HK40019652B - Method and apparatus for decoding a bitstream including encoded hoa representations, and medium - Google Patents

Description

This includes methods and apparatus for decoding bitstreams represented by HOA codes, as well as media.

This application is a divisional application of the invention patent application with application number 201480072725.X, application date December 19, 2014, entitled "Method and apparatus for encoding side information required to improve the encoding of high-order high-fidelity stereo sound reproduction representation of sound field".

Technical Field

The present invention relates to a method and apparatus for encoding side information required to improve the encoding of higher-order ambisonics representations of sound fields.

Background Technology

In addition to other techniques such as wave field synthesis (WFS) or channel-based methods such as 22.2 multichannel audio formats, high-order high-fidelity stereo reproduction (HOA) also offers the possibility of representing three-dimensional sound. Compared to channel-based methods, HOA representation offers the advantage of being independent of specific speaker setups. However, this flexibility comes at the cost of the decoding processing required for playback of the HOA representation on a specific speaker setup. HOA signals can also be presented to setups with only a few speakers, unlike the WFS method, which typically requires a very large number of speakers. Another advantage of HOA is that the same representation can be used without any modification to the binaural presentation of headphones.

HOA is a representation of the spatial density of complex plane harmonic amplitudes based on the expansion of truncated spherical harmonics (SH). Each expansion coefficient is a function of angular frequency, which can be equivalently represented by a time-domain function. Therefore, without loss of generality, the entire HOA sound field representation can actually be assumed to contain O time-domain functions, where O denotes the number of expansion coefficients. Hereinafter, these time-domain functions will be equivalently referred to as the HOA coefficient sequence or HOA channel.

As the highest order N of the expansion increases, the spatial resolution of the HOA representation improves. Unfortunately, the number of expansion coefficients O grows quadratically with order N, specifically O = (N+1) ^² . For example, a typical HOA representation with order N = 4 requires O = 25 HOA (expansion) coefficients. Based on the preceding considerations, given a desired single-channel sampling rate f _{_s} and bit depth N_{_b} per sample, the total bit rate for transmitting the HOA representation is determined by O·f _{_s} ·N_{_b} . Therefore, transmitting an order N = 4 HOA representation at a sampling rate of f _{_s} = 48 kHz using N _{_b} = 16 bits per sample results in a bit rate of 19.2 MBits/s, which is very high for many practical applications such as streaming. Therefore, compressing the HOA representation is highly desirable.

Compression of HOA sound field representations is proposed in WO 2013/171083A1, EP 13305558.2, and PCT/EP2013/075559. These processes share the common feature of performing sound field analysis and decomposing a given HOA representation into directional and residual environmental components. On the one hand, the final compressed representation is assumed to contain several quantized signals derived from the correlation coefficient sequence of the environmental HOA components and the perceptual encoding of the directional signals. On the other hand, it is assumed to contain additional side information associated with the quantized signals, which is necessary to reconstruct the HOA representation from its compressed version.

A key part of this side information is the description of multiple parts of the original HOA representation predicted from the directional signal. Since the original HOA representation is assumed to be equivalently represented by several spatially dispersed general plane waves impacting from spatially uniformly distributed directions for this prediction, the prediction is referred to below as the spatial prediction.

The encoding of such side information related to spatial prediction is described in ISO/IEC JTC1/SC29/WG11, N14061, “Working Draft Text of MPEG-H 3D Audio HOA RM0”, November 2013, Geneva, Switzerland. However, the existing encoding of this side information is quite inadequate.

Summary of the Invention

One problem this invention aims to solve is to provide a more efficient way to encode side information related to spatial prediction.

This problem is solved by the methods disclosed in this invention. Apparatus utilizing these methods is also disclosed in this invention.

A bit is pre-allocated to the encoded side information representation data ζ _COD , and this bit is used to indicate whether any prediction should be performed. This feature reduces the average bit rate of transmitting ζ _COD data over time. Furthermore, in certain situations, transmitting or relaying the number of predictions and indicators of the activity is more efficient than using a bit array indicating whether a prediction should be performed in each direction. A single bit can be used to indicate how the indicator presumed to be the direction of prediction execution is encoded. On average, this operation further reduces the bit rate of transmitting ζ _COD data over time.

