RU2803142C1 - Audio upmixing device with possibility of operating in a mode with or without prediction - Google Patents

Abstract

FIELD: computer technology.

SUBSTANCE: effect is achieved by calculating the second representation of the downmix signal in the frequency domain based on its first representation in the frequency domain, and the second representation in the frequency domain includes the second spectral components representing the spectral composition of the signal expressed in the second subspace of the multidimensional space, which includes a part of the multidimensional space not included in the first subspace, wherein the second spectral components of the downmix signal are determined by applying a finite impulse response (FIR) filter to the first spectral components of the downmix signal; calculating the side signal (S) based on the first and second representations of the downmix signal in the frequency domain, the first representation of the residual signal in the frequency domain, and the complex prediction factor (α) encoded in the bitstream signal; and calculating a stereo signal based on the first representation of the frequency domain downmix signal and the side signal.

EFFECT: increase in audio processing efficiency.

20 cl, 19 dwg, 2 tbl

Description

Field of technical application

The invention disclosed herein generally relates to stereo coding and, more specifically, to stereo coding methods using complex frequency domain prediction.

BACKGROUND OF THE INVENTION

Joint encoding of the left (L) and right (R) channels of a stereo signal allows for more efficient encoding compared to independent L and R encoding. A common approach for joint stereo encoding is mid/side (M/S) encoding. Here the average (M) signal is formed by adding the L and R signals, for example, the M signal can have the form

M = (L+R)/2.

Also, by subtracting two channels L and R, a side signal (S) is formed, for example, the signal S can be of the form

S = (L-R)/2.

In the case of M/S encoding, M and S signals are encoded instead of L and R signals.

In the MPEG (Moving Picture Experts Group) AAC (advanced audio coding) standard (see ISO/IEC standard document 13818-7), L/R stereo coding and M/S stereo coding can be selected to vary with time or in a manner that varies with frequency. Thus, a stereo encoder may use L/R coding for some frequency bands of a stereo signal, while using M/S (varying with frequency) coding to encode other frequency bands of a stereo signal. In addition, the encoder can switch between L/R and M/S encoding over time (variation over time). In the MPEG AAC standard, stereo coding is performed in the frequency domain, more specifically in the MDCT (Modified Discrete Cosine Transform) domain. This allows either L/R or M/S encodings to be adaptively selected in a frequency-varying as well as time-varying manner.

Parametric stereo coding is a method of efficiently encoding a stereo audio signal as a mono signal plus a small amount of additional information for the stereo parameters. It forms part of the MPEG-4 Audio standard (see document ISO/IEC 14496-3). A mono signal can be encoded using any audio encoder. Stereo parameters can be embedded in the auxiliary part of the mono bitstream, and thus full forward and backward compatibility is achieved. In the decoder, the mono signal is first decoded, after which the stereo signal is reconstructed using stereo parameters. A decorrelator version of the decoded mono signal, which has zero cross-correlation with the mono signal, is generated by a decorrelator, such as an appropriate phase filter, which may include one or more delay lines. Essentially, a decorrelated signal has the same spectral and temporal energy distribution as a monaural signal. The mono signal, together with the decorrelated signal, is input to an upmixing process that is controlled by stereo parameters and which reconstructs the stereo signal. For more information, see the article "Low Complexity Parametric Stereo Coding in MPEG-4", H. Purnhagen, Proc. of the ^7th Int. Conference on Digital Audio Effects (DAFx'04), Naples, Italy, October 5-8, 2004, pages 163-168.

MPEG Surround (MPS; see ISO/IEC 23003-1 and "MPEG Surround - The ISO/MPEG Standard for Efficient and Compatible Multi-Channel Audio Coding", J. Herre et al., Audio Engineering Convention Paper 7084, 122 ^nd Convention, May 5-8, 2007) allows the principles of parametric stereo coding to be combined with residual coding, which replaces the decorrelated signal with a transmitted residual and thus improves the perceived quality. Residual coding can be performed by downmixing the multi-channel signal and optionally by extracting spatial labels. During the downmixing process, residual signals representing the error signal are calculated, then encoded and transmitted. In the decoder they can replace decorrelated signals. In a hybrid approach, they can replace decorrelated signals in some frequency bands, preferably in relatively low frequency bands.

In accordance with the current MPEG Unified Speech and Audio Coding (USAC) system, two examples of which are shown in FIG. 1A and Fig. 1B, the decoder includes a complex-valued quadrature mirror filter (QMF) bank located downstream of the base decoder. The QMF representation resulting from the output of the filter bank is complex - and thus oversampled by a factor of two - and can be organized as a downmix signal (or equivalently an average signal) M and a residual signal D, to which an upmix matrix is applied mixing with complex elements. The L and R signals (in the QMF domain) are obtained as:

,

where g is the real-valued gain and α is the complex-valued prediction coefficient. Preferably, α is chosen to minimize the energy of the residual signal D. The gain can be determined by normalization, i.e. so as to ensure that the power of the total signal is equal to the sum of the powers of the left and right signals. The real and imaginary parts of each of the L and R signals are mutually redundant—in principle, each can be calculated from the other—but they are useful in allowing subsequent application of a spectral band replication (SBR) decoder without causing audible artifacts due to aliasing. For similar reasons, the use of an oversampled signal representation may also be chosen to prevent artifacts associated with other time- or frequency-adaptive signal processing (not shown), such as, for example, upmixing a mono signal into a stereo signal. The last stage of processing in the decoder is inverse QMF filtering. It is noted that the QMF representation of a bandlimited signal allows for bandlimited difference and "residual padding" methods that can be integrated into decoders of this type.

The above encoding structure is well suited for low bit rates, typically below 80 Kbps, but is not optimal for higher bit rates in terms of computational complexity. More precisely, at higher bit rates the SBR tool is generally not used (since it will not improve encoding efficiency). Therefore, in a decoder without an SBR stage, only the presence of a complex-valued upmix matrix justifies the presence of a QMF filter bank, which is computationally expensive and introduces latency (with a frame length of 1024 samples, the analysis/synthesis QMF filter bank introduces a delay of 961 samples). This clearly indicates the need for a more efficient coding structure.

Brief description of the invention

An object of the present invention is to provide methods and apparatus for stereo encoding that are computationally efficient also in the high bit rate range.

The invention accomplishes the stated purpose by providing an encoder and a decoder, encoding and decoding methods, and, accordingly, computer program products for encoding and decoding, as defined by the independent claims. Dependent claims define embodiments of the invention.

In a first feature, the invention provides a decoder system for generating a stereo signal by complex prediction stereo coding, and the decoder system includes: upmixing adapted to generate a stereo signal based on first representations of a downmix (M) signal and a residual signal (M). D) in the frequency domain, wherein each of the first frequency domain representations includes first spectral components representing the spectral composition of the corresponding signal expressed in a first subspace of the multidimensional space, and the upmixing stage includes: - - a module for computing a second representation of the downmixing signal in a frequency domain based on its first frequency domain representation, wherein the second frequency domain representation includes second spectral components representing the spectral composition of a signal expressed in a second subspace of the multidimensional space, which includes a portion of the multidimensional space not included in the first subspace; - - a weighted adder for calculating a side signal (S) based on the first and second frequency domain representations of the downmix signal, the first frequency domain representation of the residual signal, and the complex prediction coefficient (α) encoded in the bitstream signal; and - a sum-difference stage for calculating a stereo signal based on a first frequency domain representation of the downmix signal and the side signal, wherein the upmix stage may also operate in a relay mode in which said downmix signal and residual signal are applied directly to the summation signal. - difference stage.

In a second aspect, the invention provides an encoder system for encoding a stereo signal by a bitstream signal by complex prediction stereo coding, which includes: an estimator for estimating a complex prediction coefficient; - an encoding stage operable to: (a) convert the stereo signal into frequency domain representations of the downmix signal and the residual signal in a relationship determined by the value of the complex prediction coefficient; and a multiplexer for receiving outputs of the encoding and estimator stages and encoding them using said bitstream signal.

The third and fourth aspects of the invention provide methods for encoding a stereo signal into a bit stream and decoding the bit stream into at least one stereo signal. The technical features of each method are similar to those for the encoder system and the decoder system, respectively. In fifth and sixth features, the invention also provides a computer program product containing instructions for executing each of the methods on a computer.

The invention takes advantage of the advantages of unified stereo coding in the MPEG USAC system. These benefits persist at higher bit rates, where SBR is typically not used, without the significant increase in computational complexity that would accompany a QMF-based approach. This is possible because the critically sampled MDCT transform, which is the basis of transform coding in the MPEG USAC system, in accordance with the invention can be used for stereo complex prediction coding, at least in cases where the bandwidths of the encoded audio channel signals The downmix and residual channel are the same, and the upmix process does not include decorrelation. This means that additional QMF conversion is no longer required. An exemplary implementation of complex predictive stereo coding in the QMF domain would actually significantly increase the number of operations per unit time compared to traditional L/R or M/S stereo. Therefore, the encoder according to the invention is competitive at the specified bit rates, providing high audio quality at moderate computational costs.

As those skilled in the art will appreciate, the fact that the upmixing stage can also operate in the relay mode allows the decoder to adaptively perform decoding in accordance with traditional forward or joint coding and complex prediction coding as determined at the encoder end. Therefore, in cases where the decoder cannot positively improve the quality level above that of traditional direct L/R stereo encoding, or joint M/S stereo encoding, it can at least guarantee that the same level will be maintained. Thus, the decoder according to this aspect of the invention can, from a functional point of view, be considered extended relative to the premises.

As an advantage over QMF predictively encoded stereo signals, perfect reconstruction of the signal is possible (not counting quantization errors, which can be made arbitrarily small).

Thus, the invention provides an encoder for stereo encoding a signal based on a complex prediction transform. Preferably, the device according to the invention is not limited to complex prediction stereo encoding, but can also operate in direct L/R stereo encoding mode or joint M/S stereo encoding mode as appropriate so that the most suitable encoding method can be selected for specific application or over a specific period of time.

