RU2841624C1 - Backward-compatible integration of high-frequency reconstruction techniques for audio signals - Google Patents

Abstract

FIELD: computer equipment.

SUBSTANCE: invention relates to computer engineering for processing audio data. Result is achieved due to the stages, at which the decoded audio signal in the low-frequency band is filtered in order to generate the filtered audio signal in the low-frequency band, wherein filtering is performed using analysis filter bank, which includes analysis filters, h_k(n), which are modulated versions of the prototype filter, p₀(n), and regenerating part of the high frequency band of the audio signal using the filtered audio signal in the low frequency band and high frequency reconstruction metadata, wherein the regeneration includes spectral transfer if the patch insertion mode parameter has a first value, and the regeneration includes harmonic transposition by frequency extension of the phase vocoder if the patch insertion mode parameter has a second value.

EFFECT: improved spectral range replication.

8 cl, 7 dwg

Description

Cross reference to related application

This application claims priority from the following priority application: U.S. Provisional Application No. 62/622,205, filed January 26, 2018, which is incorporated herein by reference.

Field of technology

Embodiments relate to audio signal processing and, more particularly, to encoding, decoding, or transcoding audio bitstreams with control data specifying that either a basic form of high frequency reconstruction (“HFR”) or an enhanced form of HFR is to be performed on the audio data.

Prior art

A typical audio bitstream includes both audio data (e.g., encoded audio data) indicating one or more channels of audio content and metadata indicating at least one characteristic of the audio data or audio content. One well-known format for generating an encoded audio bitstream is the MPEG-4 Advanced Audio Coding (AAC) format, described in the ISO/IEC 14496-3:2009 MPEG standard. In the MPEG-4 standard, AAC stands for “Advanced Audio Coding,” and HE-AAC stands for “High Efficiency Advanced Audio Coding.”

The MPEG-4 AAC standard defines several audio profiles that determine which objects and coding tools are present in a compliant encoder or decoder. Three of these audio profiles are (1) the AAC profile, (2) the HE-AAC profile, and (3) the HE-AAC v2 profile. The AAC profile includes the AAC Low Complexity (or “AAC-LC”) object type. The AAC-LC object is equivalent to the MPEG-2 AAC Low Complexity profile, with some adjustments, and does not include either the Spectral Range Replication (“SBR”) object type or the Parametric Stereo (“PS”) object type. The HE-AAC profile is a superset of the AAC profile and additionally includes the SBR object type. The HE-AAC v2 profile is a superset of the HE-AAC profile and additionally includes the PS object type.

The SBR object type contains the spectral range replication tool, which is an important tool in high-frequency reconstruction (“HFR”) coding, which significantly improves the compression efficiency of perceptual audio codecs. SBR reconstructs the high-frequency components of the audio signal at the receiver side (e.g., in the decoder). Thus, the encoder only needs to encode and transmit the low-frequency components, providing increased audio quality at low bit rates. SBR is based on the replication of harmonic sequences, previously truncated to reduce the bit rate, from the available bandwidth-limited signal and control data received from the encoder. The relationship between tonal and noise-like components is maintained by adaptive inverse filtering and optional addition of noise and sine waves. In the MPEG-4 AAC standard, the SBR tool performs spectral patching (also called linear transfer or spectral transfer) where a number of consecutive quadrature mirror filter (QMF) subbands are copied (or "patched") from the transmitted low-band portion of the audio signal to the high-band portion of the audio signal generated at the decoder.

Spectral patch insertion or linear transfer may not be ideal for some types of audio, such as music content with relatively low crossover frequencies. Therefore, methods to improve spectral range replication are needed.

Brief description of embodiments of the invention

A first class of embodiments is disclosed that relates to a method for decoding an encoded audio bitstream. The method includes receiving an encoded audio bitstream and decoding audio data to generate a decoded audio signal in a low frequency band. The method further includes extracting high-frequency reconstruction metadata and filtering the decoded audio signal in the low frequency band with an analysis filter bank to generate a filtered audio signal in the low frequency band. The method further includes extracting a label indicating whether a spectral transfer or a harmonic transposition should be performed with respect to the audio data, and regenerating a portion of the high frequency band of the audio signal using the filtered audio signal in the low frequency band and the high-frequency reconstruction metadata according to the label. Finally, the method includes combining the filtered audio signal in the low frequency band and the regenerated portion of the high frequency band to form a wideband audio signal.

A second class of embodiments relates to an audio decoder for decoding an encoded audio bitstream. The decoder includes an input interface for receiving an encoded audio bitstream, where the encoded audio bitstream includes audio data representing a low-frequency band portion of an audio signal, and a core decoder for decoding the audio data to generate a decoded audio signal in the low frequency band. The decoder also includes a demultiplexer for extracting high-frequency reconstruction metadata from the encoded audio bitstream, where the high-frequency reconstruction metadata includes operating parameters for a high-frequency reconstruction process that linearly maps a number of successive subbands from a low-frequency band portion of the audio signal to a high-frequency band portion of the audio signal, and an analysis filter bank for filtering the decoded audio signal in the low frequency band to generate a filtered audio signal in the low frequency band. The decoder further includes a demultiplexer for extracting from the encoded audio bitstream a mark indicating whether a linear transfer or a harmonic transposition should be performed with respect to the audio data, and a high-frequency regenerator for regenerating a portion of the high-frequency band of the audio signal using the filtered audio signal in the low-frequency band and high-frequency reconstruction metadata in accordance with the mark. Finally, the decoder includes a synthesis filter bank for combining the filtered audio signal in the low-frequency band and the regenerated portion of the high-frequency band to form a wideband audio signal.

Other classes of embodiments relate to encoding and transcoding audio bitstreams containing metadata identifying whether to perform enhanced spectral range replication (eSBR) processing.

Brief description of the drawings

Fig. 1 is a block diagram of an embodiment of a system that can be configured to perform an embodiment of the method according to the invention.

Fig. 2 is a block diagram of an encoder which is an embodiment of an audio processing module according to the invention.

Fig. 3 is a block diagram of a system including a decoder, which is an embodiment of the audio processing module according to the invention, and optionally also a post-processor associated therewith.

Fig. 4 is a block diagram of a decoder which is an embodiment of an audio processing module according to the invention.

Fig. 5 is a block diagram of a decoder which is another embodiment of the audio processing module according to the invention.

Fig. 6 is a block diagram of another embodiment of an audio processing module according to the invention.

Fig. 7 is a diagram of an MPEG-4 AAC bitstream block including the segments into which it is divided.

Designations and Terminology

Throughout this disclosure, including in the claims, the expression performing an operation “on” a signal or data (e.g., filtering, scaling, transforming, or applying gain to a signal or data) is used broadly to mean performing the operation directly on the signal or data or on a processed version of the signal or data (e.g., on a version of the signal that has been pre-filtered or pre-processed prior to performing the operation on it).

Throughout this disclosure, including the claims, the term “audio processing module” or “audio processor” is used in a broad sense to refer to a system, device, or apparatus configured to process audio data. Examples of audio processing modules include, but are not limited to, encoders, transcoders, decoders, codecs, pre-processing systems, post-processing systems, and bitstream processing systems (sometimes referred to as bitstream processing tools). Virtually all consumer electronics, such as mobile phones, televisions, laptops, and tablets, include an audio processing module or audio processor.

