Cross-reference section to related application
[0001] This application is a European divisional application of Euro-PCT patent application
EP 14815731.6 (reference: A16025EP01), filed 19 December 2014.
Technical field
[0002] The invention relates to a method and to an apparatus for improving the coding of side information required for coding a Higher Order Ambisonics representation of a sound field.
Background
[0003] Higher Order Ambisonics (HOA) offers one possibility to represent three-dimensional sound among other techniques like wave field synthesis (WFS) or channel based approaches like the 22.2 multichannel audio format. In contrast to channel based methods, the HOA representation offers the advantage of being independent of a specific loudspeaker set-up. This flexibility, however, is at the expense of a decoding process which is required for the playback of the HOA representation on a particular loudspeaker set-up. Compared to the WFS approach, where the number of required loudspeakers is usually very large, HOA signals may also be rendered to set-ups consisting of only few loudspeakers. A further advantage of HOA is that the same representation can also be employed without any modification for binaural rendering to head-phones.
[0004] HOA is based on the representation of the spatial density of complex harmonic plane wave amplitudes by a truncated Spherical Harmonics (SH) expansion. Each expansion coefficient is a function of angular frequency, which can be equivalently represented by a time domain function. Hence, without loss of generality, the complete HOA sound field representation actually can be assumed to consist of
O time domain functions, where
O denotes the number of expansion coefficients. These time domain functions will be equivalently referred to as HOA coefficient sequences or as HOA channels in the following.
The spatial resolution of the HOA representation improves with a growing maximum order
N of the expansion. Unfortunately, the number of expansion coefficients
O grows quadratically with the order
N, in particular
O = (
N + 1)
^{2}. For example, typical HOA representations using order
N = 4 require
O = 25 HOA (expansion) coefficients. According to the previously made considerations, the total bit rate for the transmission of HOA representation, given a desired single-channel sampling rate
f_{s} and the number of bits
N_{b} per sample, is determined by
O·
f_{s}·N_{b}. Consequently, transmitting an HOA representation of order
N = 4 with a sampling rate of
f_{s} = 48kHz employing
N_{b} = 16 bits per sample results in a bit rate of 19.2MBits/s, which is very high for many practical applications like e.g. streaming. Thus, compression of HOA representations is highly desirable.
The compression of HOA sound field representations is proposed in
WO 2013/171083 A1,
EP 13305558.2 and
PCT/EP2013/075559. These processings have in common that they perform a sound field analysis and decompose the given HOA representation into a directional component and a residual ambient component. On one hand the final compressed representation is assumed to consist of a number of quantised signals, resulting from the perceptual coding of the directional signals and relevant coefficient sequences of the ambient HOA component. On the other hand it is assumed to comprise additional side information related to the quantised signals, which side information is necessary for the reconstruction of the HOA representation from its compressed version.
An important part of that side information is a description of a prediction of portions of the original HOA representation from the directional signals. Since for this prediction the original HOA representation is assumed to be equivalently represented by a number of spatially dispersed general plane waves impinging from spatially uniformly distributed directions, the prediction is referred to as spatial prediction in the following.
The coding 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, this state-of-the-art coding of the side information is rather inefficient.
Summary of invention
[0005] A problem to be solved by the invention is to provide a more efficient way of coding side information related to that spatial prediction.
This problem is solved by the methods disclosed in claims 1 and 6. An apparatus that utilises these methods is disclosed in claims 2 and 7.
[0006] A bit is prepended to the coded side information representation data
ζ_{COD}, which bit signals whether or not any prediction is to be performed. This feature reduces over time the average bit rate for the transmission of the
ζ_{COD} data. Further, in specific situations, instead of using a bit array indicating for each direction if the prediction is performed or not, it is more efficient to transmit or transfer the number of active predictions and the respective indices. A single bit can be used for indicating in which way the indices of directions are coded for which a prediction is supposed to be performed. On average, this operation over time further reduces the bit rate for the transmission of the ζ
_{COD} data.
[0007] In principle, the inventive method is suited for improving the coding of side information required for coding a Higher Order Ambisonics representation of a sound field, denoted HOA, with input time frames of HOA coefficient sequences, wherein dominant directional signals as well as a residual ambient HOA component are determined and a prediction is used for said dominant directional signals, thereby providing, for a coded frame of HOA coefficients, side information data describing said prediction, and wherein said side information data can include:
- a bit array indicating whether or not for a direction a prediction is performed;
- a bit array in which each bit indicates, for the directions where a prediction is to be performed, the kind of the prediction;
- a data array whose elements denote, for the predictions to be performed, indices of the directional signals to be used;
- a data array whose elements represent quantised scaling factors,
said method including the step: - providing a bit value indicating whether or not said prediction is to be performed;
- if no prediction is to be performed, omitting said bit arrays and said data arrays in said side information data;
- if said prediction is to be performed, providing a bit value indicating whether or not, instead of said bit array indicating whether or not for a direction a prediction is performed, a number of active predictions and a data array containing the indices of directions where a prediction is to be performed are included in said side information data.