In principle, the method of the present invention is suitable for improving the encoding of side information required for HOA representation of a sound field encoded using an input time frame of a high-order high-fidelity stereo reproduction (labeled HOA) coefficient sequence, wherein a dominant direction signal and residual environmental HOA components are determined, and a prediction is used for the dominant direction signal, thereby providing side information data describing the prediction to the encoded frame of the HOA coefficients, wherein the side information data may include:

- A bit array indicating whether to perform direction prediction;

- Each bit in this array indicates the type of prediction for the direction in which the prediction is to be performed;

- Its elements are data arrays of indicators of the directional signals to be used in relation to the prediction to be performed;

- Its elements represent a data array of quantized scaling factors.

The method includes the following steps:

- Provides an indication of whether the predicted bit value should be performed;

- If prediction is not performed, then the bit array and the data array are omitted from the edge information data;

- If the prediction is to be performed, then, instead of the bit array indicating whether a prediction is to be performed on the direction, a data array indicating the number of predictions to be performed and a data array containing an index of the direction to be predicted are included in the bit values of the side information data.

In principle, the apparatus of the present invention is suitable for improving the encoding of side information required for HOA representation of a sound field encoded using an input time frame of a high-order high-fidelity stereo reproduction (labeled HOA) coefficient sequence, wherein a dominant direction signal and residual environmental HOA components are determined, and a prediction is used for the dominant direction signal, thereby providing side information data describing the prediction to the encoded frame of the HOA coefficients, wherein the side information data may include:

- A bit array indicating whether to perform direction prediction;

- Each bit in this array indicates the type of prediction for the direction in which the prediction is to be performed;

- Its elements are data arrays of indicators of the directional signals to be used in relation to the prediction to be performed;

- Its elements represent a data array of quantized scaling factors.

The device includes the following components:

- Provides an indication of whether the predicted bit value should be performed;

- If prediction is not performed, then the bit array and the data array are omitted from the edge information data;

- If the prediction is to be performed, then, instead of the bit array indicating whether a prediction is to be performed on the direction, a data array indicating the number of predictions to be performed and a data array containing an index of the direction to be predicted are included in the bit values of the side information data.

Further advantageous embodiments of the invention are disclosed in the individual claims.

Attached Figure Description

Exemplary embodiments of the present invention are described with reference to the accompanying drawings, wherein,

Figure 1 shows an exemplary encoding of side information related to spatial prediction in the HOA compression process described in EP 13305558.2;

Figure 2 illustrates an exemplary decoding of side information related to spatial prediction in the HOA decompression process described in patent application EP 13305558.2;

Figure 3 illustrates the HOA decomposition described in patent application PCT/EP2013/075559;

Figure 4 shows a diagram representing the direction of the general plane wave (shown as a cross) and the direction of the dominant sound source (shown as a circle) representing the residual signal. These directions are represented as sampling positions on a unit sphere in a three-dimensional coordinate system;

Figure 5 illustrates the existing technology coding for spatial prediction edge information;

Figure 6 illustrates the encoding of the present invention for spatial prediction edge information;

Figure 7 illustrates the decoding of the encoded spatial prediction edge information of this invention;

Figure 8 is a continuation of Figure 7.

Detailed Implementation

The following is a recap of the HOA compression and decompression process described in patent application EP 13305558.2, in order to provide context for the encoding of the present invention using edge information related to spatial prediction.

HOA compression

Figure 1 illustrates how the encoding of side information related to spatial prediction can be embedded into the HOA compression process described in patent application EP13305558.2. For HOA representation compression, frame-like processing is employed for non-overlapping input frames C(k) of a HOA coefficient sequence of length L, where k denotes the frame index. The first step or stage 11/12 in Figure 1 is optional and includes concatenating the non-overlapping k-th and (k-1)-th frames of the HOA coefficient sequence C(k) into a long frame as follows:

This long frame overlaps with adjacent long frames by 50%, and is subsequently used to estimate the dominant sound source direction. Similar to the notation used in [previous method], the tilde is used in the following description to indicate that the quantities refer to long overlapping frames. If step/stage 11/12 does not exist, the tilde has no specific meaning. Bold parameters represent a set of values, such as a matrix or vector.

As described in EP 13305558.2, long frames are used successively in step or stage 13 to estimate the dominant sound source direction. This estimation provides a data set of indices of the detected relevant directional signals and a data set D of corresponding direction estimates of the directional signals, representing the maximum number of directional signals that must be set before the start of HOA compression and can be handled in subsequent known processing.