Throughout the present disclosure, including the claims, the term "connects" or "connected" is used in a broad sense to mean a direct or indirect connection. Thus, if a first device is connected to a second device, this connection may be carried out as a direct connection or an indirect connection through other devices and connections. Moreover, components integrated into other components or with other components are also connected to each other.

Detailed description of embodiments of the invention

The MPEG-4 AAC standard implies that an encoded MPEG-4 AAC bitstream includes metadata indicating each type of high frequency reconstruction (“HFR”) processing that is to be applied (if any) by a decoder to decode the audio content of the bitstream, and/or that controls such HFR processing and/or indicates at least one characteristic or parameter of at least one MPEG-4 AAC tool used to decode the audio content of the bitstream. The expression “SBR metadata” is used here to denote metadata of the type described or referred to in the MPEG-4 AAC standard for use with spectral range replication (“SBR”). As will be understood by one skilled in the art, SBR is a form of HFR.

SBR is preferably used as a double-rate system, with the base codec running at half the original sampling rate, while SBR runs at the original sampling rate. The SBR encoder runs in parallel with the main base codec, albeit at a higher sampling rate. Although SBR is primarily a post-processing step in the decoder, important parameters are extracted in the encoder to ensure the most accurate high-frequency reconstruction in the decoder. The encoder estimates the spectral envelope of the SBR band for a time and frequency range/resolution appropriate to the current characteristics of the input signal segments. The spectral envelope is estimated by means of a complex QMF analysis and subsequent energy calculation. The time and frequency resolutions of the spectral envelopes can be chosen with a high level of freedom to provide the most appropriate time/frequency resolution for a given input segment. The envelope estimation must take into account that the transient in the original, mainly high-frequency region (e.g. high-hat) will be present to a lesser extent in the SBR-generated high range before envelope correction, since the high range in the decoder is based on the low range, where the transient is much less pronounced compared to the high range. This aspect imposes different requirements for the time/frequency resolution of the spectral envelope data compared to the usual spectral envelope estimation as used in other audio coding algorithms.

In addition to the spectral envelope, several additional parameters are extracted that represent the spectral characteristics of the input signal for different time and frequency domains. Since the encoder naturally has access to the original signal as well as to information about how the SBR module in the decoder will create the high band, given a specific set of control parameters, the system will be able to cope with situations where the low band constitutes a strong harmonic series and the reconstructed high band mainly constitutes arbitrary signal components, as well as situations where strong tonal components are present in the original high band with no equivalents in the low band on which the high band region is based. Moreover, the SBR encoder works closely with the underlying core codec to estimate which frequency range should be covered by SBR at a given time. The SBR data is efficiently encoded prior to transmission by using entropy coding as well as channel dependencies of the control data, in the case of stereo signals.

Control parameter extraction algorithms usually need to be carefully tuned to the underlying codec at a given bit rate and a given sampling frequency. This is due to the fact that a lower bit rate usually implies a larger SBR range than a higher bit rate, and different sampling frequencies correspond to different SBR frame temporal resolutions.

An SBR decoder typically includes several different parts. It comprises a bitstream decoding module, a high-frequency reconstruction (HFR) module, an additional high-frequency components module, and an envelope equalizer module. The system is based on a complex-valued QMF filter bank (for high-quality SBR) or a real-valued QMF filter bank (for low-power SBR). Embodiments of the invention are applicable to both high-quality and low-power SBR. In the bitstream extraction module, control data is read from the bitstream and decoded. A time-frequency grid is obtained for the current frame, before reading the envelope data from the bitstream. The main core decoder decodes the audio signal of the current frame (albeit at a low sampling rate) to form time-domain audio samples. The resulting audio data frame is used for high-frequency reconstruction by the HFR module. The decoded signal in the low-pass band is then analyzed using the QMF filter bank. High-frequency reconstruction and envelope correction are then performed on the sub-band samples of the QMF filter bank. High frequencies are reconstructed from the low band in a flexible manner, based on the given control parameters. Furthermore, the reconstructed high band is adaptively filtered on a sub-band channel basis according to the control data, in order to provide suitable spectral characteristics for the given time/frequency domain.

The top layer of the MPEG-4 AAC bitstream is a sequence of data blocks (raw_data_block elements), each of which is a data segment (referred to here as a block) that contains audio data (typically for a time period of 1024 or 960 samples) and associated information and/or other data. Here, we use the term block to refer to a segment of the MPEG-4 AAC bitstream containing audio data (and associated metadata and optionally also other associated data) that specifies or points to one (but not more than one) raw_data_block element.

Each MPEG-4 AAC bitstream block may include a number of syntax elements (each of which is also materialized in the bitstream as a data segment). Seven types of such syntax elements are defined in the MPEG-4 AAC standard. Each syntax element is identified by a different value of the “id_syn_ele” data element. Examples of syntax elements include “single_channel_element()”, “channel_pair_element()”, and “fill_element()”. A single channel element is a container that includes audio data from a single audio channel (a monophonic audio signal). A channel pair element includes audio data from two audio channels (i.e., a stereophonic audio signal).

A padding element is a container of information that includes an identifier (such as the value of the “id_syn_ele” element noted above) followed by data, which is referred to as “padding data.” Padding elements have historically been used to adjust the instantaneous bit rate of bit streams that are to be transmitted over a constant-rate channel. By adding an appropriate amount of padding data to each block, a constant data rate can be achieved.

According to embodiments of the invention, the fill data may include one or more extension payloads that extend the type of data (e.g., metadata) that can be transmitted in the bitstream. A decoder that receives bitstreams with fill data containing a new type of data may optionally be used by a device receiving the bitstream (e.g., a decoder) to extend the functionality of the device. Thus, as may be understood by a person skilled in the art, fill elements represent a special type of data structure and differ from data structures typically used for transmitting audio data (e.g., audio payloads containing channel data).

In some embodiments of the invention, the identifier used to identify the content element may consist of a three-bit unsigned integer with the most significant bit transmitted first (“uimsbf”), having a value of 0x6. Multiple instances of the same type of syntax element (e.g., multiple content elements) may appear in one block.

Another standard for coding audio bitstreams is the MPEG Unified Speech and Audio Coding (USAC) standard (ISO/IEC 23003-3:2012). The MPEG USAC standard describes the encoding and decoding of audio content using spectral range replication processing (including SBR processing as described in the MPEG-4 AAC standard, but also including other extended forms of spectral range replication processing). This processing applies the spectral range replication tools (sometimes referred to here as the “extended SBR tools” or “eSBR tools”) to an extended and improved version of the SBR toolset described in the MPEG-4 AAC standard. Thus, eSBR (as defined in the USAC standard) is an enhancement to SBR (as defined in the MPEG-4 AAC standard).

Here, we use the expression “extended SBR processing” (or “eSBR processing”) to refer to spectral range replication processing using at least one eSBR tool (e.g., at least one eSBR tool that is described or referenced in the MPEG USAC standard) that is not described or referenced in the MPEG-4 AAC standard. Examples of such eSBR tools are harmonic transposition and additional QMF patch insertion preprocessing or “pre-dithering”.

The harmonic transposer of integer order T maps a sinusoid of frequency ω to a sinusoid of frequency Tω, while preserving the signal duration. Three orders, T=2, 3, 4, are typically used in series to generate each portion of the desired output frequency range using the smallest possible transposition order. If output above the fourth-order transposition range is required, it can be generated by frequency shifts. Whenever possible, closely sampled baseband time domains are created for processing to minimize computational complexity.