As the basis for the complex prediction according to the invention, a resampled (eg complex) representation of the signal including said first and second spectral components is used, and therefore modules for calculating said resampled representation are located in the encoder system and in the decoder system according to the invention. The spectral components refer to the first and second subspaces of a multidimensional space, which may represent a plurality of time-dependent features over an interval of a given length (eg, a predetermined time frame length) sampled at a finite sampling rate. It is well known that functions in such a special multidimensional space can be approximated by a finite weighted sum of basis functions.

As will be understood by those skilled in the art, an encoder adapted to operate in conjunction with a decoder is equipped with equivalent modules to create a resampled representation on which predictive coding is based so as to enable reliable reproduction of the encoded signal. Said equivalent modules may be identical or similar modules having identical or similar transmission characteristics. In particular, the encoder and decoder modules, respectively, may be similar or dissimilar processing units executing corresponding computer programs that perform equivalent sets of mathematical operations.

In some embodiments of a decoder system or encoder system, the first spectral components have real values expressed in a first subspace and the second spectral components have imaginary values expressed in a second subspace. The first and second spectral components together form a complex spectral representation of the signal. The first subspace may be a linear span of a first set of basis functions, while the second subspace may be a span of a set of second basis functions, some of which are linearly independent of the first set of basis functions.

In one embodiment of the invention, the module for calculating the complex representation is a real-to-imaginary transformation, i.e. a module for calculating the imaginary parts of the spectrum of a discrete signal based on the real spectral representation of this signal. The transformation can be based on exact or approximate mathematical relationships, such as formulas from harmonic analysis or heuristic relationships.

In some embodiments of a decoder system or encoder system, the first spectral components may be obtained by transforming the discrete time domain signal from the time domain to the frequency domain, preferably through a Fourier transform such as a discrete cosine transform (DCT), a modified discrete cosine transform (MDCT) , Discrete Sine Transform (DST), Modified Discrete Sine Transform (MDST), Fast Fourier Transform (FFT), Prime Factor Fourier Algorithm, etc. In the first four cases, the second spectral components can then be obtained by DST, MDST, DCT and MDCT, respectively. As is well known, the linear hull of cosines that are periodic on the unit interval forms a subspace that does not completely fit within the linear hull of sines that are periodic on the same interval. Preferably, the first spectral components can be obtained by MDCT, and the second spectral components can be obtained by MDST.

In one embodiment of the invention, the decoder system includes at least one temporal noise limiter (TNS module, or TNS filter) that is located upstream of the upmix stage. Generally speaking, the use of TNS improves the perceived audio quality for signals with transient-type components, and this also applies to embodiments of the decoder system of the invention containing TNS as a feature. In traditional L/R or M/S stereo encoding, the TNS filter may be applied as the last stage of frequency domain processing just before downconversion. In the case of complex prediction stereo coding, however, it is often more advantageous to apply a TNS filter on the downmix signal and the residual signal, i.e. before the upmix matrix. In other words, TNS is applied to linear combinations of left and right channels, which has some advantages. First, it may turn out that in this situation TNS is only useful for, say, a downmix signal. The TNS filtering may then be suppressed or bypassed for the residual signal, and it is necessary to transmit the TNS filter coefficients only for the downmix signal, which may mean more economical use of the available bandwidth. Second, computing the resampled representation of the downmix signal (eg, the MDST data derived from the MDCT data so that a complex frequency domain representation can be generated) that is needed for complex prediction coding may require that it be possible to calculate time domain representation of a buck signal. In turn, this means that the downmix signal should preferably be available as a time series of MDCT spectra obtained in a uniform manner. If a TNS filter was applied in the decoder after an upmix matrix that converts the downmix/residual representation to a left/right representation, only the sequence of post-TNS residual MDCT spectra of the downmix signal will be available. This can make efficient calculation of the corresponding MDST spectra quite difficult, especially if the left and right channels used TNS filters with different characteristics.

It should be emphasized that the availability of a time sequence of MDCT spectra is not an absolute criterion for obtaining an MDST representation suitable to serve as the basis for complex prediction coding. In addition to experimental evidence, this fact can be explained by the fact that TNS is generally only applicable to higher frequencies, such as those above a few kilohertz, so the residual signal filtered by TNS approximately corresponds to the unfiltered residual signal for lower frequencies. Thus, the invention can be implemented as a decoder for complex prediction stereo coding, in which, as discussed below, the TNS filters have a different arrangement than in the upstream direction relative to the upmixing stage.

In one embodiment of the invention, the decoder system includes at least one additional TNS module located downstream of the upmixing stage. The selector position selects either the TNS module(s) upstream of the upmix stage or the TNS module(s) downstream of the upmix stage. In some circumstances, computing a complex frequency domain representation does not require that a time domain representation of the downmix signal can be computed. In addition, as discussed below, the decoder can selectively operate in direct or joint encoding mode without using complex prediction encoding, in which case it may be more appropriate to use the traditional arrangement of TNS modules, i.e. as one of the last stages of processing in the frequency domain.

In one embodiment of the invention, the decoder system is adapted to save processing resources and possibly energy by disabling the module for computing the second frequency domain representation of the downmix signal when the latter is not needed. It is assumed that the downmix signal is divided into successive time blocks, each of which is associated with a complex prediction coefficient value. This value may be determined by a decision made for each time block by the encoder in cooperation with the decoder. Moreover, in this embodiment, the module for calculating the second frequency domain representation of the downmix signal is adapted to turn itself off if, for a given time block, the absolute value of the imaginary part of the complex prediction coefficient is zero or does not exceed a predetermined tolerance. Disabling the module may imply that for a given time block, a second frequency domain representation of the downmix signal is not calculated. If shutdown does not occur, a second frequency domain representation (eg, a set of MDST coefficients) may be multiplied by zero, or by a number having substantially the same order of magnitude as the decoder's machine epsilon (rounding unit), or whatever another suitable threshold value.

In a further development of the previous embodiment of the invention, savings in processing resources are achieved at the sub-level of the time block into which the downmix signal is divided. For example, such a sublayer within a time block may be a frequency band, where the decoder determines the value of the complex prediction coefficient for each frequency band within the time block. Similarly, the module for generating the second frequency domain representation is adapted to suppress its operation for a frequency band within a time block where the complex prediction coefficient is zero or has an absolute value less than a tolerance.

In one embodiment of the invention, the first spectral components are transform coefficients located in one or more time blocks of transform coefficients, where each block is generated by applying a transform to a time segment of the time domain signal. In addition, a module for calculating a second representation of the downmix signal in the frequency domain is adapted for

• obtaining one or more first intermediate components from at least some of the first spectral components;

• generating a combination of said one or more first spectral components in accordance with at least a portion of the one or more impulse responses to produce one or more second intermediate components; And

• obtaining said one or more second spectral components from said one or more second intermediate components. This procedure calculates the second frequency domain representation directly from the first frequency domain representation, as described in more detail in US Pat. No. 6,980,933 B2, especially in sections 8-28 and, in particular, in equation 41. As will be understood by those skilled in the art, the calculation does not is performed across the time domain, as opposed to, for example, an inverse transform followed by another transform.

For an example implementation of complex prediction stereo coding according to the invention, it has been estimated that the computational complexity increases only slightly (significantly less than the increase caused by complex prediction stereo coding in the QMF domain) compared to traditional L/R or M/S stereo . One embodiment of this type of invention, involving accurate calculation of the second spectral components, introduces a delay that is typically only a few percent longer than the delay introduced by a QMF-based implementation (assuming a time block length of 1024 samples and comparing it to delay of the hybrid block of analyzing/synthesizing QMF filters, which is equal to 961 discrete values).

Accordingly, in at least some of the previous embodiments of the invention, the impulse responses are adapted to the transformation by which the first frequency domain representation can be obtained, and more precisely, adapted according to its frequency characteristics.

In some embodiments of the invention, a first frequency domain representation of the downmix signal is obtained by a transform that is applied in connection with one or more analysis window functions (or cutoff functions, e.g., rectangular window, sine window, Kaiser-Bessel window, etc.), one of the goals of which is temporal segmentation without introducing harmful amounts of noise or unwanted changes in the spectrum. It is possible that the specified window functions are partially overlapping. In such a case, preferably, the frequency characteristics of the transformation depend on the characteristics of said one or more analysis window functions.

Continuing with embodiments of the invention that feature the computation of a second frequency domain representation within a frequency domain, it is possible to reduce the computational load by using an approximate second frequency domain representation. This approximation can be achieved by waiving the requirement for complete information on which the calculation is based. According to the teachings of US Pat. No. 6,980,933 B2, for example, to accurately calculate a second frequency domain representation of a downmix signal in one block, first requires frequency domain data from three time blocks, namely: the block simultaneous with the output block, the previous block and subsequent block. For the purposes of complex prediction coding according to the present invention, suitable approximations can be obtained by omitting - or replacing by zero - data originating from a subsequent block (whereby the action of the module can become causal, i.e. it will not introduce delay) and/or from the previous block so that the calculation of the second frequency domain representation is based on data from only one or two time blocks. It should be noted that even though skipping the input data may imply a rescaling of the second frequency domain representation - in the sense that, for example, it will no longer represent equal power - it can still be used as a basis for complex prediction coding , since it, as stated above, is calculated at the encoder and decoder ends in an equivalent manner. In reality, a possible change in scale of this type will be compensated for by changing the value of the prediction coefficient accordingly.

Another approximate method of calculating the spectral component portion of a second frequency domain representation of a downmix signal may involve combining at least two components from the first frequency domain representation. The latter components may be adjacent in time and/or frequency. Alternatively, they can be combined via finite impulse response (FIR) filtering with a relatively small number of links. For example, in a system using a time block size of 1024, these filters could include 2, 3, 4, etc. links Descriptions of methods for approximate calculations of this kind can be found, for example, in US patent application No. 2005/0197831 A1. If a window function is used, such as, for example, a non-rectangular function that gives relatively small weights to the surroundings of each of the time block boundaries, it may be appropriate to base the second spectral components in a time block only on combinations of the first spectral components in the same time block, implying that for For edge components, not the same amount of information is available. The approximation error possibly introduced by such a practical implementation is to some extent suppressed, or hidden, by the form of the window function.