In step or phase 14, the current (long) frame of the HOA coefficient sequence is decomposed (as proposed in EP 13305156.5) into several directional signals X _DIR (k-2) belonging to the directions contained in the group and the residual environment HOA component C _AMB (k-2). To obtain a smooth signal, a delay of two frames is introduced as a result of the overlap-addition process. It is assumed that X _DIR (k-2) contains a total of D channels, however, only those corresponding to the active directional signals are non-zero. The indices of these channels are assumed to be output in the data group J _DIR,ACT (k-2). In addition, the decomposition in step/phase 14 provides some parameters ζ (k-2) that can be used on the decomposition side for predicting multiple parts of the original HOA representation from the directional signals (see EP 13305156.5 for more details). The HOA decomposition is described in more detail in the later section "HOA Decomposition" to explain the meaning of the spatial prediction parameter ζ (k-2).

In step or stage 15, the number of coefficients for the environmental HOA component _CAMB (k-2) is reduced to only O _RED + DN _DIR,ACT (k-2) non-zero HOA coefficient sequences, where DN _DIR,ACT (k-2) = |J _DIR,ACT (k-2)| represents the cardinality of the data set J _DIR,ACT (k-2), i.e., the number of active directional signals in frame k-2. Since the environmental HOA component is considered to always be represented by the minimum number of HOA coefficient sequences O _RED , the problem can actually be simplified to selecting the remaining DN _DIR,ACT (k-2) HOA coefficient sequences from the possible O _RED HOA coefficient sequences. To obtain a smooth, simplified representation of the environmental HOA, this selection is performed such that as few changes as possible occur compared to the selection made in the preceding frame k-3.

The final environmental HOA representation with a reduced number of O _RED + N _DIR,ACT (k-2) non-zero coefficient sequences is represented by C _AMB,RED (k-2). The index of the selected environmental HOA coefficient sequence is output in the data set J _AMB,ACT (k-2). In step/stage 16, as described in EP 13305558.2, the activity direction signal contained in X _DIR (k-2) and the HOA coefficient sequence contained in C _AMB,RED (k-2) are assigned to a single perceptually encoded l-channel frame Y (k-2). Perceptual coding step/stage 17 encodes the l-channels of frame Y (k-2) and outputs the encoded frame.

According to the present invention, after the decomposition of the original HOA representation in step/stage 14, in order to provide the encoded data representation ζ _COD (k-2), the spatial prediction parameters or side information data ζ (k-2) obtained from the decomposition of the HOA representation are losslessly encoded in step or stage 19 by using a set of indices that are delayed by two frames in delay 18.

HOA decomposition

Figure 2 exemplarily illustrates how the decoding of the received encoded side information data ζ _COD (k-2) related to spatial prediction is embedded in step or stage 25 into the HOA decomposition process described in Figure 3 of patent application EP 13305558.2. The decoding of the encoded side information data ζ _COD (k-2) is achieved by using a set of indicators whose reception is delayed by two frames in delay 24 before the decoded version ζ(k-2) of the encoded side information data ζ _COD (k-2) is incorporated into the composition of the HOA representation in step or stage 23.

In step or stage 21, in order to obtain l decoded signals in, perceptual decoding of l signals contained in is performed.

In the signal redistribution step or stage 22, the perceptual decoded signals in the frames of the directional signal and the environmental HOA component are redistributed to recreate the directional signal frames. The redistribution operation performed on the HOA compression is reproduced using the index data set and J _AMB,ACT (k-2), providing information on how the signals were redistributed. In the composition step or stage 23, the current frame of the desired total HOA representation is recomposed (according to the processing described in Figures 2b and 4 of PCT/EP2013/075559, using the frame of the directional signal active directional signal index set along with the corresponding directional set from the parameter ζ(k-2) of the predicted portion of the directional signal HOA representation, and the frame of the reduced environmental HOA component HOA coefficient sequence).

Corresponding to the components in PCT/EP2013/075559, and, and corresponding to the components in PCT/EP2013/075559, the activity direction signal index can be obtained by acquiring those indices of the row containing the effective elements. That is, the direction signal with respect to the uniformly distributed direction is predicted from the direction signal using the received parameter ζ(k-2) for such prediction, and then the current decompressed frame is reconstructed from the frame of the direction signal, from and from the predicted portion and the reduced environmental HOA components.

HOA decomposition

Regarding Figure 3, the HOA decomposition process is described in detail to explain the meaning of the spatial predictions therein. This process is derived from the process described in Figure 3 of patent application PCT/EP2013/075559.

