added to each of i = 0, 1,..., N-1. For each sample i having an index greater than or equal to a value T satisfying the relationship Σ _{t = T−1} ^M bitcount (t)> Maxbit ≧ Σt _{= T} ^M bitcount (t), where M is the maximum value of the index k (i). On the other hand, the smaller number of bits of k (i) -T + 1 and L are allocated. One bit is assigned to each Maxbit-Σt _{= T} ^M bitcount (t) samples among samples having an index equal to or greater than the value T−1 and not yet assigned L bits.
According to the bit allocating device according to another aspect of the present invention, the maximum number of bits that can be allocated to one sample is L, and Maxbit bits are allocated to N samples. The N samples are rearranged in the order of importance based on the index indicating the degree of importance. The sample numbers of the rearranged N samples are newly set to 0, 1,..., I,..., N−1, and the index indicating the importance given to the sample with the sample number i is k (i). The number of possible values of the index k (i) + the initial value of the L−1 bit counter bitcount is 0, and each bit counter corresponding to the index k (i),..., K (i) − (L−1) .., bitcount (k (i) − (L−1)) is incremented by 1 for each of i = 0, 1,..., N−1. For each sample i having an index greater than or equal to a value T satisfying the relationship Σ _{t = T−1} ^M bitcount (t)> Maxbit ≧ Σt _{= T} ^M bitcount (t), where M is the maximum value of the index k (i). On the other hand, the smaller number of bits of k (i) -T + 1 and L are allocated. Among samples having an index equal to or greater than the value T-1 and not yet assigned L bits, Maxbit-Σt _{= T} ^M bitcount (t) samples from the one with the smaller sample number before rearrangement. On the other hand, one bit is allocated.

  By allocating bits using a one-dimensional array instead of a two-dimensional array, it is possible to reduce the memory consumption compared to the conventional one.

The functional block diagram of the example of a bit allocation apparatus. The flowchart which shows the example of the bit allocation method. The flowchart which shows the example of step S2. The flowchart which shows the example of step S3. The flowchart which shows the example of step S4. The flowchart which shows the example of step S5. The figure for demonstrating the example of the bit allocation by this invention. The figure for demonstrating the example of the bit allocation by this invention. The figure for demonstrating the example of the bit allocation by this invention. The figure for demonstrating the example of the bit allocation by this invention. The figure for demonstrating the example of the bit allocation by this invention. The figure for demonstrating the example of the bit allocation by a prior art. The figure for demonstrating the example of the bit allocation by a prior art.

  Hereinafter, a bit allocation apparatus and method according to an embodiment of the present invention will be described in detail.

  In the present invention, the maximum number of bits that can be assigned to one sample is L, and Maxbit bits are assigned to N samples. L, N and Maxbit are predetermined positive integers.

  As shown in FIG. 1, the bit allocation device includes, for example, a sorting unit 10, a bit count unit 20, a threshold value determination unit 30, a first allocation unit 40, and a second allocation unit 50. As shown in FIG. 2, the bit allocation method includes, for example, steps S1 to S5.

  N samples are assigned sample numbers 0, 1,..., N,. The sample of sample number n is represented as sample n, where n = 0, 1,. Each sample 0, 1,..., N,..., N−1 is associated with an index representing the importance of each sample. As n = 0, 1,..., N−1, an index representing the importance of the sample n is represented as k (n). k (n) is, for example, an output value f (log | x (n) |) when a log value having a magnitude of the sample value x (n) of the sample n is input to a function f described later. The base of log is 2. f (·) is a function that rounds down, rounds up, or rounds off the decimal point of •, or a function that outputs the maximum integer less than or equal to •. k (n) is an output value f () when a value obtained by adding a predetermined constant c to logx (n) is input to the function f so that the log value of the sample value x (n) of the sample n becomes positive. log | x (n) | + c).

<Step S1>
Sample and index pairs (0, k (0)), (1, k (1)),..., (N−1, k (N−1)) are prepared and input to the sorting unit 10. . The sorting unit 10 rearranges the N samples 0, 1,..., N−1 in an important order based on the indices k (0), k (1),..., K (N−1) (Step S1). ). If a more important sample is given an index having a larger value, the samples 0, 1,..., N−1 have the indices k (0), k (1),. Rearranged in descending order.

  The samples after the rearrangement are assigned new sample numbers 0, 1,..., I,. The original sample number assigned to the sample with the new sample number i is substituted into the variable pos (i). Further, k (pos (i)) is assigned to the variable exp (i). That is, exp (i) = k (pos (i)).