[0008] In principle the inventive apparatus is suited for improving the coding of side information required for coding a Higher Order Ambisonics representation of a sound field, denoted HOA, with input time frames of HOA coefficient sequences, wherein dominant directional signals as well as a residual ambient HOA component are determined and a prediction is used for said dominant directional signals, thereby providing, for a coded frame of HOA coefficients, side information data describing said prediction, and wherein said side information data can include:
- a bit array indicating whether or not for a direction a prediction is performed;
- a bit array in which each bit indicates, for the directions where a prediction is to be performed, the kind of the prediction;
- a data array whose elements denote, for the predictions to be performed, indices of the directional signals to be used;
- a data array whose elements represent quantised scaling factors,
said apparatus including means which: - provide a bit value indicating whether or not said prediction is to be performed;
- if no prediction is to be performed, omit said bit arrays and said data arrays in said side information data;
- if said prediction is to be performed, provide a bit value indicating whether or not, instead of said bit array indicating whether or not for a direction a prediction is performed, a number of active predictions and a data array containing the indices of directions where a prediction is to be performed are included in said side information data.
[0009] Advantageous additional embodiments of the invention are disclosed in the respective dependent claims.
Brief description of drawings
[0010] Exemplary embodiments of the invention are described with reference to the accompanying drawings, which show in:
- Fig. 1
- Exemplary coding of side information related to spatial prediction in the HOA compression processing described in EP 13305558.2;
- Fig. 2
- Exemplary decoding of side information related to spatial prediction in the HOA decompression processing described in patent application EP 13305558.2;
- Fig. 3
- HOA decomposition as described in patent application PCT/EP2013/075559;
- Fig. 4
- Illustration of directions (depicted as crosses) of general plane waves representing the residual signal and the directions (depicted as circles) of dominant sound sources. The directions are presented in a three-dimensional coordinate system as sampling positions on the unit sphere;
- Fig. 5
- State of art coding of spatial prediction side information;
- Fig. 6
- Inventive coding of spatial prediction side information;
- Fig. 7
- Inventive decoding of coded spatial prediction side information;
- Fig. 8
- Continuation of Fig. 7.
Description of embodiments
[0011] In the following, the HOA compression and decompression processing described in patent application
EP 13305558.2 is recapitulated in order to provide the context in which the inventive coding of side information related to spatial prediction is used.
HOA compression
[0012] In Fig. 1 it is illustrated how the coding of side information related to spatial prediction can be embedded into the HOA compression processing described patent application
EP 13305558.2. For the HOA representation compression, a frame-wise processing with non-overlapping input frames
C(
k) of HOA coefficient sequences of length
L is assumed, where
k denotes the frame index. The first step or stage 11/12 in Fig. 1 is optional and consists of concatenating the non-overlapping
k-th and (
k - 1)-th frames of HOA coefficient sequences
C(
k) into a long frame
C̃(
k) as
which long frame is 50% overlapped with an adjacent long frame and which long frame is successively used for the estimation of dominant sound source directions. Similar to the notation for
C̃(
k), the tilde symbol is used in the following description for indicating that the respective quantity refers to long overlapping frames. If step/stage 11/12 is not present, the tilde symbol has no specific meaning.
A parameter in bold means a set of values, e.g. a matrix or a vector.
The long frame
C̃(
k) is successively used in step or stage 13 for the estimation of dominant sound source directions as described in
EP 13305558.2. This estimation provides a data set
of indices of the related directional signals that have been detected, as well as a data set
of the corresponding direction estimates of the directional signals.
D denotes the maximum number of diretional signals that has to be set before starting the HOA compression and that can be handled in the known processing which follows.
In step or stage 14, the current (long) frame
C̃(
k) of HOA coefficient sequences is decomposed (as proposed in
EP 13305156.5) into a number of directional signals
X_{DIR}(
k - 2) belonging to the directions contained in the set
, and a residual ambient HOA component
C_{AMB}(
k - 2). The delay of two frames is introduced as a result of overlap-add processing in order to obtain smooth signals. It is assumed that
X_{DIR}(
k - 2) is containing a total of
D channels, of which however only those corresponding to the active directional signals are non-zero. The indices specifying these channels are assumed to be output in the data set
Additionally, the decomposition in step/stage 14 provides some parameters
ζ(
k -2) which can be used at decompression side for predicting portions of the original HOA representation from the directional signals (see
EP 13305156.5 for more details). In order to explain the meaning of the spatial prediction parameters
ζ(
k - 2), the HOA decomposition is described in more detail in the below section
HOA decomposition.
In step or stage 15, the number of coefficients of the ambient HOA component
C_{AMB}(
k - 2) is reduced to contain only
O_{RED} +
D -
N_{DIR,ACT}(
k - 2) non-zero HOA coefficient sequences, where
indicates the cardinality of the data set
i.e. the number of active directional signals in frame
k - 2. Since the ambient HOA component is assumed to be always represented by a minimum number
O_{RED} of HOA coefficient sequences, this problem can be actually reduced to the selection of the remaining
D -
N_{DIR,ACT}(
k - 2) HOA coefficient sequences out of the possible
O -
O_{RED} ones. In order to obtain a smooth reduced ambient HOA representation, this choice is accomplished such that, compared to the choice taken at the previous frame
k - 3, as few changes as possible will occur.