In one embodiment of a decoder that is designed to output a stereo signal in the time domain, it includes the ability to switch between direct or joint stereo coding and complex prediction coding. This is achieved by supplying:

• a switch that can selectively act either as a relay stage (not modifying the signals) or as a sum-difference transform;

• an inverse conversion stage designed to perform the frequency-time conversion; And

• a selector circuit designed to supply to the inverse transform stage either a directly (or jointly) encoded signal or a signal encoded with complex prediction.

As those skilled in the art will appreciate, this flexibility of the decoder portion provides the encoder with the freedom to choose between traditional direct or joint coding and complex prediction coding. Therefore, in cases where the quality level of traditional L/R stereo encoding or joint M/S stereo encoding cannot be improved, this embodiment can at least ensure that the same level is maintained. Thus, the decoder according to the invention can be considered an extension of the current state of the art.

Another group of embodiments of the decoder system performs calculation of second spectral components in a second frequency domain representation via time domain. More precisely, a transformation is applied that is the inverse of the one by which the first spectral components are (or can be) obtained, followed by another transformation that produces the second spectral components as an output. In particular, the inverse MDCT may be followed by MDST. In order to reduce the number of conversions and deconversions, the inverse MDCT output signal in such an embodiment of the invention may be provided to both the MDST and the decoding system data output terminals (which may be preceded by additional processing steps).

For an example implementation of complex prediction coding according to the invention, it has been estimated that the computational complexity increases only slightly compared to traditional L/R or M/S stereo coding (even less than the increase caused by complex prediction coding in the QMF domain ).

As a further development of the embodiment referred to in the preceding paragraph, the upmixing stage may include an additional down-conversion stage for processing the side signal. In this case, the sum-difference stage is provided with a time domain representation of the side signal, which is generated by said additional inverse conversion stage, and a time domain representation of the downmix signal, which is generated by the already mentioned inverse conversion stage. It should be noted again that it is advantageous from the point of view of computational complexity to supply the last signal both to the sum-difference stage and to the specified other conversion stage, which was mentioned above.

In one embodiment of the invention, a decoder designed to output a stereo time domain signal includes the ability to switch between L/R stereo coding, or M/S joint stereo coding, and complex prediction coding. This is achieved by equipping:

• a switch capable of acting either as a relay stage or as a sum-difference stage;

• an additional inverse conversion stage for calculating the time domain representation of the spurious signal;

• a selector circuit for coupling the inverse conversion stages or a sum-difference stage coupled to a point upstream of the upmix stage and downstream of a switch (preferably where the switch is actuated to act as a bandpass filter, as may be the case for decoding a stereo signal generated by complex prediction encoding), or for combining a downmix signal from a switch with a side signal from a weighted adder (preferably where the switch is actuated to perform the function of a sum-difference stage, as may be in the case of decoding a directly encoded stereo signal).

As those skilled in the art will appreciate, this provides the encoder with the freedom to choose between traditional direct or joint encoding and complex prediction encoding, which implies the ability to ensure that the level of quality is at least equivalent to that of direct or joint stereo encoding

In one embodiment of the invention, the encoder system according to the second aspect of the invention may include an estimator for estimating a complex prediction coefficient to reduce, or minimize, signal power, or average signal power, for a residual signal. The minimization may occur over a period of time, preferably a time slice, or time block, or time frame of the signal that is being encoded. The square of the amplitude can be chosen as a criterion for the instantaneous signal power, and the integral of the squared amplitude (signal shape) over a period of time can be chosen as a criterion for the average signal power in this interval. Accordingly, the complex prediction coefficient is determined based on the time block and frequency band, i.e. its value is set such that it reduces the average power (ie total energy) of the residual signal in that time block and frequency band. In particular, the output signal, on which the complex prediction coefficient can be calculated in accordance with mathematical relationships known to those skilled in the art, can be created by modules designed to estimate parameters of parametric stereo coding, such as IID, ICC and IPD or similar ones.

In one embodiment of the invention, an encoding stage in the encoder system may also operate to function as a relay stage so as to enable direct stereo encoding. By selecting direct stereo encoding in situations where higher quality is expected, the encoder system can ensure that the encoded stereo signal will have at least the same quality as direct encoding. Likewise, in situations where the greater computational effort introduced by complex prediction coding is not motivated by significant improvements in quality, computational resource savings are thus readily available to the encoder system. The decision to choose between joint coding, direct coding, real prediction coding and complex prediction coding in the encoder is primarily based on rate/distortion optimization considerations.

In one embodiment of the invention, the encoder system may include a module configured to compute a second frequency domain representation directly (ie, without applying an inverse transform to the time domain and without using time domain signal data) based on the first spectral components. With respect to the corresponding embodiments of the decoder system described above, this module may have a similar design, namely, include similar processing operations, but in a different order so that the encoder is adapted to output data suitable as input at the decoder side. For the purpose of illustrating this embodiment of the invention, it is assumed that the stereo signal that is encoded includes the middle and side channels, or has been converted into such a design, and the encoding stage is adapted to receive the first frequency domain representation. The encoding stage includes a module for calculating a second frequency domain representation of the average signal. (The first and second frequency domain representations referred to herein are the same as those defined above; in particular, the first frequency domain representations may be an MDCT representation and the second frequency domain representation may be an MDST representation.) The encoding stage also includes a weighted adder, designed to calculate the residual signal as a linear combination formed from the spurious signal and two frequency domain representations of the average signal, weighted, respectively, by the real and imaginary parts of the complex prediction coefficient. The average signal, or a suitable first frequency domain representation thereof, can be used directly as a downmix signal. Moreover, in this embodiment, the estimator determines the value of the complex prediction coefficient to minimize the power, or average power, of the residual signal. The last operation (optimization) can be performed either through feedback control, where the estimator can receive a residual signal obtained through the current values of the prediction coefficient and intended for further adjustment if necessary, or, in a feedforward manner - through calculations performed directly on the left /right channels of the original stereo signal or on the middle/side channels. Preferred is a feedforward method in which the complex prediction coefficient is determined directly (in particular, not iteratively or without feedback) based on the first and second representations of the average frequency domain signal and the first representation of the side signal in the frequency domain. It should be noted that the determination of the complex prediction coefficient may be followed by a decision as to whether direct, joint, real or complex prediction coding is used, where the resulting quality is taken into account (preferably the perceived quality taking into account, for example, signal/mask ) for each of the available options; therefore, the above statements should not be interpreted as saying that there is no feedback mechanism within the encoder.

In one embodiment of the invention, the encoder system includes modules for calculating a second representation of the average (or downmix) of the frequency domain signal through the time domain. It should be understood that the implementation details related to this embodiment of the invention, at least to the extent that the calculation of the second frequency domain representation is considered, are similar or can be handled similarly to corresponding decoder embodiments. In this embodiment of the invention, the encoding stage includes:

• a sum-difference stage designed to convert the stereo signal into a form that includes the middle and side channels;

• a transform stage for creating a frequency domain representation of the side channel and a complex valued (and therefore oversampled) frequency domain representation of the middle channel; And

• a weighted adder designed to calculate the residual signal, where the complex prediction coefficient is used as a weighting coefficient. Here, the estimator may receive the residual signal and determine, perhaps in a feedback control manner, a complex prediction coefficient so as to reduce, or minimize, the power, or average power, of the residual signal. Preferably, however, the estimator receives the stereo signal that is being encoded and determines a prediction coefficient based on it. From a computational savings point of view, it is advantageous to use a critically sampled frequency domain representation of the side channel, since the latter will not be subject to complex number multiplication in this embodiment. Accordingly, the conversion stage may include an MDCT stage and a parallel MDST stage, where both stages take as input a time domain representation of the middle channel. Thus, an oversampled frequency domain representation of the middle channel and a critically sampled frequency domain representation of the side channel are generated.

It should be noted that the methods and apparatus disclosed in this section, after appropriate modifications within the capabilities of specialists, including typical experiments, can be used to encode signals containing more than two channels. Modifications intended to provide suitability for the specified multi-channel operation may follow, for example, the directions described in sections 4 and 5 of the above-cited article by J. Herre et al.

Characteristic features from two or more embodiments of the invention described above may be combined in further embodiments of the invention unless they are expressly additional. The fact that two characteristic features are set out in different claims does not prevent them from being advantageously combined. Likewise, further embodiments of the invention may also be provided omitting certain features that are not necessary or not essential for the intended purpose. For example, the decoding system of the invention may be implemented without a dequantization stage in cases where the encoded signal being processed is not quantized or is already available in a form suitable for processing by the upmixing stage.

Brief description of graphic materials

The invention will be further illustrated below by means of the embodiments described in the following section with reference to the accompanying drawings, in which:

Figures 1A and 1B are general block diagrams showing QMF-based decoders according to the current state of the art;

Figure 2 is a generalized block diagram of a complex prediction MDCT stereo decoder system in accordance with one embodiment of the present invention, wherein a complex channel representation of a signal being decoded is calculated in the frequency domain;

Figure 3 is a generalized block diagram of a MDCT-based complex prediction stereo decoder system in accordance with one embodiment of the present invention, wherein a complex channel representation of a signal being decoded is computed in the time domain;

Figure 4 shows an alternative embodiment of the decoder system of Figure 2, where the position of the TNS active stage can be selected;

Figure 5 includes generalized block diagrams showing MDCT-based complex prediction stereo encoder systems in accordance with embodiments of another feature of the present invention;

Figure 6 is a general block diagram of a complex predictive MDCT stereo encoder system in accordance with one embodiment of the present invention, wherein a complex channel representation of a signal being encoded is computed based on its time domain representation;

Figure 7 shows an alternative embodiment of the encoder system of Figure 6, which can also operate in direct L/R encoding mode;

Figure 8 is a generalized block diagram of a complex predictive MDCT based stereo encoder system in accordance with one embodiment of the present invention, wherein a complex channel representation of a signal being decoded is computed based on its first frequency domain representation whose decoder system can operate also in direct L/R encoding mode;

Figure 9 shows an alternative embodiment of the encoder system of Figure 7, which also includes a TNS stage located downstream of the encoding stage;

Figure 10 shows alternative embodiments of the portion marked A in Figures 2-8;

Figure 11 shows an alternative embodiment of the encoder system of Figure 8, which also includes two frequency domain modification devices located, respectively, in the downstream and upstream directions of the encoding stage;

Figure 12 is a graphical representation of test listening results at 96 Kbps for six sites, showing various possibilities for trade-offs between complexity and quality when calculating, or approximating, the MDST spectrum, where data points marked "+" refer to the hidden reference, "x " refers to the 3.5 KHz band-limited anchor, "*" refers to traditional USAC stereo coding (M/S or L/R), "□" refers to unified stereo coding in the MDCT domain via imaginary-part-locked complex prediction prediction coefficient (i.e., with valid prediction not requiring MDST), "■" refers to unified stereo coding in the MDCT domain by complex prediction using the current MDCT frame to calculate the MDST approximation, " _° " refers to unified stereo coding in the domain MDCT by complex prediction using the current and previous MDCT frames to calculate the MDST approximation, and "●" refers to unified stereo coding in the MDCT domain by complex prediction using the current, previous and next MDCT frames to calculate the MDST;

Figure 13 presents the data of Figure 12, but as a differential estimate with respect to unified stereo coding in the MDCT domain through complex prediction using the current MDCT frame to calculate the MDST approximation;

Figures 14A, 14B and 14C are general block diagrams showing three embodiments of a decoder system in accordance with embodiments of the invention;

Figure 15 is a flowchart showing a decoding method in accordance with one embodiment of the invention; And

Figure 16 is a flowchart showing an encoding method in accordance with one embodiment of the invention.