  As illustrated in FIG. 7, the set of the sample n and the index k (n) is (0,5), (1,5), (2,3), (3,3), (4,2), In the case of (5, 7), (6, 6), (7, 2), (8, 7), (9, 5), the set of these sample n and index k (n) is shown in FIG. Sorted as shown in

  At least the number of possible values of the index k (i) + L−1 bit counters bitcount are prepared. Each bit counter bitcount is a variable that has an initial value of 0 and takes an integer value. As i = 0, 1,..., N−1, the bit counter corresponding to the index k (i) is expressed as bitcount (k (i)).

<Step S2>
The bit count unit 20 includes bit counters (bit (k (i)),..., Bitcount (k (i) − (L−) corresponding to the indices k (i),..., K (i) − (L−1)). 1)) is added to each of i = 0, 1,..., N−1 (step S2).

  The bit count unit 20 performs step S2 by performing, for example, steps S21 to S27 shown in FIG.

  The bit count unit 20 substitutes 0 for i (step S21).

  The bit count unit 20 substitutes 0 for j (step S22).

  The bit count unit 20 adds 1 to the bit counter bitcount (exp (i) -j) (step S23).

  The bit count unit 20 determines whether j = L−1 (step S24).

  If j = L−1 is not true, the bit count unit 20 substitutes j + 1 for j (step S25), and then proceeds to step S23.

  If j = L-1, the bit count unit 20 determines whether i = N-1 (step S26).

  If j = N−1 is not true, the bit count unit 20 substitutes i + 1 for i (step S27), and then proceeds to step S22.

  If j = N-1, step S2 is ended and the process proceeds to step S3 described later.

  When pairs of samples and indices are given as shown in FIGS. 7 and 8, and Maxbit = 10 and L = 3, the value of each bit counter bitcount is calculated as shown in the center of FIG. The left side of FIG. 9 represents the position of the bit counter bitcount to which 1 is added. The left side of FIG. 9 is an image diagram for explaining the calculation of the bit counter bitcount. Actually, such a two-dimensional array table is unnecessary.

<Step S3>
Threshold determination section 30 determines the threshold Thres satisfying the relation ^{_{Σ t = Thres-1 M bitcount}} (t)> Maxbit ≧ Σ t = Thres M bitcount (t) ( step S3). M is the maximum value of the index k (i). Information about the determined threshold value Thres is sent to the first allocation unit 40.
More precisely, the above relationship can be expressed as the following formula (1). Note that Thres may be written as T.

  The threshold value determination unit 30 performs step S3 by performing, for example, step S31 to step S35 illustrated in FIG.

  The threshold value determination unit 30 substitutes M for i, substitutes Maxbit for bitsrv, and substitutes bitcount (M) for bitsum (step S31). bitsrv indicates the number of assignable bits.

  The threshold value determination unit 30 substitutes i−1 for i (step S32).

  The threshold value determination unit 30 compares bitsum and bitsrv (step S33).

  If bitsum> bitsrv is not satisfied, the threshold determination unit 30 substitutes bitsum + bitcount (i) for bitsum (step S34), and proceeds to step S32.

  If bitsum> bitsrv, the threshold determination unit 30 substitutes bitsum-bitcount (i + 1) for bitsum and i + 2 for Thres (step S35). Then, it progresses to step S4 mentioned later.

Threshold Thres obtained by Step S35 from Step S31 satisfy the relationship of ^{_{Σ t = Thres-1 M bitcount}} (t)> Maxbit ≧ Σ t = Thres M bitcount (t).

  In the example of FIG. 9, the bitsum calculated in step S <b> 34 is 11 when i = 5, and the bitsum exceeds the Maxbit value 10 in step S <b> 33 when i = 4 thereafter. Therefore, the threshold value Thres is 4 + 2 = 6.

  As shown by a broken line in FIG. 4, instead of step S32, step S321 for substituting i-1 for i after step S31 and step S322 for substituting i-1 for i after step S34 are used as threshold values. The determination unit 30 may execute.

<Step S4>
The first assigning unit 40 assigns the smaller number of bits of k (i) −Thres + 1 and L to each sample i having an index equal to or greater than the threshold Thres (step S4).

  The first allocation unit 40 performs Step S4 by performing, for example, Step S41 to Step S47 illustrated in FIG.

  The first assigning unit 40 substitutes 0 for i, and substitutes 0 for bitalloc (0), bitalloc (1),..., Bitalloc (N−1) (step S41). bitalloc (i) indicates the number of bits assigned to the sample of sample number i.