The final ambient HOA representation with the reduced number of
O_{RED} +
N_{DIR,ACT}(
k - 2) non-zero coefficient sequences is denoted by
C_{AMB,RED}(
k - 2). The indices of the chosen ambient HOA coefficient sequences are output in the data set
In step/stage 16, the active directional signals contained in
X_{DIR}(
k - 2) and the HOA coefficient sequences contained in
C_{AMB,RED}(
k - 2) are assigned to the frame
Y(
k - 2) of
I channels for individual perceptual encoding as described in
EP 13305558.2. Perceptual coding step/stage 17 encodes the
I channels of frame
Y(
k - 2) and outputs an encoded frame
[0013] According to the invention, following the decomposition of the original HOA representation in step/stage 14, the spatial prediction parameters or side information data
ζ(
k - 2) resulting from the decomposition of the HOA representation are losslessly coded in step or stage 19 in order to provide a coded data representation
ζ_{COD}(
k - 2), using the index set
delayed by two frames in delay 18.
HOA decompression
[0014] In Fig. 2 it is exemplary shown how to embed in step or stage 25 the decoding of the received encoded side information data
ζ_{COD}(
k - 2) related to spatial prediction into the HOA decompression processing described in Fig. 3 of patent application
EP 13305558.2. The decoding of the encoded side information data
ζ_{COD}(
k - 2) is carried out before entering its decoded version
ζ(
k - 2) into the composition of the HOA representation in step or stage 23, using the received index set
delayed by two frames in delay 24.
[0015] In step or stage 21 a perceptual decoding of the
I signals contained in
is performed in order to obtain the
I decoded signals in
Ŷ(
k - 2).
In signal re-distributing step or stage 22, the perceptually decoded signals in
Ŷ(
k - 2) are re-distributed in order to recreate the frame
X̂_{DIR}(
k - 2) of directional signals and the frame
Ĉ_{AMB,RED}(
k - 2) of the ambient HOA component. The information about how to re-distribute the signals is obtained by reproducing the assigning operation performed for the HOA compression, using the index data sets
and
In composition step or stage 23, a current frame
Ĉ(
k - 3) of the desired total HOA representation is re-composed (according to the processing described in connection with
Fig. 2b and Fig. 4 of
PCT/EP2013/075559 using the frame
X̂_{DIR}(
k - 2) of the directional signals, the set
of the active directional signal indices together with the set
of the corresponding directions, the parameters
ζ(
k - 2) for predicting portions of the HOA representation from the directional signals, and the frame
Ĉ_{AMB,RED}(
k - 2) of HOA coefficient sequences of the reduced ambient HOA component.
Ĉ_{AMB,RED}(
k - 2) corresponds to component
D̂_{A}(
k - 2) in
PCT/EP2013/ 075559, and
and
correspond to
A_{Ω̂}(
k) in PCT/
EP2013/075559, wherein active directional signal indices can be obtained by taking those indices of rows of
A_{Ω̂}(
k) which contain valid elements. I.e., directional signals with respect to uniformly distributed directions are predicted from the directional signals
X̂_{DIR}(
k -
2) using the received parameters
ζ(
k - 2) for such prediction, and thereafter the current decompressed frame
Ĉ(
k - 3) is re-composed from the frame of directional signals
X̂_{DIR}(
k - 2), from
and
, and from the predicted portions and the reduced ambient HOA component
Ĉ_{AMB,RED}(
k - 2).
HOA decomposition
[0016] In connection with Fig. 3 the HOA decomposition processing is described in detail in order to explain the meaning of the spatial prediction therein. This processing is derived from the processing described in connection with Fig. 3 of patent application
PCT/EP2013/075559.
First, the smoothed dominant directional signals
X_{DIR}(
k - 1) and their HOA representation
C_{DIR}(
k - 1) are computed in step or stage 31, using the long frame
C̃(
k) of the input HOA representation, the set
of directions and the set
of corresponding indices of directional signals. It is assumed that
X_{DIR}(
k - 1) contains a total of
D channels, of which however only those corresponding to the active directional signals are non-zero. The indices specifying these channels are assumed to be output in the set
1).
In step or stage 33 the residual between the original HOA representation
C̃(
k - 1) and the HOA representation
C_{DIR}(
k - 1) of the dominant directional signals is represented by a number of
O directional signals
X̃_{RES}(
k - 1), which can be considered as being general plane waves from uniformly distributed directions, which are referred to a uniform grid.
In step or stage 34 these directional signals are predicted from the dominant directional signals
X_{DIR}(
k - 1) in order to provide the predicted signals
together with the respective prediction parameters
ζ(
k - 1). For the prediction only the dominant directional signals
x_{DIR,d}(
k - 1) with indices
d, which are contained in the set
are considered. The prediction is described in more detail in the below section
Spatial prediction.
[0017] In step or stage 35 the smoothed HOA representation
Ĉ_{RES}(
k - 2) of the predicted directional signals
is computed. In step or stage 37 the residual
C_{AMB}(
k - 2) between the original HOA representation
C̃(
k - 2) and the HOA representation
C_{DIR}(
k - 2) of the dominant directional signals together with the HOA representation
Ĉ_{RES}(
k - 2) of the predicted directional signals from uniformly distributed directions is computed and is output.
The required signal delays in the Fig. 3 processing are performed by corresponding delays 381 to 387.
Spatial prediction
[0018] The goal of the spatial prediction is to predict the
O residual signals
from the extended frame
of smoothed directional signals (see the description in above section
HOA decomposition and in patent application
PCT/EP2013/075559) .
Each residual signal
x̃_{RES,GRID,q}(
k - 1),
q = 1,...,
0, represents a spatially dispersed general plane wave impinging from the direction
Ω_{q}, whereby it is assumed that all the directions
Ω_{q},
q = 1,
...,0 are nearly uniformly distributed over the unit sphere. The total of all directions is referred to as a 'grid' .