Detailed Description of Embodiments of the Invention

I. Decoder systems

Figure 2, in general block diagram form, shows a decoding system for decoding a bitstream including at least one complex prediction coefficient value. and an MDCT representation of a stereo signal containing a downmix channel M and a residual channel D. Real and imaginary parts , prediction coefficients may be quantized and/or jointly encoded. It is preferred, however, that the real and imaginary parts are quantized independently and uniformly, typically with a step size of 0.1 (a dimensionless number). According to the MPEG standard, the resolution of the spectral band used for the complex prediction coefficient is not necessarily equal to the resolution of scale factor bands ( sfb ; i.e., a group of MDCT lines that uses the same quantization step size and quantization range). In particular, the resolution of the frequency band for the prediction coefficient can be psychoacoustically justified, such as the Bark scale. The demultiplexer 201 is adapted to extract the specified MDCT representations and prediction coefficient (part of the Control Information shown in the figure) from the bit stream that is supplied to it. In fact, more control information than just the complex prediction coefficient information may be encoded in the bit stream, such as instructions on whether the bit stream should be decoded in predictive or non-predictive mode, TNS information, etc. The TNS information may include TNS parameter values for use by the decoder system's TNS (synthesis) filters. If multiple TNS filters are expected to use the same TNS parameter sets, it is more economical to receive this information in the form of a bit indicating that the parameter sets are identical, rather than receiving the two parameter sets independently. Information may also be included regarding whether to apply TNS before or after the upmixing stage, as appropriate based on, for example, a psychoacoustic assessment of these two possibilities. In addition, the control information can then indicate separately limited bandwidths for the downmix signal and the residual signal. For each channel, frequency bands above the bandwidth limit will not be decoded, but will be set to zero. In some cases, the highest frequency bands have so little energy that they are already quantized to zero. Standard practice (cf. the max_sfb parameter in the MPEG standard) typically uses the same bandwidth limit for both the downmix signal and the residual signal. However, the residual signal, more than the downmix signal, contains its energy reserves localized in lower frequency bands. Therefore, by placing a dedicated bandwidth cap on the residual signal, it is possible to lower the data bit rate without significantly reducing quality. For example, this could be controlled by two independent max_sfb parameters encoded in the bitstream: one for the downmix signal, and one for the residual signal.

In this embodiment, the MDCT representation of the stereo signal is segmented into successive time frames (time blocks) comprising a fixed number of data points (eg, 1024 points), one of several fixed numbers of data points (eg, 128 or 1024 points), or a variable number points. As those skilled in the art know, MDCT is critically sampled. The output signal of the decoding system shown on the right side of the illustration is a time domain stereo signal containing left L and right R channels. The dequantization modules 202 are adapted to manipulate the bit stream entering the decoding system, or, where necessary, the two bit streams resulting from demultiplexing the original bit stream and corresponding to each of the downmix and residual channels. The dequantized channel signals are supplied to switching node 203, operating either in relay mode or in sum-difference mode, which relate to the corresponding transformation matrices

And

.

As will be further explained in the next paragraph, the decoder system includes a second switching node 205. Both switch nodes 203, 205, like most other switches and switch nodes in this embodiment and in the embodiments that will be described below, operate in a frequency selective manner. This makes it possible to decode a wide variety of decoding modes, for example, frequency-dependent L/R or M/S decoding, as is known in the current art. Therefore, the decoder according to the invention can be considered an extension of the current state of the art.

Assuming for now that switching node 203 is in relay mode, in this embodiment, the dequantized signals are passed through appropriate TNS filters 204. TNS filters 204 are not essential to the operation of the decoding system and can be replaced by relay elements. Thereafter, the signal is supplied to the second switching node 205, which performs the same function as the switching node 203 located in the upstream direction. Provided that the input signals are as described above and the second switching unit 205 is set to the relay mode, the output signal of the first is a downmix channel signal and a residual channel signal. The downmix signal, still represented by its time-sequential MDCT spectrum, is fed to a real-to-imaginary transform 206 adapted to calculate from it the MDST spectrum of the downmix signal. In this embodiment, one MDST frame is based on three MDCT frames: one previous frame, one current (or simultaneous) frame, and one subsequent frame. Symbolically indicated - that the input side of the real-to-imaginary transform 206 includes delay elements.

The MDST representation of the downmix signal obtained from the real-to-imaginary transform 206 is weighted by the imaginary part. prediction coefficient and added to the real-part weighted MDCT representation of the downmix signal prediction coefficient, and MDCT representation of the residual signal. The two additions and multiplications are performed by multipliers and adders 210, 211, together forming a (functionally) weighted adder, which are supplied with the value of the complex prediction coefficient α encoded in the bit stream initially received by the decoder system. The complex prediction coefficient may be determined once for each time frame. It may also be determined more frequently, such as once for each frequency band within a frame, where the frequency bands represent a psychoacoustically motivated separation. It may also be determined less frequently, as will be described below in connection with the coding systems of the invention. The real-to-imaginary transform 206 is synchronized with the weighted adder such that the current MDST frame of the downmix channel signal is combined with simultaneous MDCT frames for each of the signals, the downmix channel signal and the residual channel signal. The sum of these three signals is the side signal . In this expression, M includes both the MDCT and MDST representations of the downmix signal, namely: , Where is real-valued. Thus, a stereo signal is obtained containing a downmix channel and a side channel, from which sum-difference transform 207 reconstructs the left and right channels as follows: . These signals are represented in the MDCT region. The final step of the decoding system is to apply the inverse MDCT 209 to each of the channels, whereby a time domain representation of the left/right stereo signal is obtained.

A possible implementation of the real-to-imaginary conversion 206 is further described in Applicant's US Pat. No. 6,980,933 B2, as noted above. Using Formula 41 in this patent, the transform can be expressed as a finite-duration impulse response filter, for example for even points:

,

Where - -th MDST data point, is the MDCT data for each of the frames, and N is the length of the frame. Besides, are impulse responses that depend on the windowing function applied and are therefore determined for each choice of windowing function, such as rectangular, sinusoidal or Kaiser-Bessel derivative, and for each frame length. The complexity of this calculation can be reduced by omitting those impulse responses that have relatively less energy and contribute relatively less to the MDST data. As an alternative or extension of this implementation, the impulse responses themselves can be shortened, for example from the full frame length N to fewer points. For example, the length of the impulse response can be reduced from 1024 points (links) to 10 points. The most extreme truncation that can still be considered significant is . Other direct approximations can be found in US Patent Application No. 2005/0197831 A1.

You can also reduce the number of inputs on which the calculation is based. For illustration, the real-to-imaginary transformation 206 and its uplinks, which are indicated by the portion indicated in the illustration as "A", can be replaced by simplified versions, two of which, A' and A", are shown in figure 10. Option A' creates an approximate imaginary representation of the signal. In this case, the MDST calculation takes into account only the current and previous frames. With reference to the formula given above in this paragraph, this can be done by specifying when p =0, ..., N -1 (index III denotes the subsequent time frame). Since option A' does not require the MDCT spectrum of the subsequent frame as input, the MDST calculation does not suffer any time delay. Obviously, this approximation somewhat reduces the accuracy of the resulting MDST signal, but may also imply that the energy of this signal will become reduced; the latter fact can be fully compensated by increasing as a result of the essence of predictive coding.

Also shown in Figure 10 is Option A" which uses only the MDCT data for the current frame as input. It is possible that Option A" generates a less accurate MDST representation than Option A'. On the other hand, it, like option A', operates with zero delay and has less high computational complexity. As mentioned, the waveform encoding properties are not affected since the same approximation is used by the encoder system and the decoder system.

It should be noted that, regardless of whether options A, A' or A" or any improvement thereto are used, it is necessary to calculate only those parts of the MDST spectrum for which the imaginary part of the complex prediction coefficient is non-zero. . In practical terms, this means that the absolute value The imaginary part must be greater than a predefined threshold value, which can be related to the rounding unit of the hardware used. In the case where the imaginary part of the coefficient is zero for all frequency bands within a time frame, there is no need to calculate any MDST data for that frame. Thus, accordingly, the real to imaginary conversion 206 adapts to respond to the occurrence of values that are too small by not generating an MDST output signal, whereby computing resources can be saved. However, in embodiments of the invention where more frames than the current frame are used to generate one frame of MDST data, any elements upstream of transform 206 should continue to act accordingly even when MDST spectrum is not required—in particular, second switch node 205 should continue sending MDCT spectra so that enough input data is already available for real-to-imaginary conversion 206 when the next time frame associated with a non-zero prediction coefficient appears; Of course, this frame could be the next frame.