  The first allocation unit 40 compares exp (i) with a threshold value Thres (step S42).

  If exp (i) ≦ Thres−1, step S4 is ended, and the process proceeds to step S6 described later.

If exp (i) ≦ Thres−1 is not satisfied, the first assigning unit 40 substitutes the smaller number of exp (i) −Thres + 1 and L into bitalloc (pos (i)) (step S43). In other words, the first allocation unit 40 allocates min (exp (i) −Thres + 1, L) bits to the sample of the sample number pos (i). min (x ₁ , x ₂ ) is a function that outputs the smaller number of x ₁ and x ₂ .

  The first allocation unit 40 substitutes bitsrv-bitalloc (pos (i)) for bitsrv (step S44). That is, the first allocating unit 40 subtracts the number of bits allocated to the sample of the sample number pos (i) bitloc (pos (i)) from the number of bits that can be allocated bitsrv.

  The first allocation unit 40 compares bitalloc (pos (i)) with L (step S45).

  If binaryloc (pos (i)) <L, the first allocation unit 40 substitutes Thres-1 for k (pos (i)) (step S46). Thereafter, the process proceeds to step S47.

  If binaryloc (pos (i)) <L, or after step S46, the first assigning unit 40 substitutes i + 1 for i (step S47). Thereafter, the process proceeds to step S42.

  In this example, the smaller number of bits of exp (i) −Thres + 1 and L is assigned to each sample i having a value equal to or greater than the threshold Thres, but exp (i) = k (pos (i) ), Assigning a smaller number of bits of exp (i) −Thres + 1 and L to each sample i having a value greater than or equal to the threshold Thres, and having a value greater than or equal to the threshold Thres Assigning the smaller number of bits of k (i) -Thres + 1 and L to each sample i is synonymous.

  An example of bit allocation by the first allocation unit 40 corresponding to the examples of FIGS. 7 to 9 is shown in FIG. As shown in FIG. 10, the smaller number of bits of exp (i) −Thres + 1 and L is assigned to each sample i having a value of threshold Thres = 6 or more. According to the definition of the threshold Thres, if exp (i) ≧ Thres, bitsum does not exceed Maxbit. Therefore, there is no problem even if bits are assigned to each sample i having an index equal to or greater than the threshold Thres.

<Step S5>
The second assigning unit 50 has 1 for each Maxbit _{-Σt = Thres} ^M bitcount (t) samples among samples having an index equal to or greater than the threshold Thres-1 and not yet assigned L bits. Bits are allocated (step S51).
That is, the second allocation unit 50 has one index for the number of samples defined by the following equation (2) among samples having an index equal to or greater than the threshold Thres-1 and not yet allocated L bits. Allocate bits.

  The second allocation unit 50 performs Step S5 by performing, for example, Step S51 to Step S54 illustrated in FIG.

The second allocation unit 50 substitutes 0 for i (step S51).
The second allocation unit 50 compares bitsrv with 0 (step S52).

  If bitsrv> 0 is not satisfied, the second assignment unit 50 ends the process. The value of bit loc (i) at this time is the number of bits finally assigned to the sample of sample number i.

  If bitsrv> 0, the second allocation unit 50 determines whether k (i) = Thres−1 and bitalloc (i) <L (step S53).

  If k (i) = Thres−1 and bitalloc (i) <L, the second assigning unit 50 substitutes bitloc (i) +1 for bitloc (i) and bitsbit−1 for bitsrv (step) S54). Thereafter, the process proceeds to step S55.

  If k (i) = Thres−1 and bitalloc (i) <L, or after step S54, the second assigning unit 50 substitutes i + 1 for i (step S55). Thereafter, the process proceeds to step S52.

An example of bit allocation by the second allocation unit 50 corresponding to the examples of FIGS. 7 to 10 is shown in FIG. As shown in FIG. 11, Maxbit−Σ _{t =} from the sample having a smaller sample number before rearrangement among samples having an index equal to or greater than the threshold Thres−1 (5 in this example) and not yet assigned L bits. _One bit is assigned to each of _Thres ^M bitcount (t) samples (5 in this example).
Note that the samples to which one bit is assigned are Maxbit-Σt _{= Thres} ^M bitcount (t) of samples having an index equal to or greater than the threshold Thres-1 and not yet assigned L bits. if the sample, may not be ^{_{Maxbit-Σ t = Thres M bitcount}} (t) samples from people Sort front of the sample number is small.