The harmonic transposer can be either QMF-based or DFT-based. When using a QMF-based harmonic transposer, the bandwidth expansion of the time domain signal of the core coder is performed entirely in the QMF domain, using a modified phase vocoder structure, performing decimation followed by time stretching for each QMF subband. Transposition using multiple transposition factors (e.g. T=2, 3, 4) is performed in the normal QMF analysis/synthesis transform stage. Since the QMF-based harmonic transposer does not exhibit signal-adaptive frequency domain oversampling, the corresponding flag in the bitstream (sbrOversamplingFlag[ch]) can be ignored.

When using a DFT-based harmonic transposer, the coefficient transposers of 3 and 4 (3rd and 4th order transposers) are preferably integrated into the coefficient transposer of 2 (2nd order transposer) via interpolation to reduce complexity. For each frame (corresponding to coreCoderFrameLength core encoder samples), the nominal size of the “full-size” transposer transform is first determined by the signal-adaptive frequency domain oversampling flag (sbrOversamplingFlag[ch]) in the bitstream.

When sbrPatchingMode==1, indicating that linear transposition should be used to generate the high band, an additional stage may be introduced to avoid discontinuities in the shape of the spectral envelope of the high frequency signal input to the subsequent envelope equalizer. This improves the operation of the subsequent envelope equalization stage, resulting in a high band signal that is perceived as more stable. The additional preprocessing operation is advantageous for signal types where the coarse spectral envelope of the low band signal used to reconstruct the high frequencies exhibits large variations in level. However, the meaning of the bitstream element may be determined in the encoder by applying any kind of signal-dependent classification. The additional preprocessing is preferably activated via the one-bit bitstream element, bs_sbr_preprocessing. When bs_sbr_preprocessing is set to one, the additional preprocessing is enabled. When bs_sbr_preprocessing is set to zero, the additional preprocessing is disabled. Additional processing preferably uses a preGain curve that is used by the high-frequency generator to scale the low range, XLow, for each patch insertion. For example, the preGain curve might be calculated according to:

where k ₀ is the first QMF sub-band in the frequency band summary table, and lowEnvSlope is calculated using a function that calculates the coefficients of a best-fit polynomial (in the least-squares sense), such as polyfit(). For example,

can be applied (using a third degree polynomial), and where

where x_lowband(k)=[0…k ₀ -1], numTimeSlot is the number of SBR envelope time segments that exist in the frame, RATE is a constant indicating the number of QMF subband samples per time segment (e.g. 2), ϕ _k is the coefficient of the linear prediction filter (potentially derived from the covariance method), and where

A bitstream generated in accordance with the MPEG USAC standard (sometimes referred to herein as a “USAC bitstream”) includes encoded audio content and typically includes metadata indicating each type of spectral range replication processing to be applied by a decoder to decode the audio content of the USAC bitstream and/or metadata that controls such spectral range replication processing and/or indicates at least one characteristic or parameter of at least one SBR tool and/or eSBR tool to be used to decode the audio content of the USAC bitstream.

Here, we use the expression “extended SBR metadata” (or “eSBR metadata”) to denote metadata that indicates each type of spectral range replication processing to be applied by a decoder to decode the audio content of an encoded audio bitstream (e.g., a USAC bitstream), and/or that controls such spectral range replication processing and/or indicates at least one characteristic or parameter of at least one SBR tool and/or eSBR tool to be used to decode such audio content, but that is not described or mentioned in the MPEG-4 AAC standard. An example of eSBR metadata is metadata (intended to indicate or control spectral range replication processing) that is described or mentioned in the MPEG USAC standard, but not in the MPEG-4 AAC standard. Thus, eSBR metadata here refers to metadata that is not SBR metadata, and SBR metadata here refers to metadata that is not eSBR metadata.

The USAC bitstream may include both SBR metadata and eSBR metadata. More specifically, the USAC bitstream may include eSBR metadata that controls the execution of eSBR processing by the decoder, and SBR metadata that controls the execution of SBR processing by the decoder. According to exemplary embodiments of the present invention, the eSBR metadata (e.g., eSBR-specific configuration data) is included (in accordance with the present invention) in the MPEG-4 AAC bitstream (e.g., in the sbr_extension() container at the SBR payload end).

Performing eSBR processing, during decoding of the encoded bitstream using the eSBR toolkit (containing at least one eSBR tool), the decoder regenerates the high-frequency range of the audio signal based on the replication of the harmonic sequences that were cut off during encoding. Such eSBR processing typically corrects the spectral envelope of the generated high-frequency range and applies inverse filtering, and adds noise and sinusoidal components to recreate the spectral characteristics of the original audio signal.

According to exemplary embodiments of the invention, the eSBR metadata is included (e.g., a small number of control bits that represent the eSBR metadata are included) in one or more metadata segments of an encoded audio bitstream (e.g., an MPEG-4 AAC bitstream), which also includes encoded audio data in other segments (audio data segments). Typically, at least one such metadata segment of each bitstream block is (or includes) a fill element (including an identifier indicating the beginning of the fill element), and the eSBR metadata is included in the fill element after the identifier.

Fig. 1 is a block diagram of an exemplary audio processing chain (audio data processing system), in which one or more elements of the system may be configured in accordance with an embodiment of the present invention. The system includes the following elements, interconnected as shown: encoder 1, delivery subsystem 2, decoder 3 and post-processing module 4. In various embodiments of the shown system, one or more elements are omitted, or additional audio data processing modules are included.

In some implementations, the encoder 1 (which optionally includes a pre-processing module) is configured to receive PCM (time domain) samples containing audio content as input and to output an encoded audio bitstream (having a format that is compatible with the MPEG-4 AAC standard) that indicates the audio content. The bitstream data that indicates the audio content is sometimes referred to here as "audio data" or "encoded audio data". If the encoder is configured in accordance with an exemplary embodiment of the present invention, the audio bitstream output from the encoder includes eSBR metadata (and typically also other metadata) as well as audio data.

One or more coded audio bitstreams output from the encoder 1 may be provided to the coded audio delivery subsystem 2. The subsystem 2 is configured to store and/or deliver each coded audio bitstream output from the encoder 1. The coded audio bitstream output from the encoder 1 may be stored by the subsystem 2 (for example, in the form of a DVD or Blu-ray disc) or transmitted by the subsystem 2 (which may implement a communication line or a transmission network) or may be both stored and transmitted by the subsystem 2.

The decoder 3 is configured to decode the encoded MPEG-4 AAC audio bitstream (generated by the encoder 1), which it receives via the subsystem 2. In some embodiments, the decoder 3 is configured to extract eSBR metadata from each block of the bitstream and decode the bitstream (including by performing eSBR processing using the extracted eSBR metadata) to generate decoded audio data (e.g., streams of samples of decoded PCM audio). In some embodiments, the decoder 3 is configured to extract SBR metadata from the bitstream (but to ignore the eSBR metadata included in the bitstream) and decode the bitstream (including by performing SBR processing using the extracted SBR metadata) to generate decoded audio data (e.g., streams of samples of decoded PCM audio). Typically, decoder 3 includes a buffer that stores (e.g., in a non-temporal manner) segments of the encoded audio bitstream received from subsystem 2.

The post-processing module 4 according to Fig. 1 is configured to receive a stream of decoded audio data from the decoder 3 (e.g. samples of decoded PCM audio) and to perform post-processing on them. The post-processing module may also be configured to reproduce the post-processed audio content (or decoded audio received from the decoder 3) for reproduction by one or more speakers.