First, in step or stage 31, smoothed dominant directional signals X _DIR (k-1) and their HOA representations C _DIR (k-1) are calculated using a group of long-frame directions represented by the input HOA and a group of corresponding indices of the directional signals. It is assumed that X _DIR (k-1) contains a total of D channels, but only those corresponding to the active directional signals are non-zero. The indices for these channels are assumed to be output in the group J _DIR,ACT (k-1). In step or stage 33, the residual between the original HOA representation and the HOA representation C _DIR (k-1) of the dominant directional signals is represented by the number of O directional signals (which can be considered as general plane waves from uniformly distributed directions called a uniform grid). In step or stage 34, these directional signals are predicted from the dominant directional signals X _DIR (k-1) to provide prediction signals with each prediction parameter ζ (k-1). For prediction, only the dominant directional signal x _DIR,d (k-1) with the index d included in the group is considered. The predictions will be described in more detail in the later section on "Spatial Prediction".

In step or stage 35, a smoothed HOA representation of the predicted direction signal is calculated. In step or stage 37, the residual C _AMB (k-2) between the original HOA representation, the HOA representation C _DIR (k-2) of the dominant direction signal, and the HOA representation of the predicted direction signal from the uniformly distributed direction is calculated and output.

The signal delay required in the processing shown in Figure 3 is executed by the corresponding delays of 381 to 387.

Spatial prediction

The purpose of spatial prediction is to predict O residual signals:

These O residual signals are predicted from the extended frames of the following smoothed directional signals:

(See the description of “HOA decomposition” in and above of patent application PCT/EP2013/075559).

Each residual signal q = 1, ..., O represents a spatially dispersed general plane wave impacted from direction Ω _q . Therefore, it is assumed that all directions Ω _q , q = 1, ..., O are almost uniformly distributed on a unit sphere. The entire set of all directions is called the "grid".

Assuming that the signal in direction d is active for each frame, then each direction signal d = 1, ..., D represents a general plane wave impacted by the trajectory interpolated between directions _ΩACT,d (k-3), ΩACT _{, d(k-2), ΩACT,d} (k-1) and _ΩACT,d (k).

To illustrate the meaning of spatial prediction with an example, consider the decomposition of the HOA representation of order N=3, where the maximum number of extracted directions is equal to D=4. For simplicity, further assume that only the direction signals with indices "1" and "4" are active, while those with indices "2" and "3" are inactive. Additionally, for simplicity, assume that the direction of the dominant sound source is constant for the frame under consideration, i.e.,

As a result of order N=3, there are 16 directions _Ωq for spatially dispersed general plane waves q=1, ..., O. Figure 4 shows these directions, as well as the directions ΩACT _,1 and _ΩACT,4 of the dominant active sound source.

Parameters used to describe existing technologies for spatial prediction

The aforementioned ISO/IEC document provides a method for describing spatial prediction. In this document, signals q = 1, ..., O are assumed to be predicted as a weighted sum of a predetermined maximum number of D _PREDs of the directional signals, or as a low-pass filtered version of that weighted sum. The side information related to spatial prediction is described by the parameter set ζ(k-1) = {p _TYPE (k-1), P _IND (k-1), P _{Q, F} (k-1)}, which contains the following three components:

• The vector p _TYPE (k-1) has elements p _TYPE,q (k-1), where q = 1, ..., 0. These elements indicate whether a prediction is performed for the q-th direction Ω _q , and if so, they also indicate the type of prediction. The meanings of these elements are as follows:

A matrix _PIND (k-1) contains elements _pIND,d,q (k-1), where d = 1, ..., D _PRED , and q = 1, ..., 0, indicating the direction signals used to predict the direction _Ωq . If no prediction is performed for the direction _Ωq , the corresponding column of matrix _PIND (k-1) is zero. Furthermore, if the prediction of the direction _Ωq uses less direction signals than D _PRED , the unwanted element in the q-th column of _PIND (k-1) is also zero.

• Matrix P _Q,F (k-1) contains the corresponding quantitative predictors p _Q,F,d,q (k-1), d＝1、…、D _PRED , q＝1、…、O.

In order to properly interpret these parameters, the following two parameters must be known on the decoding side:

• The maximum number of directional signals, D _PRED , which allows prediction of general plane wave signals.

• The number of bits used to quantize the predictor factors _pQ,F,d,q (k-1) is _BSC , d＝1、…、D _PRED , q＝1、…、O. The dequantization rule is given in equation (10).