  In this way, bit allocation is performed using the bit counter bitcount (i), which is a one-dimensional array, instead of the two-dimensional array used in the prior art. While realizing the same bit allocation result as the bit allocation used in the low-frequency extension encoding of 711.1, it is possible to reduce the memory consumption compared to the conventional one. The effect increases as the values of M, L, and N increase.

  The bit allocation apparatus and method can be realized by a computer. In this case, each part of the bit allocation device and each step of the bit allocation method are described by a program. Then, by executing this program on a computer, each part of the bit allocation device and each step of the bit allocation method are realized on the computer.

  This program can be recorded on a computer-readable recording medium. In this embodiment, these devices are configured by executing a predetermined program on a computer. However, at least a part of these processing contents may be realized by hardware.

  The present invention is not limited to the above-described embodiment, and can be modified as appropriate without departing from the spirit of the present invention.

  For example, in the above embodiment, after the samples are sorted in descending order in step S1, the threshold Thres is calculated by the processing in steps S2 and S3. However, the sorting in step S1 may not be performed.

20 bit count unit 30 threshold determination unit 40 first allocation unit 50 second allocation unit

According to the bit allocating device of one aspect of the present invention, the maximum number of bits that can be allocated to one sample is L, and in order to allocate Maxbit bits to N samples, the sample number of N samples is set to 0. , 1,..., I,..., N−1, an index representing the importance given to the sample of sample number i is k (i), and the number of possible values of index k (i) + L−1 the bit counter initial value 0, the index k (i), ..., k (i) - (L-1) to each corresponding bit counter bitcount (k (i)), ..., bitcount (k (i )-(L-1)) is added to each of i = 0, 1,..., N-1. For each sample i having an index greater than or equal to a value T satisfying the relationship Σ _{t = T−1} ^M bitcount (t)> Maxbit ≧ Σt _{= T} ^M bitcount (t), where M is the maximum value of the index k (i). On the other hand, the smaller number of bits of k (i) -T + 1 and L are allocated. One bit is assigned to each Maxbit-Σt _{= T} ^M bitcount (t) samples among samples having an index equal to or greater than the value T−1 and not yet assigned L bits.
According to the bit allocating device according to another aspect of the present invention, the maximum number of bits that can be allocated to one sample is L, and Maxbit bits are allocated to N samples. The N samples are rearranged in the order of importance based on the index indicating the degree of importance. The sample numbers of the rearranged N samples are newly set to 0, 1,..., I,..., N−1, and the index indicating the importance given to the sample with the sample number i is k (i). The number of possible values of the index k (i) + the initial value of the L−1 bit counter bitcount is 0, and each bit counter corresponding to the index k (i),..., K (i) − (L−1) .., bitcount (k (i) − (L−1)) is incremented by 1 for each of i = 0, 1,..., N−1. For each sample i having an index greater than or equal to a value T satisfying the relationship Σ _{t = T−1} ^M bitcount (t)> Maxbit ≧ Σt _{= T} ^M bitcount (t), where M is the maximum value of the index k (i). On the other hand, the smaller number of bits of k (i) -T + 1 and L are allocated. Among samples having an index equal to or greater than the value T-1 and not yet assigned L bits, Maxbit-Σt _{= T} ^M bitcount (t) samples from the one with the smaller sample number before rearrangement. On the other hand, one bit is allocated.
According to the bit allocating device according to another aspect of the present invention, the maximum number of bits that can be allocated to one sample is L, and Maxbit bits are allocated to N samples. The N samples are rearranged in the order of importance based on the index indicating the degree of importance. The sample numbers of the rearranged N samples are newly set to 0, 1,..., I,..., N−1, and the index indicating the importance given to the sample with the sample number i is k (i). The number of possible values of index k (i) + the initial value of L−1 bit counters is 0, and each bit counter bitcount corresponding to index k (i),..., K (i) − (L−1) The process of adding 1 to (k (i)),..., Bitcount (k (i) − (L−1)) is performed for each of i = 0, 1,. The maximum value of the index k (i) is M, as Maxbit the initial value of allocatable bits Bitsrv, the relationship of ^{_{Σ t = T-1 M bitcount}} (t)> Maxbit ≧ Σ t = T M bitcount (t) For each sample i having an index equal to or greater than the value T to be satisfied, the smaller number of bits of k (i) −T + 1 and L is assigned, and the assigned number of bits is subtracted from the assignable number of bits bitsrv. In the order from the smallest sample number before rearrangement, when the number of bits that can be allocated bitsrv is larger than 0 and has an index equal to or greater than the value T-1, L bits are not yet allocated. A process of adding 1 to the number of allocated bits for each sample and subtracting 1 from the number of bits that can be allocated bitsrv is performed.