Fig. 2 is a block diagram of an encoder (100), which is an embodiment of an audio processing module according to the invention. Any of the components or elements of the encoder 100 may be implemented as one or more processes and/or one or more circuits (e.g., an ASIC, FPGA or other integrated circuits) in hardware, software or a combination of hardware and software. The encoder 100 includes an encoder 105, a filler/format stage 107, a metadata generation stage 106 and a buffer memory 109, connected as shown. Typically, the encoder 100 also includes other processing elements (not shown). The encoder 100 is configured to transform an input audio bitstream into an encoded output MPEG-4 AAC bitstream.

The metadata generator 106 is connected to and configured to generate (and/or pass to the cascade 107) metadata (including eSBR metadata and SBR metadata) to be included by the cascade 107 in an encoded bitstream to be output from the encoder 100.

The encoder 105 is connected and configured to encode (e.g. by performing compression thereon) the input audio data and to provide the resulting encoded audio to the stage 107 for inclusion in an encoded bitstream to be output from the stage 107.

The cascade 107 is configured to multiplex the encoded audio from the encoder 105 and the metadata (including the eSBR metadata and the SBR metadata) from the generator 106 to generate an encoded bitstream to be output from the cascade 107, preferably such that the encoded bitstream has a format as specified by one embodiment of the present invention.

The buffer memory 109 is configured to store (e.g., in a non-temporal manner) at least one block of the encoded audio bitstream output from the cascade 107, and a sequence of blocks of the encoded audio bitstream is then provided from the buffer memory 109 as an output from the encoder 100 to the delivery system.

Fig. 3 is a block diagram of a system including a decoder (200), which is an embodiment of the inventive audio processing module, and optionally also a post-processor (300) associated therewith. Any of the components or elements of the decoder 200 and the post-processor 300 may be implemented as one or more processes and/or one or more circuits (e.g., an ASIC, FPGA or other integrated circuits) in hardware, software or a combination of hardware and software. The decoder 200 comprises a buffer memory 201, a bitstream payload formatter (parser) 205, an audio decoding subsystem 202 (sometimes referred to as a “core” decoding stage or a “core” decoding subsystem), an eSBR processing stage 203 and a control bit generation stage 204, all associated as shown. Typically, the decoder 200 also includes other processing elements (not shown).

The buffer memory (buffer) 201 stores (e.g., in a non-temporal manner) at least one block of the encoded MPEG-4 AAC audio bitstream received by the decoder 200. During operation of the decoder 200, a sequence of bitstream blocks is added from the buffer 201 to the formatting unit 205.

In variations of the embodiment according to Fig. 3 (or embodiments according to Fig. 4, which will be described), an APU that is not a decoder (e.g., APU 500 in Fig. 6) includes a buffer memory (e.g., a buffer memory identical to buffer 201) that stores (e.g., in a non-temporal manner) at least one block of an encoded audio bitstream (e.g., an MPEG-4 AAC audio bitstream) of the same type received by buffer 201 according to Fig. 3 or Fig. 4 (i.e., an encoded audio bitstream that includes eSBR metadata).

Referring again to Fig. 3, the formatting unit 205 is connected and configured to demultiplex each block of the bitstream to extract therefrom SBR metadata (including quantized envelope data) and eSBR metadata (and typically also other metadata) to provide at least the eSBR metadata and the SBR metadata to the eSBR processing stage 203 and typically also provide other extracted metadata to the decoding subsystem 202 (and optionally also control the bit generator 204). The formatting unit 205 is also connected and configured to extract audio data from each block of the bitstream and to provide the extracted audio data to the decoding subsystem 202 (decoding stage).

The system according to Fig. 3 optionally also includes a post-processor 300. The post-processor 300 includes a buffer memory (buffer) 301 and other processing elements (not shown), including at least one processing element associated with the buffer 301. The buffer 301 stores (e.g., in a non-temporal manner) at least one block (or frame) of decoded audio data received by the post-processor 300 from the decoder 200. The processing elements of the post-processor 300 are connected and configured to receive and adaptively process a sequence of blocks (or frames) of decoded audio output from the buffer 301 using metadata output from the decoding subsystem 202 (and/or the deformatting unit 205) and/or to manage bits output from the stage 204 of the decoder 200.

The audio decoding subsystem 202 of the decoder 200 is configured to decode the audio data extracted by the parser 205 (such decoding may be referred to as a “core” decoding operation) to generate decoded audio data and provide the decoded audio data to the eSBR processing stage 203. The decoding is performed in the frequency domain and typically includes inverse quantization followed by spectral processing. Typically, the final processing stage in the subsystem 202 applies a transform from the frequency domain to the time domain to the decoded frequency domain audio data, so that the output from the subsystem is decoded time domain audio data. The cascade 203 is configured to apply the SBR tools and the eSBR tools indicated by the eSBR and eSBR metadata (extracted by the parser 205) to the decoded audio data (i.e., to perform SBR and eSBR processing on the output of the decoding subsystem 202 using the SBR and eSBR metadata) to generate fully decoded audio data that is output (e.g., to the post-processor 300) from the decoder 200. Typically, the decoder 200 includes a memory (accessed by the subsystem 202 and the cascade 203) that stores the deformatted audio data and the metadata output from the formatting unit 205, and the cascade 203 is configured to access the audio data and the metadata (including the SBR metadata and the eSBR metadata) as needed during the SBR and eSBR processing. The SBR processing and eSBR processing in cascade 203 can be considered as post-processing on the output of the core decoding subsystem 202. Optionally, the decoder 200 also includes a final upmix subsystem (which may apply parametric stereo (“PS”) tools defined in the MPEG-4 AAC standard, using PS metadata extracted by the asformatting unit 205 and/or control bits generated in the subsystem 204), which is connected and configured to perform an upmix on the output of the stage 203 to generate fully decoded, upmixed audio, which is output from the decoder 200. Alternatively, the post-processor 300 is configured to perform an upmix on the output of the decoder 200 (e.g. using the PS metadata extracted by the asformatting unit 205 and/or control bits generated in the subsystem 204).

In response to the metadata extracted by the formatting unit 205, the control bit generator 204 can generate control data, and the control data can be used in the decoder 200 (for example, in the final upmix subsystem) and/or provided as an output of the decoder 200 (for example, to the post-processor 300 for use in post-processing). In response to the metadata extracted from the input bitstream (and optionally also in response to the control data), the stage 204 can generate (and provide to the post-processor 300) control bits indicating that the decoded audio data output from the eSBR processing stage 203 should undergo a specific type of post-processing. In some implementations, decoder 200 is configured to provide metadata extracted by formatting unit 205 from the input bitstream to post-processor 300, and post-processor 300 is configured to perform post-processing on decoded audio data output from decoder 200 using the metadata.

Fig. 4 is a block diagram of an audio processing unit (210) ("APU"), which is another embodiment of the audio processing unit according to the invention. The APU 210 is a legacy decoder that is not configured to perform eSBR processing. Any of the components or elements of the APU 210 may be implemented as one or more processes and/or one or more circuits (e.g., an ASIC, FPGA, or other integrated circuits) in hardware, software, or a combination of hardware and software. The APU 210 includes a buffer memory 201, a bitstream payload deformatter (parser) 215, an audio decoding subsystem 202 (sometimes referred to as a "core" decoding stage or a "core" decoding subsystem), and an SBR processing stage 213, connected as shown. Typically, the APU 210 also includes other processing elements (not shown). The APU 210 may represent, for example, an audio encoder, decoder, or transcoder.