These two parameters must be arbitrarily set to fixed values known to the encoder and decoder, or to fixed values to be transmitted separately, but the transmission rate is significantly less frequent than the frame rate. The latter option can be used to adapt these two parameters to the HOA representation to be compressed.

Assuming O = 16, D _PRED = 2, and B _SC = 8, an example of the parameter set might look something like the following:

p _TYpE (k-1)＝[1 0 0 0 0 0 2 0 0 0 0 0 0 0 0 0], (7)

This parameter means that, by pure multiplication with the factor obtained from dequantization of logarithm 40 (i.e., full band), the general plane wave signal from direction Ω ₁ is predicted from the directional signal from direction Ω _ACT,1. Furthermore, by low-pass filtering and multiplication with the factors obtained from dequantization of logarithms 15 and -13, the general plane wave signal from direction Ω ₇ is predicted.

Given this edge information, the prediction is assumed to be performed as follows:

First, the quantized predictors _pQ,F,d,q (k-1), d＝1、…、D _PRED , q＝1、…、O are dequantized to provide the actual predictors:

As already described, the B _SC marker is used to quantify a predetermined number of bits in the predictor. Additionally, if p _IND,d,q (k-1) equals zero, then p _F,d , _q (k-1) is assumed to be set to zero.

In the example above, assuming _BSC = 8, dequantizing the predictor vector would result in:

Furthermore, in order to perform low-pass prediction, a predetermined low-pass FIR filter h _LP of length L _h = 31 is used: = [h _LP (0)h _LP (1) … h _LP (L _h - 1)] (12). The filtering delay is given by D _h = 15 samples.

As a signal, assume the prediction signal

and direction signal

pass

and

*for: For signals composed of their samples, the sampled values of the predicted signal are given by the following formula:

*if: if

in,

As stated above, and now it can be seen from equation (17) that the signals q = 1, ..., O are assumed to be predicted by a weighted sum of a predetermined maximum number of _DPREDs of the directional signals or by a low-pass filtered version of that weighted sum.

Existing technology coding of side information related to spatial prediction

The aforementioned ISO/IEC document addresses the encoding of spatially predicted side information. This is summarized in Algorithm 1 shown in Figure 5 and will be explained below. For clarity, the frame index k-1 is ignored in all representations.

First, create a bit array ActivePred containing O bits, where bit ActivePred[q] indicates whether prediction is performed for direction Ω _q . The number of "1"s in this array is marked by NumActivePred.

Next, a bit array `PredType` of length `NumActivePred` is created, where each bit pair indicates the type of prediction (full-band or low-pass) for the direction to be predicted. Simultaneously, an unsigned integer array `PredDirSigIds` of length `NumActivePred·D <sub>PRED`</sub> is created, where each element of this array is labeled with the `D <sub>PRED`</sub> index of the direction signal to be used for each active prediction. If a direction signal with less than `D <sub>PRED`</sub> is used for the prediction, the index is assumed to be set to zero. Each element of the `PredDirSigIds` array is assumed to be represented by |log ₂ (D+1)| bits. The number of non-zero elements in the `PredDirSigIds` array is represented by `NumNonZerolds`.

Finally, an integer array QuantPredGains of length NumNonZerolds is created, whose elements are assumed to represent the quantization scaling factors _PQ,F,d,q (k-1) used in Equation (17). Dequantization is given in Equation (10) to obtain the corresponding dequantization scaling factors _PF,d,q (k-1). Each element of the array QuantPredGains is assumed to be represented by _BSC bits.

Finally, the encoded representation of the edge information ζ _COD contains four of the above arrays according to the following formula:

ζ _COD =[ActivePred PredType PredDirSiglds QuantPredGains].(19)

To illustrate this encoding with examples, we will use the encoding representations of equations (7) to (9):

ActivePred＝[1 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0] (20)

PredType = [0 1] (21)

PredDirSigIds＝[1 0 1 4] (22)

QuantPredGains＝[40 15 -13]. (23)

The number of bits required is 16 + 2 + 3·4 + 8·+3 = 54.

The encoding of side information related to spatial prediction in this invention

To improve the efficiency of encoding side information related to spatial prediction, the processing of existing techniques has been advantageously modified.