Each directional signal
x̃_{DIR,d}(
k - 1),
d = 1,...,
D represents a general plane wave impinging from a trajectory interpolated between the directions
Ω_{ACT,d}(
k - 3),
Ω_{ACT,d}(
k - 2),
Ω_{ACT,d}(
k - 1) and
Ω_{ACT,d}(
k), assuming that the d-th directional signal is active for the respective frames.
To illustrate the meaning of the spatial prediction by means of an example, the decomposition of an HOA representation of order
N = 3 is considered, where the maximum number of directions to extract is equal to
D = 4. For simplicity it is further assumed that only the directional signals with indices '1' and '4' are active, while those with indices '2' and '3' are non-active. Additionally, for simplicity it is assumed that the directions of the dominant sound sources are constant for the considered frames, i.e.
Ω_{ACT,d}(
k - 3)
= As a consequence of order
N = 3, there are
O = 16 directions
Ω_{q} of spatially dispersed general plane waves
x̃_{RES,GRID,q}(
k - 1),
q = 1,...,
0. Fig. 4 shows these directions together with the directions
Ω_{ACT,1} and
Ω_{ACT,4} of the active dominant sound sources.
State-of-the-art parameters for describing the spatial prediction
[0019] One way of describing the spatial prediction is presented in the above-mentioned ISO/IEC document. In this document, the signals
x̃_{RES,GRID,q}(
k - 1),
q = 1,...,0 are assumed to be predicted by a weighted sum of a predefined maximum number
D_{PRED} of directional signals, or by a low pass filtered version of the 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 consists of the following three components:
- The vector p_{TYPE}(k - 1) whose elements p_{TYPE,q}(k - 1), q = 1,...,0 indicate whether or not for the q-th direction Ω_{q} a prediction is performed, and if so, then they also indicate which kind of prediction. The meaning of the elements is as follows:
- The matrix P_{IND}(k - 1), whose elements p_{IND,d,q}(k - 1), d = 1,...,D_{PRED}, q = 1,...,0 denote the indices from which directional signals the prediction for the direction Ω_{q} has to be performed. If no prediction is to be performed for a direction Ω_{q}, the corresponding column of the matrix P_{IND}(k - 1) consists of zeros. Further, if less than D_{PRED} directional signals are used for the prediction for a direction Ω_{q}, the non-required elements in the q-th column of P_{IND}(k - 1) are also zero.
- The matrix P_{Q,F}(k - 1), which contains the corresponding quantised prediction factors p_{Q,F,d,q}(k - 1), d=1,...,D_{PRED}, q = 1,...,0.
The following two parameters have to be known at decoding side for enabling the appropriate interpretation of these parameters:
- The maximum number D_{PRED} of directional signals, from which a general plane wave signal x̃_{RES,GRID,q}(k - 1) is allowed to be predicted.
- The number B_{SC} of bits used for quantising the prediction factors p_{Q,F,d,q}(k - 1), d = 1,...,D_{PRED}, q = 1,...,0. The de-quantisation rule is given in equation (10).
These two parameters have to either be set to fixed values known to the encoder and decoder, or to be additionally transmitted, but distinctly less frequently than the frame rate. The latter option may be used for adapting the two parameters to the HOA representation to be compressed.
An example for a parameter set may look like the following, assuming
O = 16,
D_{PRED} = 2 and
B_{SC}= 8:
Such parameters would mean that the general plane wave signal
x̃_{RES,GRID,1}(
k - 1) from direction
Ω_{1} is predicted from the directional signal
x̃_{DIR,1}(
k - 1) from direction
Ω_{ACT,1} by a pure multiplication (i.e. full band) with a factor that results from de-quantising the value 40. Further, the general plane wave signal
x̃_{RES,GRID,7}(
k - 1) from direction
Ω_{7} is predicted from the directional signals
x̃_{DIR,}(
k - 1) and
x̃_{DIR,4}(
k - 1) by a lowpass filtering and multiplication with factors that result from de-quantising the values 15 and -13.
Given this side information, the prediction is assumed to be performed as follows:
First, the quantised prediction factors
p_{Q,F,d,q}(
k - 1),
d = 1,...,
D_{PRED},
q = 1,...,
0 are dequantised to provide the actual prediction factors
As already mentioned,
B_{SC} denotes a predefined number of bits to be used for the quantisation of the prediction factors. Additionally,
p_{F,d,q}(
k - 1) is assumed to be set to zero, if
p_{IND,d,q}(
k - 1) is equal to zero.
[0020] For the previously mentioned example, assuming
B_{SC}=8, the de-quantised prediction factor vector would result in
Further, for performing a low pass prediction a predefined low pass FIR filter
of length
L_{h} = 31 is used. The filter delay is given by
D_{h} = 15 samples.
Assuming as signals the predicted signals
and the directional signals
to be composed of their samples by
and
the sample values of the predicted signals are given by
with
As already mentioned and as now can be seen from equation (17), the signals
x̃_{RES,GRID,q}(
k - 1),
q = 1,...,
0 are assumed to be predicted by a weighted sum of a predefined maximum number
D_{PRED} of directional signals, or by a low pass filtered versions of the weighted sum.