Elements 201 and 202 of the APU 210 are identical to the identically numbered elements of the decoder 200 (Fig. 3), and their description given above will not be repeated. In operation of the APU 210, a sequence of blocks of an encoded audio bitstream (MPEG-4 AAC bitstream) received by the APU 210 is provided from the buffer 201 to the deformatting block 215.

The deformatting unit 215 is connected and configured to demultiplex each block of the bitstream to extract the SBR metadata (including the quantized envelope data) and typically also other metadata from it, but to ignore the eSBR metadata that may be included in the bitstream in accordance with any embodiment of the present invention. The deformatting unit 215 is configured to provide at least the SBR metadata to the SBR processing stage 213. The deformatting unit 215 is also connected and configured to extract audio data from each block of the bitstream and to provide the extracted audio data to the decoding subsystem 202 (decoding stage).

The decoding subsystem 202 of the audio decoder 200 is configured to decode the audio data extracted by the deformatting unit 215 (such decoding may be referred to as a "core" decoding operation) to generate decoded audio data and provide the decoded audio data to the SBR processing stage 213. The decoding is performed in the frequency domain. Typically, the final processing stage in the subsystem 202 applies a transform from the frequency domain to the time domain to the decoded frequency domain audio data, so that the output of the subsystem is decoded time domain audio data. The cascade 213 is configured to apply the SBR tools (but not the eSBR tools) indicated by the SBR metadata (extracted by the deformatting unit 215) to the decoded audio data (i.e., perform SBR processing on the output of the decoding subsystem 202 using the SBR metadata) to generate fully decoded audio data that is output (e.g., to the post-processor 300) from the APU 210. Typically, the APU 210 includes a memory (accessed by the subsystem 202 and the cascade 213) that stores the deformatted audio data and metadata output from the deformatting device 215, and the cascade 213 is configured to access the audio data and metadata (including the SBR metadata) as needed during the SBR processing. The SBR processing in the cascade 213 can be considered as post-processing on the output of the core decoding subsystem 202. Optionally, the APU 210 also includes a final upmix subsystem (which may employ parametric stereo ("PS") tools defined in the MPEG-4 AAC standard using PS metadata extracted by the deformatting unit 215), which is coupled and configured to perform an upmix on the output of the stage 213 to generate fully decoded, upmixed audio that is output from the APU 210. Alternatively, the post processor is configured to perform an upmix on the output of the APU 210 (e.g. using PS metadata extracted by the deformatting unit 215 and/or control bits generated in the APU 210).

Different implementations of the encoder 100, decoder 200 and APU 210 are configured to perform different embodiments of the inventive method.

According to some embodiments, eSBR metadata is included (e.g., a small number of control bits that represent eSBR metadata are included) in an encoded audio bitstream (e.g., an MPEG-4 AAC bitstream), so that legacy decoders (that are not configured to parse the eSBR metadata or to use any eSBR tool to which the eSBR metadata relates) can ignore the eSBR metadata, but still decode the bitstream to the extent possible without using the eSBR metadata or any eSBR tool to which the eSBR metadata relates, typically without any significant degradation in the quality of the decoded audio. However, eSBR decoders that are configured to parse the bitstream to identify the eSBR metadata and to use at least one eSBR tool in response to the eSBR metadata will benefit from using at least one such eSBR tool. Therefore, embodiments of the invention provide a means for efficiently transmitting enhanced spectral range replication (eSBR) control data or metadata in a backward compatible manner.

Typically, eSBR metadata in a bitstream specifies (e.g. specifies at least one characteristic or parameter of) one or more of the following eSBR tools (which are described in the MPEG USAC standard and which may or may not be applied by the encoder during bitstream generation):

- harmonic transposition; and

- additional pre-processing of the QMF patch insert (pre-smoothing).

For example, eSBR metadata included in the bitstream may specify the values of the parameters (described in the MPEG USAC standard and in this disclosure): sbrPatchingMode[ch], sbrOversamplingFlag[ch], sbrPitchInBins[ch], sbrPitchInBins[ch], and bs_sbr_preprocessing.

Here, the notation X[ch], where X represents some parameter, means that the parameter refers to the channel (“ch”) of the audio content of the encoded bitstream to be decoded. For simplicity, we sometimes omit the expression [ch] and assume that the relevant parameter refers to the channel of the audio content.

Here, the notation X[ch][env], where X represents some parameter, means that the parameter refers to the SBR envelope (“env”) of the channel (“ch”) of the audio content of the encoded bitstream to be decoded. For simplicity, we sometimes omit the expressions [env] and [ch] and assume that the relevant parameter refers to the SBR envelope of the audio content channel.

During decoding of an encoded bitstream, the execution of harmonic transposition at the eSBR decoding processing stage (for each channel, “ch”, of the audio content specified by the bitstream) is controlled by the following eSBR metadata parameters: sbrPatchingMode[ch]: sbrOversamplingFlag[ch]; sbrPitchInBinsFlag[ch]; and sbrPitchInBins[ch].

The “sbrPatchingMode[ch]” value specifies the type of transposer used in eSBR: sbrPatchingMode[ch]=1 specifies patch insertion using linear transposition, as described in Section 4.6.18 of the MPEG-4 AAC standard (as used with High Quality SBR or Low Power SBR); sbrPatchingMode[ch]=0 specifies patch insertion using harmonic SBR, as described in Section 7.5.3 or 7.5.4 of the MPEG USAC standard.

The “sbrOversamplingFlag[ch]” value specifies the use of signal-adaptive frequency domain oversampling in eSBR in combination with DFT-based harmonic SBR patch insertion, as described in Section 7.5.3 of the MPEG USAC standard. This flag controls the size of the DFTs that are used in the transposer: 1 specifies that signal-adaptive frequency domain oversampling is enabled, as described in Section 7.5.3.1 of the MPEG USAC standard; 0 specifies that signal-adaptive frequency domain oversampling is disabled, as described in Section 7.5.3.1 of the MPEG USAC standard.

The value of “sbrPitchInBinsFlag[ch]” controls the interpretation of the sbrPitchInBins[ch] parameter: 1 indicates that the value in sbrPitchInBins[ch] is valid and greater than zero; 0 indicates that the value of sbrPitchInBins[ch] is set to zero.

The value of “sbrPitchInBins[ch]” controls the addition of cross product terms to the SBR harmonic transposer. The value of sbrPitchinBins[ch] is an integer value in the range [0,127] and represents the distance measured in frequency bins for the 1536-linear DFT operating at the sampling frequency of the base encoder.

In case the MPEG-4 AAC bitstream specifies an SBR channel pair whose channels are not related (rather than a single SBR channel), the bitstream specifies two examples of the syntax above (for harmonic or non-harmonic transposition), one for each sbr_channel_pair_element() channel.

Harmonic transposition of the eSBR tool typically improves the quality of decoded music signals at relatively low crossover frequencies. Non-harmonic transposition (i.e., legacy spectral patch insertion) typically improves speech signals. Thus, the starting point for deciding which transposition type is preferable for encoding a particular audio content is the choice of transposition method depending on speech/music detection, with harmonic transposition to be used on music content and spectral patch insertion on speech content.