Returning to FIG. 2, the operation of the decoding system is described assuming that both switching nodes 203, 205 are set to their respective relay modes. As will be described below, the decoder system can also decode signals that are not predictively encoded. For such an application, the second switching node 205 is set to sum-difference mode and accordingly, as indicated in the illustration, the selector circuit 208 will be set to the down position, thus providing signals directly to the inverse transform 209 from the source location between the TNS filters 204 and a second switching node 205. To ensure correct decoding, the signal accordingly has an L/R form at the source location. Therefore, in order to ensure that the real-to-imaginary conversion is fed with the correct middle signal (i.e., the downmix signal) at all times (and not, say, periodically the left signal), the second switching unit 205 during decoding of the stereo signal, not subject to predictive coding, is preferably set to sum-differential mode. As discussed above, predictive coding can be replaced for some frames by traditional direct or joint coding, based, for example, on decision data about the relationship between speed and audio quality. The outcome of this decision may be communicated from the encoder to the decoder in various ways, for example, by the value of a dedicated indicator bit in each frame, or by the absence or presence of a prediction coefficient value. Once these facts are established, the role of the first switching node 203 becomes easily understood. Indeed, in the non-prediction encoding mode, the decoder system can process signals in accordance with both direct (L/R) stereo encoding and in accordance with joint (M/S) stereo encoding, and when operating the first switching node 203 or in a relay mode, or in the sum-difference mode it is possible to ensure that the source location will always be provided with a directly encoded signal. Obviously, switching unit 203, when operating as a sum-difference stage, will convert an input signal in M/S form to an input signal (supplied to optional TNS filters 204) in L/R form.

The decoder system receives a signal indicating whether the decoder system should decode a particular time frame in a predictive encoding mode or in a non-predictive encoding mode. Non-prediction mode may be signaled by the value of a dedicated indicator bit in each frame or by the absence (or zero value) of a prediction coefficient. The prediction mode may be reported similarly. A particularly favorable implementation, which makes downgrading possible without any overhead, uses the reserved fourth value of the two-bit ms_mask_present field (see MPEG-2 AAC document ISO/IEC 13818-7), which is transmitted in each frame and is determined in the following way:

Table 1: USAC ms_mask_present definition Meaning Meaning 00 L/R coding for all frequency bands 01 One signal bit per strip is used to indicate L/R or M/S 10 M/S coding for all frequency bands eleven Reserved

By redefining the value 11 to mean "complex prediction encoding", the decoder can operate in all compatibility modes, particularly M/S and L/R encoding modes, without any loss of bit rate and still is available for receiving a signal indicating a complex prediction coding mode for the corresponding frames.

Figure 4 shows a decoder system of the same general design as that shown in Figure 2, but including at least two other designs. First, the system of Figure 4 includes switches 404, 411 that enable any processing stage involving frequency domain modification to be applied upstream and/or downstream of the upmixing stage. This is accomplished, on the one hand, by a first set of frequency domain modifiers 403 (depicted in this figure as TNS synthesis filters) provided in conjunction with a first switch 404 downstream of the dequantization modules 401 and the first switch node 402, but in a downstream direction. relative to the second switching node 405 located in the upstream direction immediately before the upmixing stage 406, 407, 408, 409. On the other hand, the decoder system includes a second set of frequency domain modifiers 410 that are provided in conjunction with a second switch 411 downstream of the upmix stage 406, 407, 408, 409, but upstream of the deconversion stage 412. Advantageously, as illustrated, each frequency domain modifier is arranged in parallel with a relay line that is connected upstream of the input side of the frequency domain modifier and downstream of an associated switch. By virtue of this design, the frequency domain modifier is supplied with signal data at all times, making it possible to process the frequency domain based on more frames than just the current frame. The decision whether to apply the first 403 or the second 410 set of modifiers in the frequency domain may be made by the encoder (and transmitted in the bit stream) or may be based on whether predictive coding is applied, or may be based on any other criteria found appropriate in conditions of practical application. For example, if the frequency domain modifiers are TNS filters, then for some types of signals the use of the first set 403 may be advantageous, while the use of the second set 410 may be advantageous for other types of signals. If the result of this selection is encoded in the bitstream, then the decoder system will drive the appropriate set of TNS filters accordingly.

To facilitate understanding of the decoder system shown in Figure 4, it should be unequivocally noted that decoding of a directly encoded (L/R) signal occurs when α=0 (this implies that pseudo-L/R and L/R are identical, and that the side and residual signals are not different), the first switching node 402 is in the relay mode, the second switching node is in the sum-difference mode, which thus causes the signal between the second switching node 405 and the sum-difference stage The 409 upmix stages are M/S-shaped. Then, since the upmix stage will effectively be a relay stage, it does not matter whether the first or second set of modifiers are actuated (using the corresponding switches 404, 411).

Figure 3 illustrates a decoder system according to one embodiment of the invention, which, relative to the embodiments of Figures 2 and 4, represents a different approach to creating the MDST data needed for upmixing. Like the decoder systems already described, the system of Figure 3 includes dequantization modules 301, a first switching node 302 operating in either relay or sum-difference mode, and TNS (synthesis) filters 303, which are arranged in series from the input end. decoder systems. The modules downstream from this point are selectively used by two second switches 305, 310, which preferably operate together such that they are both in either the up positions or the down positions as indicated in the figure. At the output end of the decoder system is a sum-difference stage 312 and, immediately upstream of it, two inverse MDCT modules 306, 311 for converting the MDCT domain representation of each of the channels into a time domain representation.

In complex prediction decoding, where the decoder system is provided with a bitstream in which the downmix signal/residual stereo signal and complex prediction coefficient values are encoded, the first switch node 302 is set to the relay mode and the second switches 305, 310 are set to the up position. Downstream of the TNS filters, the two channels (dequantized, TNS filtered, MDCT) of the audio signal are processed differently. The downmix signal is supplied, on the one hand, to a multiplier and adder 308, which adds an MDCT representation of the downmix channel, weighted by the real part α _R of the prediction coefficient, to the MDCT representation of the residual channel and, on the other hand, to one of the modules 306 inverse MDCT transformation. The time domain representation of the downmix channel M, which is the output signal from the inverse MDCT transform module 306, is supplied to both the final sum-difference stage 312 and the MDST transform module 307. This dual use of the time domain representation of the downmix channel is advantageous in terms of computational complexity. The MDST representation of the downmix channel thus obtained is fed to a further multiplier and adder 309, which, after weighting by the imaginary part _αI of the prediction coefficient, adds this signal to the linear combination at the output of the adder 308; thus, the output of adder 309 is a side channel signal, . Similar to the decoder system shown in Figure 2, multipliers and adders 308, 309 can be easily combined to form a weighted multi-signal adder that outputs MDCT and MDST representations of the downmix signal, an MDCT representation of the residual signal, and a complex prediction coefficient value. Downstream from this point, in the present embodiment, all that remains is to pass through the inverse MDCT module 311 before the side channel signal is applied to the final sum-difference stage 312.

The necessary synchrony in the decoder system can be achieved by using the same transform lengths and window shapes for both inverse MDCT modules 306, 311, as is already practiced in frequency selective M/S and L/R coding. A combination of some embodiments of the inverse MDCT module 306 and some embodiments of the MDST module 307 introduces a delay of one frame. Therefore, five optional delay blocks 313 (or software instructions - to perform this action in a computer implementation) are provided so that the part of the system located to the right of the dotted line can be delayed one frame relative to the left part if necessary. Obviously, all intersections between the dotted line and the communication lines are provided with delay units with the exception of the communication line between the inverse MDCT module 306 and the MDST transform module 307, which is where the resulting delay needs to be compensated.

Computing MDST data for one time frame requires one frame's data from the time domain representation. However, the inverse MDCT is based on one (current), two (preferably previous and current), or three (preferably previous, current and next) consecutive frames. Due to the well-known TDAC rejection associated with MDCT, the three-frame capability performs complete overlap of input frames and thus provides the best (and likely perfect) accuracy, at least for frames containing TDAC. in the time domain. It is clear that three-frame inverse MDCT operates with a delay of one frame. By allowing the use of an approximate time domain representation as input to the MDST transform, this delay and thus the need to compensate for delays between different parts of the decoder system can be avoided. In a two-frame capability, aliasing/addition occurs in the earlier half of the frame, making TDAC possible, and spurious low-frequency components may only be present in the later half of the frame. In a single-frame capability, the absence of TDAC implies that low-frequency spurious components may appear in the frame; however, the MDST representation thus achieved when used as an intermediate signal in complex prediction coding can still provide satisfactory quality.

The decoding system illustrated in FIG. 3 can also operate in two non-predictive decoding modes. To decode a direct L/R stereo signal, the second switches 305, 310 are set to the down position and the first switch unit 302 is set to the relay mode. Thus, the signal has an L/R form in the upstream direction relative to the sum-difference stage 304, which converts it to the M/S form, on which the inverse MDCT transform and sum-difference operation occurs. To decode a stereo signal presented in co-encoded M/S form, the first switching node 302 is instead set to sum-difference mode so that the signal between the first switching node 302 and the sum-difference stage 304 has an L/R form that is often more suitable for TNS filtering than the M/S form would be.

The downstream processing with respect to the sum-difference stage 304 is identical to the processing in the case of direct L/R decoding.

Figure 14 consists of three general block diagrams of decoders according to embodiments of the invention. Unlike some of the other block diagrams accompanying this application, the communication line in Figure 14 may symbolize a multi-channel signal. In particular, said link may be arranged to transmit a stereo signal including left/right, mid/side, downmix/residual, pseudo left/pseudo right channels, and other combinations.

Figure 14A shows a decoder system configured to decode a frequency domain representation of an input signal (referred to for purposes of this figure as an MDCT representation). The decoder system is adapted to provide as an output a time domain representation of a stereo signal that is generated from the input signal. In order to be able to decode an input signal encoded by complex prediction stereo coding, the decoder system is equipped with an upmixing stage 1410. However, it is also capable of manipulating an input signal encoded in other formats, and possibly a signal that is interleaved in time between multiple encoding formats, for example, a sequence of time frames encoded using complex prediction coding may be followed by a time segment encoded via direct left/right encoding. The ability of the decoder system to manipulate various encoding formats is achieved by creating a communication line (relay) located in parallel with the specified upmix stage 1410. Switch 1411 can select whether the output signal of the upmix stage 1410 (lower switch position in the figure) or the dry signal available above the link (upper switch position in the figure) will be supplied to decoder modules further downstream. In this embodiment, the inverse MDCT module 1412, which converts the MDCT representation of the signal to a time domain representation, is located downstream of the switch. For example, the signal supplied to the upmix stage 1410 may be a stereo signal in downmix/residual form. The upmixing stage 1410 is adapted to receive the side signal and to perform a sum-difference operation such that the output signal is a left/right stereo signal (in the MDCT domain).