State-of-the-art coding of the side information related to spatial prediction
[0021] In the above-mentioned ISO/IEC document the coding of the spatial prediction side information is addressed. It is summarised in Algorithm 1 depicted in Fig. 5 and will be explained in the following. For a clearer presentation the frame index
k - 1 is neglected in all expressions.
First, a bit array
ActivePred consisting of
O bits is created, in which the bit ActivePred[
q] indicates whether or not for the direction
Ω_{q} a prediction is performed. The number of 'ones' in this array is denoted by NumActivePred. Next, the bit array
PredType of length NumActivePred is created where each bit indicates, for the directions where a prediction is to be performed, the kind of the prediction, i.e. full band or low pass. At the same time, the unsigned integer array
PredDirSigIds of length NumActivePred ·
D_{PRED} is created, whose elements denote for each active prediction the
D_{PRED} indices of the directional signals to be used. If less than
D_{PRED} directional signals are to be used for the prediction, the indices are assumed to be set to zero. Each element of the array
PredDirSigIds is assumed to be represented by ┌log
_{2}(
D + 1)┐ bits. The number of non-zero elements in the array
PredDirSigIds is denoted by NumNonZeroIds.
Finally, the integer array
QuantPredGains of length NumNonZeroIds is created, whose elements are assumed to represent the quantised scaling factors
P_{Q,F,d,q}(
k - 1) to be used in equation (17). The dequantisation to obtain the corresponding dequantised scaling factors
P_{F,d,q}(
k - 1) is given in equation (10). Each element of the array
QuantPredGains is assumed to be represented by
B_{SC} bits.
In the end, the coded representation of the side information
ζ_{COD} consists of the four aforementioned arrays according to
For explaining this coding by an example, the coded representation of equations (7) to (9) is used:
The number of required bits is equal to 16+2+3·4+8·3=54.
Inventive coding of the side information related to spatial prediction
[0022] In order to increase the efficiency of the coding of the side information related to spatial prediction, the state-of-the-art processing is advantageously modified.
- A) When coding HOA representations of typical sound scenes, the inventors have observed that there are often frames where in the HOA compression processing the decision is taken to not perform any spatial prediction at all. However, in such frames the bit array ActivePred consists of zeros only, the number of which is equal to O. Since such frame content occurs quite often, the inventive processing prepends to the coded representation ζ_{COD} a single bit PSPredictionActive, which indicates if any prediction is to be performed or not. If the value of the bit PSPredictionActive is zero (or '1' as an alternative), the array ActivePred and further data related to the prediction are not to be included into the coded side information ζ_{COD}. In practise, this operation reduces over time the average bit rate for the transmission of ζ_{COD}.
- B) A further observation made while coding HOA representations of typical sound scenes is that the number NumActivePred of active prediction is often very low. In such situation, instead of using the bit array ActivePred for indicating for each direction Ω_{q} whether or not the prediction is performed, it can be more efficient to transmit or transfer instead the number of active predictions and the respective indices. In particular, this modified kind of coding the activity is more efficient in case that NumActivePred ≤ M_{M}, (24)
where M_{M} is the greatest integer number that satisfies
The value of M_{M} can be computed only with the knowledge of the HOA order N: O = (N + 1)^{2} as mentioned above.
In equation (25), ┌log_{2}(M_{M})┐ denotes the number of bits required for coding the actual number NumActivePred of active predictions, and M_{m} · ┌log_{2}(O)┐ is the number of bits required for coding the respective direction indices. The right hand side of equation (25) corresponds to the number of bits of the array ActivePred, which would be required for coding the same information in the known way.
According to the aforementioned explanations, a single bit KindOfCodedPredIds can be used for indicating in which way the indices of those directions, where a prediction is supposed to be performed, are coded. If the bit KindOfCodedPredIds has the value '1' (or '0' in the alternative), the number NumActivePred and the array PredIds containing the indices of directions, where a prediction is supposed to be performed, are added to the coded side information ζ_{COD}. Otherwise, if the bit KindOfCodedPredIds has the value '0' (or '1' in the alternative), the array ActivePred is used to code the same information.
On average, this operation reduces over time the bit rate for the transmission of ζ_{COD}. - C) To further increase the side information coding efficiency, the fact is exploited that often the actually available number of active directional signals to be used for prediction is less than D. This means that for the coding of each element of the index array PredDirSigIds less than ┌log_{2}(D + 1)┐ bits are required. In particular, the actually available number of active directional signals to be used for prediction is given by the number D̃_{ACT} of elements of the data set
, which contains the indices ι̃_{ACT,1},...,ι̃_{ACT,D̃ACT} of the active directional signals. Hence, ┌log_{2}(|D̃_{ACT} + 1|)┐ bits can be used for coding each element of the index array PredDirSigIds, which kind of coding is more efficient. In the decoder the data set
is assumed to be known, and thus the decoder also knows how many bits have to be read for decoding an index of a directional signal. Note that the frame indices of ζ_{COD} to be computed and the used index data set
have to be identical.
The above modifications A) to C) for the known side information coding processing result in the example coding processing depicted in Fig. 6.
Consequently, the coded side information consists of the following components:
Remark: in the above-mentioned ISO/IEC document e.g. in section 6.1.3,
QuantPredGains is called
PredGains, which however contains quantised values.
The coded representation for the example in equations (7) to (9) would be:
and the required number of bits is 1+1+2+2·4+2+2-4+8·3=46. Advantageously, compared to the state of the art coded representation in equations (20) to (23), this representation coded according to the invention requires 8 bits less.