The execution of pre-dithering during eSBR processing is controlled by the value of a one-bit eSBR metadata parameter known as “bs_sbr_preprocessing”, in that pre-dithering is either performed or not performed depending on the value of this one bit. When the QMF SBR patch insertion algorithm is used, as described in Section 4.6.18.6.3 of the MPEG-4 AAC standard, a pre-dithering stage may be performed (when specified by the “bs_sbr_preprocessing” parameter) in an attempt to avoid discontinuities in the shape of the spectral envelope of the high-frequency signal input to the subsequent envelope equalizer (the envelope equalizer performs another stage of eSBR processing). Pre-dithering typically improves the performance of the subsequent envelope equalizer stage, resulting in a high-band signal that is perceived as more stable.

It is expected that the overall bit rate requirement for inclusion in the eSBR metadata of the MPEG-4 AAC bitstream indicating the above-mentioned eSBR tools (harmonic transposition and pre-dithering) will be on the order of several hundred bits per second, since only the differential control data necessary to perform the eSBR processing is transmitted in accordance with some embodiments of the invention. Legacy decoders may ignore this information, since it is included in a backward-compatible manner (as will be explained below). Therefore, the adverse impact on bit rate associated with the inclusion of eSBR metadata is negligible, for a number of reasons, including the following:

- the bit rate penalty (due to the inclusion of eSBR metadata) is a very small fraction of the full bit rate, since only the differential control data required to perform eSBR processing is transmitted (and not the simultaneous transmission of SBR control data); and

- the setting of the SBR-related control information is generally independent of the transposition details. Examples of when the control data depends on the transposer operation are discussed later in this application.

Thus, embodiments of the invention provide a means for efficiently transmitting enhanced spectral range replication (eSBR) control data or metadata in a backwards compatible manner. This efficient transmission of eSBR control data reduces the memory requirements of decoders, encoders and transcoders employing aspects of the invention, while not having a noticeable negative impact on bit rate. Moreover, the complexity and processing requirements associated with performing eSBR in accordance with embodiments of the invention are also reduced, since the SBR data only needs to be processed once and not transmitted concurrently, which would be the case if eSBR were handled as a completely separate object type in MPEG-4 AAC instead of being integrated into the MPEG-4 AAC codec in a backwards compatible manner.

Next, with reference to Fig. 7, elements of a block (“raw_data_block”) of an MPEG-4 AAC bitstream in which eSBR metadata is included according to some embodiments of the present invention are described. Fig. 7 is a diagram of a block (“raw_data_block”) of an MPEG-4 AAC bitstream showing some of its segments.

An MPEG-4 AAC bitstream block may include at least one “single_channel_element()” (e.g., a single channel element shown in Fig. 7) and/or at least one “channel_pair_element()” (not shown specifically in Fig. 7, although may be present) including audio data for an audio program. The block may also include a number of “fill_elements” (e.g., fill element 1 and/or fill element 2 in Fig. 7) including data (e.g., metadata) related to the program. Each “single_channel_element()” includes an identifier (e.g., “ID1” in Fig. 7) indicating the start of a single channel element and may include audio data indicating another channel of a multi-channel audio program. Each “channel_pair_element” includes an identifier (not shown in Fig. 7) indicating the start of a channel pair element and may include audio data indicating two channels of the program.

The fill_element (referred to herein as the fill element) of the MPEG-4 AAC bitstream includes an identifier (“ID2” in Fig. 7) indicating the start of the fill element and fill data following the identifier. The ID2 identifier may consist of a first three-bit unsigned integer with the most significant bit transmitted first (“uimsbf”) having a value of 0x6. The fill data may include an extension_payload() element (sometimes referred to herein as the extension payload), the syntax of which is shown in Table 4.57 of the MPEG-4 AAC standard. There are several types of extension payloads, identified by the “extension_type” parameter, which is a four-bit unsigned integer with the most significant bit transmitted first (“uimsbf”).

The payload data (e.g., its extension payload) may include a header or identifier (e.g., “header1” in Fig. 7) that specifies a segment of the payload data that specifies an SBR object (i.e., the header initializes the “SBR object” type, called sbr_extension_data() in the MPEG-4 AAC standard). For example, a spectral range replication (SBR) extension payload is identified using the value ‘1101’ or ‘1110’ for the extension_type field in the header, where the identifier ‘1101’ identifies the extension payload with SBR data, and ‘1110’ identifies the extension payload with SBR data with a cyclic redundancy check (CRC) for verifying the correctness of the SBR data.

When a header (e.g., an extension_type field) initializes an SBR object type, SBR metadata (sometimes referred to herein as “spectral range replication data” and referred to as sbr_data() in the MPEG-4 AAC standard) follows the header, and at least one spectral range replication extension element (e.g., an “SBR extension element” of fill element 1 in Fig. 7) may follow the SBR metadata. Such a spectral range replication extension element (bitstream segment) is referred to as an “sbr_extension()” container in the MPEG-4 AAC standard. The spectral range replication extension element optionally includes a header (e.g., an “SBR extension header” of fill element 1 in Fig. 7).

The MPEG-4 AAC standard assumes that a spectral range replication extension element may include PS (parametric stereo) data for program audio data. The MPEG-4 AAC standard assumes that when a fill element header (e.g., its extension payload) initializes an SBR object type (as “header1” does in Fig. 7), and the spectral range replication extension element of the fill element includes PS data, the fill element (e.g., its extension payload) includes spectral range replication data, and the “bs_extension_id” parameter, the value of which (i.e., bs_extension_id=2) indicates that the PS data is included in the spectral range replication extension element of the fill element.

According to some embodiments of the present invention, eSBR metadata (e.g., a label indicating whether extended spectral range replication (eSBR) processing should be performed on the content of the audio block) is included in a spectral range replication extension element of a filling element. For example, such a label is indicated in filling element 1 in Fig. 7, where the label appears after a header (the "SBR extension header" of filling element 1) of the "SBR extension element" of filling element 1. Optionally, such a label and additional eSBR metadata are included in a spectral range replication extension element after the header of the spectral range replication extension element (e.g., in the SBR extension element of filling element 1 in Fig. 7, after the SBR extension header). According to some embodiments of the present invention, a content element that includes eSBR metadata also includes a “bs_extension_id” parameter, the value of which (e.g., bs_extension_id=3) indicates that eSBR metadata is included in the content element and that eSBR processing is to be performed on the audio content of the relevant block.

According to some embodiments of the invention, the eSBR metadata is included in a fill element (e.g., fill element 2 in Fig. 7) of the MPEG-4 AAC bitstream that is different from the spectral range replication extension element (SBR extension element) of the fill element. This is so because fill elements containing extension_payload() with SBR data or SBR data with CRC do not contain any other extension payload of any other extension type. Therefore, in embodiments where the eSBR metadata is stored with its own extension payload, a separate fill element is used to store the eSBR metadata. Such a fill element includes an identifier (e.g., “ID2” in Fig. 7) indicating the start of the fill element, and fill data following the identifier. The fill data may include an extension_payload() element (sometimes referred to herein as an extension payload), the syntax of which is shown in Table 4.57 of the MPEG-4 AAC standard. The fill data (e.g., its extension payload) includes a header (e.g., “header2” of fill element 2 in Fig. 7) that indicates an eSBR object (i.e., the header initializes the type of an extended spectral range replication (eSBR) object), and the fill data (e.g., its extension payload) includes eSBR metadata after the header. For example, fill element 2 in Fig. 7 includes such a header (“header2”) and also includes, after the header, eSBR metadata (i.e., a “tag” in fill element 2 that indicates whether extended spectral range replication (eSBR) processing should be performed on the audio content of the block). Optionally, additional eSBR metadata is also included in the fill data of fill element 2 in Fig. 7, after header2. In the embodiments described in the previous paragraph, the header (e.g., header2 in Fig. 7) has an identification value that is not one of the traditional values specified in Table 4.57 of the MPEG-4 AAC standard, and instead indicates an eSBR extension payload (so that the extension_type field of the header indicates that the payload data includes eSBR metadata).