Figure 14B shows a decoder system similar to that of Figure 14A. The present system is adapted to receive a bit stream as an input signal. First, the bitstream is processed by a combined demultiplexer and dequantization module 1420, which produces as a first output an MDCT representation of a multi-channel stereo signal for further processing, which is determined by the position of a switch 1422 having the same function as switch 1411 in FIG. 14A. More specifically, switch 1422 determines whether the first output signal from the demultiplexer and dequantization should be processed by the upmixing stage 1421 and the inverse MDCT module 1423 (lower position) or only the inverse MDCT module 1423 (higher position). The combined demultiplexer and dequantization module 1420 also outputs control information. In the present case, the control information associated with the stereo signal may include data indicating which position of the switch 1422, lower or higher, is suitable for decoding the signal, or, more abstractly, according to which encoding format the stereo signal should be decoded. The control information may also include parameters for adjusting properties of the upmix stage 1421, for example, the value of the complex prediction coefficient α used in complex prediction coding, as already described above.

Figure 14C shows a decoder system that, in addition to entities similar to those in Figure 14B, includes first and second frequency domain modification devices 1431, 1435, respectively, located upstream and downstream of the upmix stage 1433. For purposes of this figure, each frequency domain modification device is illustrated by a TNS filter. However, the term "frequency domain modification device" can also refer to processes other than TNS filtering, which can be used either before or after the upmixing stage. Examples of modifications in the frequency domain include prediction, interference accumulation, bandwidth stretching, and nonlinear processing. Psychoacoustic considerations and similar reasons, which possibly include the properties of the signal being processed and/or the configuration or settings of the frequency domain modification apparatus, sometimes indicate that it is advantageous to apply said frequency domain modification in an upstream rather than downstream direction relative to upmix stages 1433. In other cases, similar considerations may determine that a downstream location of the frequency domain modification is preferable to an upstream location. By means of switches 1432, 1436, frequency domain modification devices 1431, 1435 can be selectively actuated so that, in response to control information, the decoder system can select the desired configuration. For example, Figure 14C shows a configuration in which a stereo signal from a combined demultiplexer and dequantization module 1430 is first processed by a first frequency domain modification device 1431, then supplied to an upmixer stage 1433, and finally sent directly to an inverse MDCT module 1437 without passing through the second device. 1435 modifications in the frequency domain. As explained in the Summary of the Invention section, in complex prediction coding, this configuration is preferred over the option of performing TNS after upmixing.

II. Encoder systems

An encoder system according to the invention will be described below with reference to FIG. 5, which is a generalized block diagram of an encoder system for encoding a left/right (L/R) stereo signal as an output bitstream through complex prediction encoding. The encoder system receives a time or frequency domain representation of the signal and feeds it to a downmixing stage and a prediction coefficient estimator. The real and imaginary parts of the prediction coefficients are passed to a downmix stage to control the conversion of the left and right channels into a downmix channel and a residual channel, which are then fed to the final MUX. If the signal is not supplied to the encoder as a frequency domain representation, then it is converted to such a representation in a downmixing stage or multiplexer.

One of the principles of predictive coding is to transform the left/right signal into a middle/side form, i.e.:

,

and then using the residual correlation between these channels, i.e. exercise

,

where α is the complex prediction coefficient to be determined, andD - residual signal. You can choose α so that the energy of the residual signal minimized. Energy minimization can be performed relative to instantaneous power, short-term or long-term energy (average power), which in the case of a discrete signal is equivalent to optimization in the rms value.

Real and imaginary parts of α_R, α_Iprediction coefficients may be quantized and/or jointly encoded. It is preferable, however, that the real and imaginary parts are quantized independently and uniformly, typically with a step size of 0.1 (a dimensionless number). According to the MPEG standard, the resolution of the frequency band used for the complex prediction coefficient is not necessarily the same as the resolution of the scale factor bands (sfb; those. groups of MDCT lines that use the same MDCT quantization step size and quantization range). In particular, the resolution of the frequency band for the prediction coefficient can be justified psychoacoustically, such as the Bark scale. It should be noted that the resolution of the frequency band may change if the conversion length changes.

As already noted, the encoder system of the invention may have the freedom to use predictive stereo coding or not, where the latter case implies a downgrade back to L/R or M/S coding. This decision may be made on a time frame basis or, more precisely, on a frequency band within a time frame. As noted above, a negative decision result may be communicated to the decoding entity in various ways, for example, by the value of a dedicated indicator bit in each frame, or by the absence (or zero value) of a prediction coefficient value. A positive decision can be conveyed in a similar way. A particularly advantageous implementation, which makes downgrading possible without any overhead, uses the reserved fourth value of the two-bit ms_mask_present field (see MPEG-2 AAC document ISO/IEC 131818-7), which is transmitted in each frame and is determined in the following way:

Table 1: USAC ms_mask_present definition Meaning Meaning 00 L/R coding for all frequency bands 01 One signal bit per strip is used to indicate L/R- or M/S 10 M/S coding for all frequency bands eleven Reserved

By redefining the value 11 to mean "complex prediction coding", the encoder can operate in all compatibility modes, particularly M/S and L/R coding modes, without any loss of bit rate and is still capable of signaling complex prediction coding for those frames where it is advantageous.

The decision may be based on data transfer speed versus audio quality considerations. Data obtained using a psychoacoustic model embedded in the encoder can be used as a measure of quality (as is often the case with commercially available MDCT-based audio encoders). In particular, some encoder embodiments select a prediction coefficient that is optimized based on rate versus distortion. Accordingly, in such embodiments, the imaginary part, and possibly also the real part, of the prediction coefficient is set to zero in the event that the prediction gain does not save enough bits to encode the residual signal so as to justify the cost of the bits required to prediction coefficient encoding.

Embodiments of the encoder may encode TNS-related information into a bit stream. This information may include TNS parameter values that are applied by TNS (synthesis) filters on the decoder side. If identical TNS parameter sets are used for both channels, it is more economical to include a signal bit indicating this parameter set identity rather than transmitting the two parameter sets independently. Information may also be included as to whether TNS should be applied before or after the upmixing stage, as appropriate based on, for example, a psychoacoustic assessment of both available capabilities.

As another optional feature that is potentially beneficial in terms of complexity and bit rate, the encoder may be adapted to use a separately limited bandwidth when encoding the residual signal. Frequency bands above this limit will not be transmitted to the decoder, but will be set to zero. In some cases, the highest frequency bands have such low energy content that they are quantized to zero. Standard practice (cf. the max_sfb parameter in the MPEG standard) typically uses the same bandwidth limit for both the downmix signal and the residual signal. To date, the inventors have experimentally discovered that the residual signal, to a greater extent than the downmix signal, has energy content localized in lower frequency bands. Therefore, by placing a dedicated bandwidth upper limit on the residual signal, it is possible to reduce the data bit rate without significant loss of quality. For example, this could be accomplished by passing two independent max_sfb parameters: one for the downmix signal, and one for the residual signal.

It should be noted that, although the issues of optimal determination of the prediction coefficient, its quantization and encoding, downshifting to the M/S or L/R mode, TNS filtering, upper bandwidth limitation, etc. have been discussed with respect to the decoder system shown in Figure 5, the same facts apply equally to the embodiments of the invention that will be disclosed below with reference to the following figures.

Figure 6 shows another encoder system according to the invention adapted to perform stereo complex prediction coding. The system takes as input a time domain representation of a stereo signal, segmented into successive, possibly overlapping time frames and including left and right channels. A sum-difference stage 601 converts the signal into the middle and side channels. The middle channel is supplied to both the MDCT module 602 and the MDST module 603, while the side channel is supplied only to the MDCT module 604. The prediction coefficient estimator 605 estimates for each time frame—and possibly for individual frequency bands within the frame—the value of the complex prediction coefficient α, as explained above. The value of the coefficient α is supplied as a weighting factor to the weighted adders 606, 607, which generate the residual signal D as a linear combination of the MDCT and MDST representations of the average signal and the MDCT representation of the spurious signal. Preferably, the complex prediction coefficient is supplied to the weighted adders 606, 607 represented in the same quantization scheme that will be used when encoding it into the bit stream; this obviously provides a more reliable reconstruction since both the encoder and decoder apply the same value of the prediction coefficient. The residual signal, the average signal (which is more appropriately called a downmix signal when it occurs in combination with the residual signal), and the prediction coefficient are supplied to a combined quantization and multiplexer stage 608, which encodes them and possible additional information into the output bitstream.

Figure 7 shows a modification of the encoder system of Figure 6. As can be seen from the similarity of symbols in the figure, it has a similar design, but also contains the additional function of operating in the reverse downshift mode of direct L/R encoding. The encoder system is driven between a complex prediction encoding mode and a downshift mode by a switch 710 provided in the upstream direction immediately before the combined quantization and multiplexer stage 709. In the up position, as shown in the figure, switch 710 will cause the encoder to operate in a downshift mode. From points in the downstream direction directly behind the MDCT modules 702, 704, the average/side signal is fed to a sum-difference stage 705 which, after being converted to left/right form, passes it to a switch 710 which couples it to a combined quantization and multiplexer stage 709.

Figure 8 shows an encoder system according to the present invention. Unlike the encoder systems of Figures 6 and 7, this embodiment of the invention obtains the MDST data required for complex prediction coding directly from the MDCT data, i.e. by transforming the real into the imaginary in the frequency domain. The real-to-imaginary conversion employs any of the approaches described in connection with the decoder systems of Figures 2 and 4. It is important that the calculation method in the decoder matches the calculation method in the encoder so that reliable decoding can be performed; Preferably, the encoder side and the decoder side use the same real-to-imaginary conversion techniques. As with the decoder embodiments, the portion A enclosed by the dotted line and including the real-to-imaginary transform 804 may be replaced by approximations using fewer input time frames as input. Likewise, encoding can be simplified using any of the methods described above, or some other approximate method.