A) When encoding a typical sound field HOA representation, the inventors of this invention observed that there are often multiple frames in the HOA compression process where no spatial prediction is performed at all. However, in these frames, the bit array ActivePred contains only zeros, with the number of zeros equal to 0. Since this type of frame content occurs frequently, the processing of this invention pre-schedules a single bit PSPredictionActive for the encoded representation ζ _COD , indicating whether any prediction should be performed. If the value of the bit PSPredictionActive is zero (or, alternatively, "1"), then the array ActivePred and other prediction-related data are not included in the encoded side information ζ _COD . In effect, this operation reduces the average bit rate of ζ _COD transmission over time.

B) Further observations made when encoding the HOA representation of a typical sound field show that the number of active predictions, NumActivePred, is often very low. In this case, as an alternative to using the bit array ActivePred to indicate whether prediction should be performed in each direction _Ωq , transmitting or relaying the number of active predictions and their indices may be more efficient. In particular, this type of modified encoding of activity...

In the case of NumActivePred≤M _M (24), it is more efficient.

Here, _M is the largest integer that satisfies the following formula:

The value of _MM can be calculated solely from the knowledge of the HOA order N:O = (N+1) ^² . In Equation (25), | _log² ( _MM )| marks the number of bits required to encode the actual number NumActivePred of the active prediction, and _MM ·| _log² (O)| is the number of bits required to encode the indicators for each direction. The right side of Equation (25) corresponds to the number of bits in the array ActivePred, which is required to encode the same information in a known manner. Based on the above explanation, a single bit KindOfCodedPredIds can be used to indicate how the indicators of those directions presumed to be performing the prediction are encoded. If the bit KindOfCodedPredIds has a value of "1" (or alternatively, "0"), then the number NumActivePred and the array PredIds containing the indicators presumed to be performing the prediction are added to the encoded side information _ζCOD . Otherwise, if the bit KindOfCodedPredIds has a value of "0" (or, alternatively, "1"), then the array ActivePred is used to encode the same information.

On average, this operation reduces the bit transfer rate of ζ _COD over time.

C) To further improve the efficiency of side information coding, the fact that the actual number of available activity direction signals used for prediction is often less than D is utilized. This means that less than one bit is needed to encode each element of the index array PredDirSigIds. Specifically, the actual number of available activity direction signals used for prediction is given by the number of elements in the data group containing the index of the activity direction signal. Thus, one bit can be used to encode each element of the index array PredDirSigIds, making this type of coding more efficient. In the decoder, the data group is assumed to be known, therefore, the decoder also knows how many bits must be read for the index of the decoding direction signal. Note that the frame index for calculating ζ _COD and the index data group used must be the same.

The above modifications A) to C) to the known edge information encoding process result in the exemplary encoding process shown in Figure 6.

Therefore, the encoded side information contains the following components: ζ _COD = (26)

Note: In the aforementioned ISO/IEC document, for example in section 6.1.3, QuantPredGains is referred to as PredGains, but it includes quantized values.

The encoded representation of the examples in equations (7) to (9) will be:

PSPredictionActive＝1 (27)

KindOfCodedPredlds＝1 (28)

NumActivePred = 2 (29)

Predlds = [1 7] (30)

PredType = [0 1] (31)

PredDirSiglds＝[1 0 1 4] (32)

QuantPredGains＝[40 15 -13], (33)

The required number of bits is 1 + 1 + 2 + 2·4 + 2 + 2·4 + 8·3 = 46. Advantageously, compared with the prior art encoding representation in equations (20) to (23), the representation encoded according to the present invention requires 8 fewer bits. It is also possible not to provide the bit array PredType on the encoder side.

Decoding of side information encoding related to spatial prediction modifications

In the exemplary decoding process shown in Figures 7 and 8 (the process shown in Figure 8 is a continuation of the process in Figure 7), the decoding of the modified side information related to spatial prediction is summarized and explained below. First, all elements of the vector _pTYPE and the matrix _PIND with _PQ,F are initialized to zero. Then, the bit PSPredictionActive is read, which indicates whether spatial prediction should be performed. In the case of spatial prediction (i.e., PSPredictionActive = 1), the bit KindOfCodedPredIds is read, which indicates the type of encoding of the index of the direction to be predicted.

With KindOfCodedPredIds = 0, a bit array ActivePred of length O is read, where the q-th element indicates whether a prediction is performed for direction Ω _q . In the next step, the number of predictions NumActivePred is calculated from the ActivePred array, and a bit array PredType of length NumActivePred is read, where the elements indicate the type of prediction performed for each of the relevant directions. Using the information contained in ActivePred and PredType, the elements of vector p _TYPE are calculated.