In a first class of embodiments, the invention is an audio processing module (e.g., a decoder) comprising:

a memory (e.g., buffer 201 in Fig. 3 or 4) configured to store at least one block of an encoded audio bitstream (e.g., at least one block of an MPEG-4 AAC bitstream);

a bit stream payload deformatting unit (for example, element 205 in Fig. 3 or element 215 in Fig. 4) associated with the memory and configured to demultiplex at least one portion of said bit stream block; and

a decoding subsystem (for example, elements 202 and 203 in Fig. 3 or elements 202 and 213 in Fig. 4) connected and configured to decode at least one portion of the audio content of said bitstream block, wherein the block includes:

a fill element comprising an identifier indicating the start of the fill element (e.g. the identifier “id_syn_ele” having the value 0×6, Table 4.85 of the MPEG-4 AAC standard), and fill data following the identifier, wherein the fill data comprises:

at least one flag identifying whether enhanced spectral range replication (eSBR) processing is to be performed on the audio content of the block (e.g. using spectral range replication data and eSBR metadata included in the block).

The label represents eSBR metadata, and an example of a label is the sbrPatchingMode label. Another example of a label is the harmonicSBR label. Both of these labels indicate whether the basic form of spectral replication or the extended form of spectral replication should be performed on the audio block data. The basic form of spectral replication is spectral patch insertion, and the extended form of spectral replication is harmonic transposition.

In some embodiments, the content data also includes additional eSBR metadata (i.e., eSBR metadata other than the label).

The memory may be a buffer memory (e.g., an implementation of buffer 201 in Fig. 4) that stores (e.g., in a non-temporal manner) at least one block of the encoded audio bitstream.

The complexity of performing eSBR processing (using eSBR harmonic transposition and pre-dithering) by an eSBR decoder during decoding of an MPEG-4 AAC bitstream that includes eSBR metadata (indicating these eSBR tools) is estimated to be as follows (for a typical decoding with the specified parameters):

Harmonic transposition (16 kbps, 14400/28800 Hz)

- DFT-based: 3.68 WMOPS (weighted million operations per second);

- based on QMF: 0.98 WMOPS;

QMF patch insertion pre-processing (pre-smoothing): 0.1WMOPS.

It is known that DFT-based transposition generally performs better than QMF-based transposition for transients.

According to some embodiments of the present invention, a fill element (of an encoded audio bitstream) that includes eSBR metadata also includes a parameter (e.g., the “bs_extension_id” parameter) whose value (e.g., bs_extension_id=3) signals that eSBR metadata is included in the fill element and that eSBR processing should be performed on the audio content of the relevant block, and/or a parameter (e.g., the same “bs_extension_id” parameter) whose value (e.g., bs_extension_id=2) signals that the sbr_extension() container of the fill element includes PS data. For example, as shown in Table 1 below, such a parameter with a value of bs_extension_id=2 may signal that the sbr_extension() container of the content element includes PS data, and such a parameter with a value of bs_extension_id=3 may signal that the sbr_extension() container of the content element includes eSBR metadata:

Table 1

bs_extension_id Meaning 0 Reserved 1 Reserved 2 EXTENSION_ID_PS 3 EXTENSION_ID_ESBR

According to some embodiments of the invention, the syntax of each spectral range replication extension element that includes eSBR metadata and/or PS data is as specified in Table 2 below (in which “sbr_extension()” denotes a container that represents a spectral range replication extension element, “bs_extension_id” is as described in Table 1 above, “ps_data” denotes PS data, and “esbr_data” denotes eSBR metadata):

Table 2

Note 1: ps_data() returns the number of bits read. Note 2: esbr_data() returns the number of bits read.

In the exemplary embodiment, esbr_data(), referenced in Table 2 above, specifies the values of the following metadata parameters:

1. A one-bit metadata parameter, “bs_sbr_preprocessing”; and

2. For each channel (“ch”) of the audio content of the encoded bitstream to be decoded, each of the above parameters: “sbrPatchingMode[ch]”; “sbrOversamplingFlag[ch]”; “sbrPitchInBinsFlag[ch]”; and “sbrPitchInBins[ch]”.

For example, in some embodiments, esbr_data() may have the syntax shown in Table 3 to specify these metadata parameters:

Table 3

Note: bs_sbr_preprocessing is defined as described in Section 6.2.12 of ISO/IEC 23003-3:2012. sbrPatchingMode[ch], sbrOversamplingFlag[ch], sbrPitchInBinsFlag[ch], and sbrPitchInBins[ch] are defined as described in Section 7.5 of ISO/IEC 23003-3:2012.

The syntax described above enables efficient implementation of an extended form of spectral range replication, such as harmonic transposition, as an extension of the legacy decoder. Specifically, the eSBR data of Table 3 includes only those parameters necessary to perform the extended form of spectral range replication that are either not already supported in the bitstream or are not directly derived from parameters already supported in the bitstream. All other parameters and processing data necessary to perform the extended form of spectral range replication are derived from pre-existing parameters at already defined locations in the bitstream.

For example, an MPEG-4 HE-AAC or HE-AAC v2 compliant decoder can be extended to include an extended form of spectral range replication, such as harmonic transposition. This extended form of spectral range replication complements the basic form of spectral range replication already supported by the decoder. In the context of an MPEG-4 HE-AAC or HE-AAC v2 compliant decoder, this basic form of spectral range replication is the SBR QMF spectral patch insertion tool, as defined in Section 4.6.18 of the MPEG-4 AAC standard.

When performing the extended form of spectral range replication, the extended HE-AAC decoder can reuse many bitstream parameters already included in the SBR bitstream extension payload. Specific parameters that can be reused include, for example, various parameters that define the frequency band summary table. These parameters include bs_start_freq (the parameter that defines the start of the frequency summary table parameter), bs_stop_freq (the parameter that defines the end of the frequency summary table), bs_freq_scale (the parameter that defines the number of frequency bands per octave), and bs_alter_scale (the parameter that changes the scale of the frequency bands). Parameters that can be reused also include the parameters that define the noise band table (bs_noise_bands) and the limiter band table parameters (bs_limiter_bands). Accordingly, in various embodiments, at least some of the equivalent parameters specified in the USAC standard are omitted from the bitstream, thereby reducing control overhead in the bitstream. Typically, if a parameter specified in the AAC standard has an equivalent parameter specified in the USAC standard, then the equivalent parameter specified in the USAC standard has the same value as the parameter specified in the AAC standard, such as an envelope scaling factor E _OrigMapped . However, the equivalent parameter specified in the USAC standard typically has a different value that is “tuned” for the extended SBR processing specified in the USAC standard, rather than for the SBR processing specified in the AAC standard.

In order to improve the subjective quality for audio content with harmonic frequency structure and strong tonal characteristics, in particular at low bit rates, the activation of extended SBR is recommended. The values of the corresponding bitstream element (i.e. esbr_data()) controlling these tools can be determined in the encoder by applying a signal-dependent classification mechanism. In general, the use of the harmonic patching method (sbrPatchingMode==1) is preferable for encoding music signals at very low bit rates, where the underlying codec may be significantly limited in audio bandwidth. This is especially true if these signals include a structure with pronounced harmonic oscillations. In contrast, the use of the regular SBR patching method is preferable for speech and mixed signals, since it provides better preservation of the temporal structure in speech.