At a higher level, system 8 has a design that is different from the design that would follow the direct action of replacing the MDST module in Figure 7 with a (properly connected) real-to-imaginary conversion module. The present architecture is straightforward and performs the function of switching between predictive coding and direct L/R coding in a robust and computationally efficient manner. The input stereo signal is supplied to MDCT conversion modules 801, which output a frequency domain representation of each of the channels. It is supplied to both the limit switch 808 for driving the encoder system between predictive and direct coding modes, and to the sum-difference stage 802. In direct L/R coding, or joint M/S coding, which is carried out in time frame for which the prediction coefficient is set to zero, this embodiment of the invention subjects the input signal only to MDCT, quantization and multiplexing, where the last two steps are performed by a combined quantization and multiplexer stage 807 located at the output end of the system to which the bit stream is supplied. In predictive coding, each of the channels undergoes further processing between the sum-difference stage 802 and the switch 808. From the MDCT representation of the average signal, the real-to-imaginary transform 804 obtains the MDST data and feeds it to both the prediction coefficient estimator 803 and the weighted adder 806. As in the encoder systems of FIGS. 6 and 7, a further weighted adder 805 is used to combine the side signal with the weighted MDCT and MDST representations of the average signal to form a residual channel signal that is co-encoded with the average channel signal (i.e. downmix) and with a prediction coefficient using a combined quantizer and multiplexer 807.

With reference to Figure 9, it will be illustrated that each of the encoder system embodiments may be combined with one or more TNS (analysis) filters. Consistent with previous discussions, it is often advantageous to apply TNS filtering to the signal in its downmix form. Therefore, as shown in Figure 9, adapting the encoder system of Figure 7 to include TNS is accomplished by adding TNS filters 911 upstream just before the combined quantizer and multiplexer module 909.

Instead of the right/residual TNS filter 911b in the downstream direction immediately behind the portion of the switch 910 adapted for manipulation of the right or residual channel, two separate TNS filters (not shown) may be provided. Thus, each of the two TNS filters will be supplied with the signal data of the corresponding channel at all times, which makes it possible to TNS filtering based on one or more time frames, and not just the current frame. As noted, TNS filters are just an example of frequency domain modification devices, particularly devices that base processing on more frames than just the current frame, which when so positioned can benefit as much or more than the TNS does. filter.

As another possible alternative to the embodiment of the invention shown in Figure 9, TNS filters designed for selective actuation may be placed at more than one point for each channel. This design is similar to the decoder system design shown in Figure 4, where different sets of TNS filters can be coupled via switches. This allows you to select the most appropriate TNS filtering stage for each time frame. In particular, it may be advantageous to switch between different TNS positions in connection with switching between complex prediction stereo coding and other coding modes.

Figure 11 shows a modification based on the encoder system of Figure 8, in which a second frequency domain representation of the downmix signal is obtained by real-to-imaginary transformation 1105. Similar to the decoder system shown in Figure 4, this encoder system also includes selectively driven frequency domain modifier modules: one 1102 provided upstream of the downmix stage, and one 1109 provided downstream of that stage. Frequency domain modules 1102, 1109, exemplified by TNS filters in this figure, may communicate with each of the signal paths using four switches 1103a, 1103b, 1109a, and 1109b.

III. Non-hardware embodiments of the invention

Embodiments of the third and fourth features of the invention are shown in Figures 15 and 16. Figure 15 shows a method for decoding a bit stream into a stereo signal, which includes the following steps: 1. The input signal is a bit stream. 2. The bit stream is dequantized, whereby a frequency domain representation of the downmix channel and the residual channel is obtained. 3. A second frequency domain representation of the downmix channel is calculated. 4. Based on the three frequency domain channel representations, the side channel signal is calculated. 5. Based on the side channel and the downmix channel, a stereo signal is calculated, preferably in left/right form. 6. The output signal is the stereo signal thus obtained. Steps 3-5 can be thought of as an upmixing process. Each of steps 1-6 is analogous to the corresponding function in any of the decoder systems disclosed in previous parts of this text, and other details relating to their implementation can be found in the same parts.

Figure 16 shows a method for encoding a stereo signal as a bitstream signal, which includes the following steps: 1. The input signal is a stereo signal. 2. The stereo signal is converted into a first frequency domain representation. 3. The coefficient of complex prediction is determined. 4. The frequency domain representation is downmixed. 5. The downmix signal and the residual signal are encoded as a bit stream together with a complex prediction coefficient. 6. The output signal is a bit stream. Each of steps 1-5 is analogous to the corresponding function in any of the encoder systems disclosed in the previous parts of this text, and other details relating to their implementation can be found in the same parts.

Both methods can be expressed as machine-readable instructions in the form of software programs and can be executed by a computer. The scope of protection of the present invention extends to said software and computer software products intended to distribute said software.

IV. Experimental evaluation

Some of the embodiments of the invention disclosed herein have been quantitatively evaluated experimentally. This subsection will summarize the most important parts of the experimental material obtained in this process.

The embodiment of the invention used for the experiments had the following characteristics: (i) each MDST spectrum (for a time frame) was calculated by finite duration two-dimensional impulse response filtering from the current, previous and next MDCT spectra. (ii) The psychoacoustic model from the USAC stereo encoder was used. (iii) Instead of the parameters PS, ICC, CLD and IPD, the real and imaginary parts of the complex prediction coefficient were transmitted. The real and imaginary parts were processed independently, constrained to the interval [-3.0, 3.0] and quantized using a step size of 0.1. They were then time-differentiated and finally Huffman-coded using the USAC scale factor codebook. The prediction coefficients were updated every other scale factor band, resulting in a frequency resolution that is close to that of MPEG Surround (see, for example, ISO/IEC 230031). This quantization and encoding scheme results in an average bit rate of approximately 2 Kbps for said stereo ancillary information in a typical configuration with a target bit rate of 96 Kbps. (iv) The bitstream format is modified without breaking the current USAC bitstreams because the 2-bit bitstream element ms_mask_present currently has only three possible values. The use of a fourth value to indicate complex prediction allows for a mid/side coding downgrade mode without losing bits (see the previous subsection of this disclosure for further details on this topic).

Test listening was performed in accordance with the MUSHRA method, which consists of individual playback through headphones and the use of 8 test elements with a sampling frequency of 48 kHz. This test involved three, five or six test subjects. The impact of different MDST approximations was evaluated to illustrate the practical complexity-quality trade-off that exists between the specified capabilities. The results are in Figures 12 and 13, where the former shows the absolute scores obtained and the latter shows the differential scores relative to 96s USAC cp1f , i.e. relative to unified stereo coding in the MDCT domain with complex prediction using the current MDCT frame to calculate the MDST approximation. As can be seen, the quality gain achieved by MDCT-based unified stereo coding increases when computationally more complex approaches to computing the MDST spectrum are used. In terms of overall test average, the single frame based 96s USAC cp1f system provides a significant increase in encoding efficiency over traditional stereo encoding. In turn, even better results are obtained for 96s USAC cp3f , namely unified stereo coding in the MDCT domain with complex prediction using the current, previous and next MDCT frames to calculate the MDST.

V. Embodiments of the Invention

The invention may also be implemented as a decoder system for decoding a bitstream signal into a stereo signal by means of complex prediction stereo coding, wherein the decoder system includes: a dequantization stage (202; 401) for creating first representations of the downmix signal (M); and a frequency domain residual signal (D) based on the bit stream signal, where each of the first frequency domain representations includes first spectral components that represent the spectral composition of the corresponding signal expressed in a first subspace of the multidimensional space, where the first spectral components represent transform coefficients, located in one or more time frames of transform coefficients, where each block is generated by applying a transform to a time segment of the time domain signal; and - an upmixing stage (206, 207, 210, 211; 406, 407, 408, 409) located downstream of the dequantization stage, adapted to generate a stereo signal based on the downmixing signal and the residual signal and including: - - a module (206; 408) for calculating a second frequency domain representation of the downmix signal based on its first frequency domain representation, wherein the second frequency domain representation includes second spectral components representing the spectral composition of the signal expressed in a second subspace of the multidimensional space that includes a portion of the multidimensional space not included in the first subspace, wherein said module is adapted to: - - - obtain one or more first intermediate components from at least some of the first spectral components; - - - forming a combination of said one or more spectral components in accordance with at least a part of one or more impulse responses to obtain one or more second intermediate components; and - - - obtaining said one or more second spectral components from said one or more second intermediate components; - - a weighted adder (210, 211; 406, 407) for calculating the side signal (S) based on the first and second frequency domain representations of the downmix signal, the first frequency domain representation of the residual signal, and the complex prediction coefficient (α), bit stream encoded in the signal; and - a sum-difference stage (207; 409) for calculating the stereo signal based on the first frequency domain representation of the downmix signal and the side signal.

Furthermore, the invention may be implemented as a decoder system for decoding a bitstream signal into a stereo signal by means of complex prediction stereo coding, wherein the decoder system includes: a dequantization stage (301) for creating first frequency domain representations of the downmix signal ( M) and a residual signal (D) based on the bitstream signal, where each of the first frequency domain representations includes first spectral components representing the spectral composition of the corresponding signal expressed in a first subspace of the multidimensional space; and - an upmixing stage (306, 307, 308, 309, 312), located in a downstream direction relative to the dequantization stage, adapted to generate a stereo signal based on the downmixing signal and the residual signal and including: - - a module (306, 307), configured to compute a second frequency domain representation of the downmix signal based on its first frequency domain representation, wherein the second frequency domain representation includes second spectral components representing the spectral composition of the signal expressed in a second subspace of the multidimensional space that includes a portion of the multidimensional space, not included in the first subspace, where the module includes: - - - an inverse transform stage (306) for calculating a time domain representation of the downmix signal based on the first frequency domain representation of the downmix signal in the first subspace of the multidimensional space; and - - - a conversion stage (307) for calculating a second frequency domain representation of the downmix signal based on the time domain representation of the signal; - - a weighted adder (308, 309) for calculating the side signal (S) based on the first and second frequency domain representations of the downmix signal, the first frequency domain representation of the residual signal, and the complex prediction coefficient (α) encoded in the bit signal flow; and - a sum-difference stage (312) for calculating the stereo signal based on the first representation of the frequency domain downmix signal and the side signal.