Decoding of the modified side information coding related to spatial prediction
[0023] The decoding of the modified side information related to spatial prediction is summarised in the example decoding processing depicted in Fig. 7 and Fig. 8 (the processing depicted in Fig. 8 is the continuation of the processing depicted in Fig. 7) and is explained in the following. Initially, all elements of vector
p_{TYPE} and matrices
P_{IND} and
P_{Q,F} are initialised by zero. Then the bit PSPredictionActive is read, which indicates if a spatial prediction is to be performed at all. In the case of a spatial prediction (i.e. PSPredictionActive = 1), the bit KindOfCodedPredIds is read, which indicates the kind of coding of the indices of directions for which a prediction is to be performed.
In the case that KindOfCodedPredIds = 0, the bit array
ActivePred of length
O is read, of which the
q-th element indicates if for the direction
Ω_{q} a prediction is performed or not. In a next step, from the array
ActivePred the number NumActivePred of predictions is computed and the bit array
PredType of length NumActivePred is read, of which the elements indicate the kind of prediction to be performed for each of the relevant directions. With the information contained in
ActivePred and
PredType, the elements of the vector
p_{TYPE} are computed.
In case KindOfCodedPredIds = 1, the number NumActivePred of active predictions is read, which is assumed to be coded with ┌log
_{2}(
M_{M})┐ bits, where
M_{M} is the greatest integer number satisfying equation (25) . Then, the data array
PredIds consisting of NumActivePred elements is read, where each element is assumed to be coded by ┌log
_{2}(
O)┐ bits. The elements of this array are the indices of directions, where a prediction has to be performed. Successively, the bit array
PredType of length NumActivePred is read, of which the elements indicate the kind of prediction to be performed for each one of the relevant directions. With the knowledge of NumActivePred, PredIds and PredType, the elements of the vector
p_{TYPE} are computed.
For both cases (i.e. KindOfCodedPredIds = 0 and KindOfCodedPredIds = 1), in the next step the array
PredDirSigIds is read, which consists of NumActivePred .
D_{PRED} elements. Each element is assumed to be coded by ┌log
_{2}(D̃
_{ACT})┐ bits. Using the information contained in
p_{TYPE},
and
PredDirSigIds, the elements of matrix
P_{IND} are set and the number NumNonZeroIds of non-zero elements in
P_{IND} is computed.
Finally, the array
QuantPredGains is read, which consists of NumNonZeroIds elements, each coded by
B_{SC} bits. Using the information contained in
P_{IND} and
QuantPredGains, the elements of the matrix
P_{Q,F} are set.
[0024] The inventive processing can be carried out by a single processor or electronic circuit, or by several processors or electronic circuits operating in parallel and/or operating on different parts of the inventive processing.
Various aspects of the present invention may be appreciated from the following enumerated example embodiments (EEEs):
- 1. Method for improving the coding of side information required for coding a Higher Order Ambisonics representation of a sound field, denoted HOA, with input time frames of HOA coefficient sequences, wherein dominant directional signals as well as a residual ambient HOA component are determined and a prediction is used for said dominant directional signals, thereby providing, for a coded frame of HOA coefficients, side information data (ζ(k - 2)) describing said prediction, and wherein said side information data (ζ(k - 2)) can include:
- a bit array (ActivePred) indicating whether or not for a direction a prediction is performed;
- a bit array (PredType) in which each bit indicates, for the directions where a prediction is to be performed, the kind of the prediction;
- a data array (PredDirSigIds) whose elements denote, for the predictions to be performed, indices of the directional signals to be used;
- a data array (QuantPredGains) whose elements represent quantised scaling factors,
said method including the step: - providing (19; 34, 384) a bit value (PSPredictionActive) indicating whether or not said prediction is to be performed;
- if no prediction is to be performed, omitting said bit arrays and said data arrays in said side information data (ζ(k - 2));
- if said prediction is to be performed, providing (19; 34, 384) a bit value (KindOfCodedPredIds) indicating whether or not, instead of said bit array (ActivePred) indicating whether or not for a direction a prediction is performed, a number (NumActivePred) of active predictions and a data array (PredIds) containing the indices of directions where a prediction is to be performed are included in said side information data (ζ(k - 2)).
- 2. Apparatus for improving the coding of side information required for coding a Higher Order Ambisonics representation of a sound field, denoted HOA, with input time frames of HOA coefficient sequences, wherein dominant directional signals as well as a residual ambient HOA component are determined and a prediction is used for said dominant directional signals, thereby providing, for a coded frame of HOA coefficients, side information data (ζ(k - 2)) describing said prediction, and wherein said side information data (ζ(k - 2)) can include:
- a bit array (ActivePred) indicating whether or not for a direction a prediction is performed;
- a bit array (PredType) in which each bit indicates, for the directions where a prediction is to be performed, the kind of the prediction;
- a data array (PredDirSigIds) whose elements denote, for the predictions to be performed, indices of the directional signals to be used;
- a data array (QuantPredGains) whose elements represent quantised scaling factors,
said apparatus including means (19; 34, 384) which: - provide a bit value (PSPredictionActive) indicating whether or not said prediction is to be performed;
- if no prediction is to be performed, omit said bit arrays and said data arrays in said side information data (ζ(k - 2)) ;
- if said prediction is to be performed, provide a bit value (KindOfCodedPredIds) indicating whether or not, instead of said bit array (ActivePred) indicating whether or not for a direction a prediction is performed, a number (NumActivePred) of active predictions and a data array (PredIds) containing the indices of directions where a prediction is to be performed are included in said side information data (ζ(k - 2)).