Alternatively, the bit array PredType can be omitted on the encoder side, and the elements of the vector _pTYPE can be calculated from the bit array ActivePred.

With KindOfCodedPredIds = 0, the number of active predictions, NumActivePred, is read, which is assumed to be encoded with |log ₂ (M _M )| bits, where M _M is the largest integer satisfying equation (25). Then, the data array PredIds containing NumActivePred features is read, where each feature is assumed to be encoded with |log ₂ (O)| bits. The features in this array are indicators of the directions in which predictions must be performed. The bit array PredType of length NumActivePred is read sequentially, where each feature represents the type of prediction performed for each of the relevant directions. The features of vector pTYPE are calculated using the knowledge of NumActivePred, PredIds, and PredType. Alternatively, the features _{of vector pTYPE can} be calculated from the number NumActivePred and the data array PredIds without providing the bit array PredType on the encoder side.

For both cases (i.e., KindOfCodedPredIds = 0 and KindOfCodedPredIds = 1), in the next step, the array PredDirSigIds containing NumActivePred·D _PRED features is read. Each feature is assumed to be encoded using individual bits. Using the information contained in _pTYPE and PredDirSigIds, the features of matrix _PIND are defined, and the number of non-zero features in _PIND, NumNonZerolds, is calculated.

Finally, the array QuanPredGains, containing NumNonZerolds elements encoded with B _SC bits, is read. The elements of matrix PQ _,F are then defined using the information contained in P _IND and QuanPredGains.

The processing of the present invention can be implemented by a single processor or electronic circuit or by several processors or electronic circuits operating in parallel and/or operating on different portions of the processing of the present invention.

Claims (7)

1. A method for decoding a bitstream including an encoded high-order high-fidelity stereo echolalia (HOA) representation, the method comprising: Evaluate the value of the bit KindOfCodedPredIds; The first array ActivePred is evaluated based on the value of the bit KindOfCodedPredIds, wherein each element in the first array ActivePred indicates whether a prediction is performed for the corresponding direction; The elements of vector _pTYPE are determined based on the evaluation of the first array ActivePred; Evaluate the second array PredDirSigIds, wherein the feature labels of the second array PredDirSigIds are used as an index for the directional signal of the activity prediction; Based on the vector _pTYPE , the data set of the direction signal index, and the elements of the second array PredDirSigIds, the elements of the matrix _PIND that mark the index of the direction signal to perform the prediction of the direction are determined. 2. The method of claim 1, wherein each element of the second array PredDirSigIds is an index of the direction signal to be used for the prediction to be performed, and wherein each element is encoded based on individual bits and is decoded accordingly, wherein the number of elements of the data group that are labeled with the index of the direction signal is... 3. A decoder for decoding a bitstream including an encoded high-order high-fidelity stereo reproduction (HOA) representation, the decoder comprising a processor configured to: Evaluate the value of the bit KindOfCodedPredIds; The first array ActivePred is evaluated based on the value of the bit KindOfCodedPredIds, wherein each element in the first array ActivePred indicates whether a prediction is performed for the corresponding direction; The elements of vector _pTYPE are determined based on the evaluation of the first array ActivePred; Evaluate the second array PredDirSigIds, wherein the feature labels of the second array PredDirSigIds are used as an index for the directional signal of the activity prediction; Based on the vector _pTYPE , the data set of the direction signal index, and the elements of the second array PredDirSigIds, the elements of the matrix _PIND that mark the index of the direction signal to perform the prediction of the direction are determined. 4. The decoder of claim 3, wherein each element of the second array PredDirSigIds is an index of the direction signal to be used for the prediction to be performed, and wherein each element is encoded based on individual bits and is decoded accordingly, wherein the number of elements of the data group that are the indexes of the direction signal are marked. 5. A computer program product comprising instructions that, when executed on a computer, perform the method according to claim 1 or 2. 6. An apparatus for decoding a bitstream including an encoded high-order high-fidelity stereo reproduction (HOA) representation, comprising: processor, and A computer program product having instructions thereon that, when executed on the processor, cause a device to perform the method according to claim 1 or 2. 7. An apparatus for decoding a bitstream including an encoded high-order high-fidelity stereo reproduction (HOA) representation, the apparatus comprising a processor or electronic circuitry for performing the method according to claim 1 or 2.