Claims (6)

In a bit allocation device that allocates Maxbit bits to N samples, where L is the maximum number of bits that can be allocated to one sample,
The sample numbers of the N samples are 0, 1,..., I,..., N−1, the index representing the importance given to the sample of the sample number i is k (i), and the index k (i). The initial value of the number of possible values + L−1 bit counter bitcount is 0, and each bit counter bitcount (k (i) corresponding to the index k (i),..., K (i) − (L−1) )),..., Bitcount (k (i) − (L−1)), and a bit count unit for performing processing for adding i = 0, 1,.
For each sample i having an index equal to or greater than the value T defined by the following equation (1), where M is the maximum value of the index k (i), the smaller number of k (i) −T + 1 and L A first allocator to allocate bits;

Second allocation in which one bit is allocated to each of the number of samples defined by the following formula (2) among samples having an index equal to or greater than the value T-1 and not yet allocated L bits. And

A bit allocation device including: In a bit allocation device that allocates Maxbit bits to N samples, where L is the maximum number of bits that can be allocated to one sample,
A sorting unit for rearranging the N samples in an important order based on an index representing the importance given to the N samples;
The sample numbers of the rearranged N samples are newly set to 0, 1,..., I,..., N−1, and the index indicating the importance given to the sample with the sample number i is k (i). , The number of possible values of the index k (i) + the initial value of the L−1 bit counter bitcount is 0, and each bit corresponding to the index k (i),..., K (i) − (L−1) , Bit count (k (i)),..., Bitcount (k (i) − (L−1)) is incremented by 1 for each of i = 0, 1,. A counting section;
For each sample i having an index equal to or greater than the value T defined by the following equation (1), where M is the maximum value of the index k (i), the smaller number of k (i) −T + 1 and L A first allocator to allocate bits;

Among the samples having an index equal to or greater than the value T-1 and not yet assigned L bits, each sample of the number defined by the following formula (2) is assigned to the sample having the smallest sample number before the rearrangement. On the other hand, a second allocation unit that allocates one bit;

A bit allocation device including: In the bit allocation method of allocating Maxbit bits to N samples, where L is the maximum number of bits that can be allocated to one sample,
The bit count unit sets the sample numbers of the N samples to 0, 1,..., I,..., N−1, and sets the index indicating the importance given to the sample of the sample number i as k (i). The number of possible values of the index k (i) + the initial value of the L−1 bit counter bitcount is 0, and each bit counter corresponding to the index k (i),..., K (i) − (L−1) Bit count (k (i)),..., bit count (k (i) − (L−1)) is incremented by 1 for each of i = 0, 1,. Steps,
The first allocator sets k (i) −T + 1 and L for each sample i having an index greater than or equal to the value T defined by the following equation (1), where M is the maximum value of the index k (i). A first allocation step of allocating the smaller number of bits,

The second assigning unit has one index for each sample defined by the following equation (2) among samples having an index equal to or greater than the value T-1 and not yet assigned L bits. A second assignment step to assign bits;

Bit allocation method including In the bit allocation method of allocating Maxbit bits to N samples, where L is the maximum number of bits that can be allocated to one sample,
A sorting step in which the sorting unit sorts the N samples in an important order based on an index representing the importance given to the N samples;
The bit count unit newly sets the sample numbers of the rearranged N samples as 0, 1,..., I,..., N−1, and indicates an index indicating the importance given to the sample of the sample number i. k (i), the number of possible values of the index k (i) + the initial value of the L−1 bit counter bitcount is 0, and the index k (i),..., k (i) − (L−1) , Bitcount (k (i) − (L−1)) corresponding to each bit counter corresponding to 1 is added to each of i = 0, 1,..., N−1. A bit count step for
The first allocator sets k (i) −T + 1 and L for each sample i having an index greater than or equal to the value T defined by the following equation (1), where M is the maximum value of the index k (i). A first allocation step of allocating the smaller number of bits,

The second allocating unit is defined by the following formula (2) from the sample having a smaller sample number before the reordering among samples having the index equal to or greater than the value T-1 and not yet allocated L bits. A second allocating step of allocating one bit for each number of samples;

Bit allocation method including   The program for making a computer perform each step of the bit allocation method described in Claim 3 or 4.   A computer-readable recording medium on which a program for causing a computer to execute each step of the bit allocation method according to claim 3 is recorded.

Images