- 3. Method according to EEE 1, or apparatus according to EEE 2, wherein in said coding of said HOA representation an estimation (13) of dominant sound source directions is carried out and provides a data set
of indices of directional signals that have been detected. - 4. Method according to the method of EEE 3, or apparatus according to the apparatus of EEE 3, wherein D is a preset maximum number of directional signals that can be used in said coding of said HOA coefficient sequences, and wherein each element of said data array (PredDirSigIds) which denote, for the predictions to be performed, indices of the directional signals to be used, is coded using ┌log_{2}(|D̃_{ACT} + 1|)┐ bits instead of ┌log_{2}(|D + 1|)┐ bits, D̃_{ACT} being the number of elements of said data set
of indices of directional signals that have been detected. - 5. Method according to the method of one of EEEs 1, 3 or 4, or apparatus according to the apparatus of one of EEEs 2 to 4, wherein said bit value (KindOfCodedPredIds) indicating that a number NumActivePred of active predictions and an array (PredIds) containing the indices of directions where a prediction is to be performed are included in said side information data (ζ(k - 2)) is provided only in case NumActivePred ≤ M_{M}, where M_{M} is the greatest integer number that satisfies ┌log_{2}(M_{M})┐ + M_{M} · ┌log_{2}(O)┐ < O, O = (N + 1)^{2}, and wherein N is the order of said HOA representation.
- 6. Method for decoding side information data (ζ(k - 2)) which was coded according to the method of EEE 3, said method including the steps:
- evaluating (25) said bit value (PSPredictionActive) indicating whether or not said prediction is to be performed;
- if said prediction is to be performed, evaluating (25) said bit value (KindOfCodedPredIds) indicating whether
- a) said bit array (ActivePred) indicating whether or not for a direction a prediction is to be performed, or
- b) said number (NumActivePred) of active predictions and said array (PredIds) containing the indices of directions where a prediction is to be performed,
are used in the decoding of said side information data (ζ(k - 2)), wherein in case a) :
evaluating said bit array (ActivePred) indicating whether or not for a direction a prediction is to be performed wherein its elements indicate if for a corresponding direction a prediction is performed;
evaluating said bit array (PredType) which elements indicate the kind of prediction for each of the corresponding directions;
computing from said bit arrays (ActivePred, PredType) the elements of a vector (p_{TYPE}),
and wherein in case b):
evaluating said number (NumActivePred) of active predictions;
evaluating said data array (PredIds) containing the indices of directions where a prediction is to be performed;
evaluating said bit array (PredType) which elements indicate the kind of prediction for each of the corresponding directions,
computing from said number (NumActivePred), said data array (PredIds) and said bit array (PredType) the elements of a vector (p_{TYPE}),
and wherein in case a) as well as b):
- evaluating said data array (PredDirSigIds) whose elements denote, for the predictions to be performed, indices of the directional signals to be used;
- computing from said vector (p_{TYPE}), said data set
of indices of directional signals and said data array (PredDirSigIds) the elements of a matrix (P_{IND}) denoting indices from which directional signals the prediction for a direction is to be performed, and the number of non-zero elements in that matrix; - evaluating said data array (QuantPredGains) whose elements represent quantised scaling factors used in said prediction.
- 7. Apparatus for decoding side information data (ζ(k - 2)) which was coded according to the apparatus of EEE 3, said apparatus including a processor which performs:
- evaluating (25) said bit value (PSPredictionActive) indicating whether or not said prediction is to be performed;
- if said prediction is to be performed, evaluating (25) said bit value (KindOfCodedPredIds) indicating whether
- a) said bit array (ActivePred) indicating whether or not for a direction a prediction is to be performed, or
- b) said number (NumActivePred) of active predictions and said array (PredIds) containing the indices of directions where a prediction is to be performed,
are used in the decoding of said side information data (ζ(k - 2)), wherein in case a) :
evaluating said bit array (ActivePred) indicating whether or not for a direction a prediction is to be performed wherein its elements indicate if for a corresponding direction a prediction is performed;
evaluating said bit array (PredType) which elements indicate the kind of prediction for each of the corresponding directions;
computing from said bit arrays (ActivePred, PredType) the elements of a vector (p_{TYPE}),
and wherein in case b):
evaluating said number (NumActivePred) of active predictions;
evaluating said data array (PredIds) containing the indices of directions where a prediction is to be performed;
evaluating said bit array (PredType) which elements indicate the kind of prediction for each of the corresponding directions,
computing from said number (NumActivePred), said data array (PredIds) and said bit array (PredType) the elements of a vector (p_{TYPE}),
and wherein in case a) as well as b):
- evaluating said data array (PredDirSigIds) whose elements denote, for the predictions to be performed, indices of the directional signals to be used;
- computing from said vector (p_{TYPE}), said data set
of indices of directional signals and said data array (PredDirSigIds) the elements of a matrix (P_{IND}) denoting indices from which directional signals the prediction for a direction is to be performed, and the number of non-zero elements in that matrix; - evaluating said data array (QuantPredGains) whose elements represent quantised scaling factors used in said prediction.