In addition, the invention may be implemented as a decoder system with the characteristics set forth in the independent claims relating to the decoder system, where the module for computing a second frequency domain representation of the downmix signal includes: an inverse conversion stage (306), for calculating a time domain representation of the downmix signal and/or side signal based on a first frequency domain representation of the corresponding signal in a first subspace of the multidimensional space; and a transform stage (307) for calculating a second frequency domain representation of the corresponding signal based on the time domain representation of that signal, where preferably the inverse transform stage (306) performs an inverse modified discrete cosine transform, and the transform stage performs a modified discrete cosine transform. sine transformation.

In the decoder system described above, the stereo signal may be represented in the time domain, and the decoder system may include: a switching unit (302) located between said dequantization stage and said upmixing stage, operable to perform the following functions: (a) a relay stage - for use in joint stereo coding; or (b) a sum-difference stage - for use in direct stereo coding; - an additional inverse conversion stage (311), located at the upmixing stage, for calculating the time domain representation of the spurious signal; - a switching node (305, 310) located in the upstream direction relative to the inverse conversion stages (306, 301), adapted for their selective coupling with: (a) an additional sum-difference stage (304), which, in turn, is connected with a point in the downstream direction relative to the switching node (302) and with a point in the upward direction relative to the upmixing stage; or (b) a downmix signal obtained from the switching node (302) and a side signal obtained from the weighted adder (308, 309).

VI. Concluding remarks

Further embodiments of the present invention will be apparent to those skilled in the art after reading the above description. Although the present description and drawings disclose embodiments and examples of the invention, the invention is not limited to these specific examples. Numerous modifications and changes may be made without departing from the scope of the present invention, which is defined by the accompanying claims.

It should be noted that the methods and apparatus disclosed in this application, after appropriate modifications, can be used within the capabilities of one skilled in the art, including typical experiments, to encode signals containing more than two channels. It is especially emphasized that any signals, parameters and matrices mentioned in connection with the described embodiments of the invention may be frequency dependent or frequency independent and/or time dependent or time independent. The described computational steps may be performed frequency by frequency or for all frequency bands at a time, and all objects may be carried out so that they have a frequency selective effect. For application purposes, any quantization schemes may be adapted according to psychoacoustic models. In addition, it should be noted that all various sum-difference transformations, i.e. conversions from downmix/residual form to pseudo-L/R form, as well as L/R to M/S conversion and M/S to L/R conversion, are of the form

,

where only the gain can change. Thus, by adjusting the gains individually, a certain encoding gain can be compensated by appropriately selecting the decoding gain. In addition, as those skilled in the art will appreciate, an even number of sum-difference transforms in succession produces the effect of a relay stage, possibly with a non-unity gain.

The systems and methods described above herein may be implemented as software, firmware, hardware, or a combination thereof. Some or all of the components may be implemented as software executed by a digital signal processor or microprocessor, or implemented as hardware, or as a special purpose integrated circuit. The software may be distributed on computer readable media, which may include computer storage media or communications media. As is well known to those skilled in the art, computer storage media includes volatile and non-volatile, removable and non-removable media implemented in any method or technology designed to store information in memory, such as machine-readable instructions, data structures, program modules or other data. Computer storage media includes, but is not limited to, RAM, ROM, EEPROM, flash memory or other memory technologies, CD-ROM, CD-ROM or other optical disk media, magnetic tape cassettes, magnetic tape, magnetic disk media or other magnetic storage devices, or any other medium that can be used to store needed information in memory and that can be accessed by a computer. In addition, those skilled in the art will recognize that communications media typically embody machine-readable instructions, data structures, program modules, or other data in a modulated data signal, such as a carrier wave or other transmission mechanism, and include any information delivery media .

Claims (59)

1. A decoder system for providing a stereo signal by complex prediction stereo coding, the decoder system comprising: an upmix stage adapted to generate a stereo signal based on first frequency domain representations of the downmix signal (M) and a residual signal (D), each of the first frequency domain representations comprising first spectral components that represent the spectral composition of the corresponding signal expressed in the first subspace of the multidimensional space, with the upmixing stage: calculates a second frequency domain representation of the downmix signal based on its first frequency domain representation, wherein the second frequency domain representation includes second spectral components representing the spectral composition of the signal expressed in a second subspace of the multidimensional space, which includes a portion of the multidimensional space not included in a first subspace, wherein second spectral components of the downmix signal are determined by applying a finite impulse response (FIR) filter to the first spectral components of the downmix signal; calculates a side signal (S) based on the first and second frequency domain representations of the downmix signal, the first frequency domain representation of the residual signal, and the complex prediction coefficient (α) encoded in the bitstream signal; And calculates a stereo signal based on the first frequency domain representation of the downmix signal and the side signal, wherein the upmix stage is adapted to apply independent bandwidth limits to the downmix signal and the residual signal. 2. The decoder system of claim 1, wherein the impulse response of the FIR filter is determined depending on a window function used to determine the first frequency domain representation of the downmix signal. 3. The decoder system of claim 1, wherein the limits of the bandwidths to be applied are signaled by two data fields indicating for each of the signals the highest frequency band to be decoded. 4. The decoder system according to claim 3, characterized in that it is adapted to receive an MPEG bitstream in which each of said data fields is encoded as the value of the max_sfb parameter of the MPEG bitstream . 5. The decoder system according to claim 1, characterized in that it additionally contains: a dequantization stage located upstream of the upmix stage for providing said first representations of the downmix signal (M) and the residual signal (D) in the frequency domain based on the bit stream signal. 6. The decoder system according to claim 1, characterized in that: the first spectral components have real values expressed in the first subspace; the second spectral components have imaginary values expressed in the second subspace. 7. The decoder system according to claim 1, characterized in that the first spectral components are configured to be obtained by one of the following: discrete cosine transform (DCT) or modified discrete cosine transform (MDCT). 8. The decoder system according to claim 1, characterized in that it further comprises at least one temporal noise shaping (TNS) module located upstream of the upmixing stage; And at least one additional TNS module located downstream of the upmixing stage; And selector circuit for selective activation or: (a) said TNS module(s) upstream of the upmix stage, or (b) said additional TNS module(s) downstream of the upmix stage. 9. The decoder system according to claim 6, characterized in that the downmix signal is divided into successive time frames, each of which is associated with a complex prediction coefficient value; And calculation of the second frequency domain representation of the downmix signal is deactivated in response to the fact that the absolute value of the imaginary part of the complex prediction coefficient is less than a predetermined tolerance for the time frame. 10. The decoder system according to claim 1, characterized in that said stereo signal is represented in the time domain, and the decoder system also comprises: a switching unit located between said dequantization stage and said upmixing stage, configured to function as: (a) a relay stage or (b) sum-difference stage, thereby allowing switching between directly and co-encoded stereo input signals; an inverse conversion stage adapted to compute a time domain representation of the stereo signal; And a selector circuit located upstream of the inverse conversion stage, adapted to selectively couple it with either: (a) at a point downstream of the upmixing stage, whereby the stereo signal obtained by complex prediction is supplied to the deconversion stage, or with (b) at a point downstream of the switching unit and upstream of the upmixing stage, whereby the stereo signal obtained by forward stereo encoding is supplied to the deconversion stage. 11. A decoding method for upmixing an input stereo signal by complex prediction stereo encoding into an output stereo signal, wherein: said input stereo signal comprises first frequency domain representations of a downmix channel (M) and a residual channel (D) and a complex prediction coefficient (α); And each of said first frequency domain representations contains first spectral components representing the spectral composition of the corresponding signal expressed in a first subspace of the multidimensional space, wherein the method is performed by an upmixing stage and includes the steps: calculating a second frequency domain representation of the downmix channel based on its first frequency domain representation, the second frequency domain representation comprising second spectral components representing the spectral composition of the signal expressed in a second subspace of the multidimensional space, which contains a portion of the multidimensional space not included in a first subspace, wherein calculating a second frequency domain representation of the downmix signal includes determining second spectral components of the downmix signal by applying a finite impulse response (FIR) filter to the first spectral components of the downmix signal; And calculating a side channel based on the first and second frequency domain representations of the downmix signal, the first frequency domain representation of the residual signal, and the complex prediction coefficient, however, independent bandwidth limits are applied to the downmix signal and the residual signal. 12. The method of claim 11, wherein the impulse response of the FIR filter is determined depending on a window function used to determine the first frequency domain representation of the downmix signal. 13. The method of claim 11, characterized in that the limits of the bandwidths to be applied are signaled by two data fields indicating for each of the signals the highest frequency band to be decoded. 14. The method of claim 13, further comprising receiving an MPEG bitstream in which each of said data fields is encoded as the value of the max_sfb parameter of the MPEG bitstream. 15. The method according to claim 11, characterized in that it additionally includes providing said first frequency domain representations of the downmix signal (M) and the residual signal (D) based on the bitstream signal. 16. Method according to claim 11, characterized in that the first spectral components have real values expressed in the first subspace; the second spectral components have imaginary values expressed in the second subspace. 17. The method according to claim 11, characterized in that the first spectral components are configured to be obtained by one of the following: discrete cosine transform (DCT) or modified discrete cosine transform (MDCT). 18. The method according to claim 11, characterized in that it additionally includes selective execution or: (a) time-shaping noise (TNS) processing upstream of the upmix stage, or (b) TNS processing downstream of the upmixing stage. 19. The method according to claim 16, characterized in that the downmix signal is divided into successive time frames, each of which is associated with a complex prediction coefficient value; And do not calculate a second frequency domain representation of the downmix signal if the absolute value of the imaginary part of the complex prediction coefficient is less than a predetermined tolerance for the time frame. 20. A non-volatile computer-readable medium on which instructions are stored that, when executed by a general-purpose computing device, carry out the method described in clause 11.