- 8. Method according to EEE 6, or apparatus according to EEE 7, wherein each element of said data array (PredDirSigIds), which denotes for the predictions to be performed indices of the directional signals to be used and which was coded using ┌log_{2}(|D̃_{ACT} + 1|)┐ bits, is correspondingly decoded, D̃_{ACT} being the number of elements of said data set
of indices of directional signals. - 9. Digital audio signal that is coded according to the method of EEE 1.
- 10. Computer program product comprising instructions which, when carried out on a computer, perform the method according to EEE 1.
1. Method for the coding of side information required for coding a Higher Order Ambisonics, HOA, representation of a sound field, with input time frames of HOA coefficient sequences, wherein dominant directional signals as well as a residual ambient HOA component are determined and a prediction is used for said dominant directional signals, thereby providing, for a coded frame of HOA coefficients, side information data (
ζ(
k - 2)) describing said prediction, including a bit array (
ActivePred) indicating whether or not for a direction a prediction is performed;
said method comprising:
- providing a bit value (PSPredictionActive) indicating whether or not said prediction is to be performed, wherein the bit value (PSPredictionActive) is provided to indicate that no prediction is to be performed if the bit array (ActivePred) consists of zeros only;
- if no prediction is to be performed, omitting said side information data (ζ(k - 2)) describing said prediction, including omitting said bit array (ActivePred);
- if said prediction is to be performed, providing said side information data (ζ(k - 2)) describing said prediction.
2. Apparatus for the coding of side information required for coding a Higher Order Ambisonics, HOA, representation of a sound field, with input time frames of HOA coefficient sequences, wherein dominant directional signals as well as a residual ambient HOA component are determined and a prediction is used for said dominant directional signals, thereby providing, for a coded frame of HOA coefficients, side information data (
ζ(
k - 2)) describing said prediction, including a bit array (
ActivePred) indicating whether or not for a direction a prediction is performed;
said apparatus including means which:
- provide a bit value (PSPredictionActive) indicating whether or not said prediction is to be performed, wherein the bit value (PSPredictionActive) is provided to indicate that no prediction is to be performed if the bit array (ActivePred) consists of zeros only;
- if no prediction is to be performed, omit said side information data (ζ(k - 2)) describing said prediction, including omitting said bit array (ActivePred);
- if said prediction is to be performed, providing said side information data (ζ(k - 2)) describing said prediction.
3. The method according to claim 1, or the apparatus according to claim 2, wherein providing said side information data (ζ(k - 2)) describing said prediction comprises: providing a bit value (KindOfCodedPredIds) indicating whether or not, instead of said bit array (ActivePred) indicating whether or not for a direction a prediction is performed, a number (NumActivePred) of active predictions and a data array (PredIds) containing the indices of directions where a prediction is to be performed are included in said side information data (ζ(k - 2)).
4. Method according to claim 1 or 3, or apparatus according to claim 2 or 3, wherein in said coding of said HOA representation an estimation of dominant sound source directions is carried out and provides a data set
of indices of directional signals that have been detected.
5. Method for decoding side information data required for decoding an encoded Higher Order Ambisonics, HOA, representation of a sound field, the encoded HOA representation comprising dominant directional signals as well as a residual ambient HOA component, wherein the side information for a coded frame of HOA coefficients describes a prediction used for said dominant directional signals, wherein the side information can include a bit array (
ActivePred) indicating whether or not for a direction a prediction is performed,
said method comprising:
- evaluating a bit value (PSPredictionActive) indicating whether or not said prediction is to be performed;
- if said prediction is to be performed, decoding the side information describing said prediction, including decoding the bit array (ActivePred).
6. Apparatus for decoding side information data required for decoding an encoded Higher Order Ambisonics, HOA, representation of a sound field, the encoded HOA representation comprising dominant directional signals as well as a residual ambient HOA component, wherein the side information for a coded frame of HOA coefficients describes a prediction used for said dominant directional signals, wherein the side information can include a bit array (
ActivePred) indicating whether or not for a direction a prediction is performed,
said apparatus including a processor which performs:
- evaluating a bit value (PSPredictionActive) indicating whether or not said prediction is to be performed;
- if said prediction is to be performed, decoding the side information describing said prediction, including the bit array (ActivePred).
7. The method of claim 5, or the apparatus of claim 6, wherein decoding the side information describing said prediction comprises:
- evaluating a bit value (KindOfCodedPredIds) indicating whether
a) said bit array (ActivePred) indicating whether or not for a direction a prediction is to be performed, or
b) a number (NumActivePred) of active predictions and an array (PredIds) containing the indices of directions where a prediction is to be performed,
are used in the decoding of said side information data (
ζ(
k - 2)),
wherein in case a):
- evaluating said bit array (ActivePred) indicating whether or not for a direction a prediction is to be performed wherein its elements indicate if for a corresponding direction a prediction is performed;
- computing from said bit array (ActivePred) the elements of a vector (p_{TYPE}),
and wherein in case b):
- evaluating said number (NumActivePred) of active predictions;
- evaluating said data array (PredIds) containing the indices of directions where a prediction is to be performed;
- computing from said number (NumActivePred) and said data array (PredIds) the elements of a vector (p_{TYPE}).
8. Digital audio signal comprising side information that is coded according to the method of claim 1.
9. Computer program product comprising instructions which, when carried out on a computer, cause the computer to perform the method according to any of claims 1, 3, 4, 5 or 7.