To improve the performance of the harmonic transposer, a preprocessing step (bs_sbr_preprocessing==1) can be activated, which aims to avoid the occurrence of spectral discontinuities in the signal entering the subsequent envelope equalizer. The operation of this tool is beneficial for signal types where the coarse spectral envelope of the low-frequency signal used for high-frequency reconstruction exhibits large variations in level.

To improve the transient response of the harmonic SBR patch insertion, signal-adaptive frequency domain oversampling (sbrOversamplingFlag==1) can be applied. Since signal-adaptive frequency domain oversampling increases the computational complexity of the transposer but only benefits frames that contain transients, the use of this tool is controlled by a bitstream element that is transmitted once per frame and per independent SBR channel.

A decoder operating in the proposed extended SBR mode must typically be able to switch between legacy and extended SBR patch insertion. Therefore, a delay may be introduced that can be as long as one basic audio frame, depending on the decoder setup. Typically, the delay for both legacy and extended SBR patch insertion will be similar.

In addition to numerous parameters, other data elements may also be reused by the extended HE-AAC decoder when performing the extended form of spectral range replication according to embodiments of the invention. For example, envelope data and noise level data may also be extracted from the bs_data_env (envelope scaling factors) and bs_noise_env (noise level scaling factors) data and used during the extended form of spectral range replication.

In essence, these embodiments use the configuration parameters and envelope data already supported by a legacy HE-AAC or HE-AAC v2 decoder in the SBR extension payload to provide an extended form of spectral range replication that requires as little additional data as possible to be transmitted. The metadata was originally configured for the basic form of HFR (e.g., the SBR spectral transfer operation), but is used for the extended form of HFR (e.g., the eSBR harmonic transposition) in accordance with the embodiments. As discussed previously, the metadata generally represents operating parameters (e.g., envelope scaling factors, noise level scaling factors, time/frequency grid parameters, sinusoid addition information, variable crossover frequency/range, inverse filtering mode, envelope resolution, smoothing mode, frequency interpolation mode) that are configured and intended for use with the basic form of HFR (e.g., linear spectral transfer). However, this metadata, in combination with additional metadata parameters specific to the extended form of HFR (such as harmonic transposition), can be used to efficiently and effectively process audio data using the extended form of HFR.

Accordingly, extended decoders that support the extended form of spectral range replication can be created in a very efficient manner by relying on already defined elements of the bitstream (e.g., elements in the SBR extension payload) and by adding only those parameters that are necessary to support the extended form of spectral range replication (in the filler element extension payload). This data reduction feature, combined with placing the newly added parameters in a reserved data field, such as an extension container, significantly reduces the barrier to creating a decoder that supports the extended form of spectral range replication by ensuring that the bitstream is backward compatible with a legacy decoder that does not support the extended form of spectral range replication. It will be understood that the reserved data field is a backward compatible data field, i.e., it is a data field that is already supported by earlier decoders, such as legacy HE-AAC or HE-AAC v2 decoders. Similarly, an extension container is backward compatible, that is, it is an extension container that is already supported by earlier decoders, such as the legacy HE-AAC or HE-AAC v2 decoders.

In Table 3, the number in the right column indicates the number of bits of the corresponding parameter in the left column.

In some embodiments, the SBR object type defined in MPEG-4 AAC is updated to contain the SBR tool and aspects of the extended SBR (eSBR) tool, as signaled in an SBR extension element (bs_extension_id==EXTENSION_ID_ESBR). If the decoder detects this SBR extension element, the decoder uses the signaled aspects of the extended SBR tool.

In some embodiments, the invention is a method that includes the step of encoding audio data to generate an encoded bitstream (e.g., an MPEG-4 AAC bitstream) that includes eSBR metadata in at least one segment of at least one block of the encoded bitstream and audio data in at least one other segment of the block. In exemplary embodiments, the method includes the step of multiplexing audio data with the eSBR metadata in each block of the encoded bitstream. In exemplary decoding of the encoded bitstream in an eSBR decoder, the decoder extracts the eSBR metadata from the bitstream (including by parsing and demultiplexing the eSBR metadata and the audio data) and uses the eSBR metadata to process the audio data to generate a stream of decoded audio data.

Another aspect of the invention is an eSBR decoder configured to perform eSBR processing (e.g., using at least one of the eSBR tools known as harmonic transposition or pre-dithering) during decoding of an encoded audio bitstream (e.g., an MPEG-4 AAC bitstream) that does not include eSBR metadata. An example of such a decoder will be described with reference to Fig. 5.

The eSBR decoder (400) according to Fig. 5 includes a buffer memory 201 (which is identical to the memory 201 in Figs. 3 and 4), a bitstream payload deformatting unit 215 (which is identical to the deformatting unit 215 in Fig. 4), an audio decoding subsystem 202 (sometimes called a “core” decoding stage or a “core” decoding subsystem and which is identical to the core decoding subsystem 202 in Fig. 3), an eSBR control data generation subsystem 401 and an eSBR processing stage 203 (which is identical to the stage 203 in Fig. 3), connected as shown. Typically, the decoder 400 also includes other processing elements (not shown).

When the decoder 400 operates, a sequence of blocks of the encoded audio bitstream (MPEG-4 AAC bitstream) received by the decoder 400 is provided from the buffer 201 to the deformatting block 215.

The deformatting unit 215 is connected and configured to demultiplex each block of the bitstream to extract SBR metadata (including quantized envelope data) and usually also other metadata from it. The deformatting unit 215 is configured to provide at least the SBR metadata to the eSBR processing stage 203. The deformatting unit 215 is also connected and configured to extract audio data from each block of the bitstream and to provide the extracted audio data to the decoding subsystem (decoding stage) 202.

The decoding subsystem 202 of the audio decoder 400 is configured to decode the audio data extracted by the deformatting unit 215 (such decoding may be referred to as a "core" decoding operation) to generate decoded audio data and provide the decoded audio data to the eSBR processing stage 203. The decoding is performed in the frequency domain. Typically, the final processing stage in the subsystem 202 applies a transform from the frequency domain to the time domain to the decoded frequency domain audio data, so that the output of the subsystem is decoded time domain audio data. The cascade 203 is configured to apply SBR tools (and eSBR tools) specified by the SBR metadata (extracted by the deformatting unit 215) and the eSBR metadata generated in the subsystem 401 to the decoded audio data (i.e., to perform SBR and eSBR processing on the output of the decoding subsystem 202 using the SBR and eSBR metadata) in order to generate fully decoded audio data, which is the output of the decoder 400. Typically, the decoder 400 includes a memory (accessed by the subsystem 202 and the cascade 203), which stores the deformatted audio data and the metadata output from the deformatting unit 215 (and optionally also the subsystem 401), and the cascade 203 is configured to access the audio data and the metadata as needed during the SBR and eSBR processing. The SBR processing in the stage 203 can be considered as post-processing of the output of the core decoding subsystem 202. Optionally, the decoder 400 also includes a final upmix subsystem (which can apply parametric stereo (“PS”) tools defined in the MPEG-4 AAC standard using PS metadata extracted by the deformatting unit 215), which is connected and configured to upmix the output of the stage 203 to generate fully decoded, upmixed audio, which is the output of the APU 210.

Parametric stereo is a coding tool that represents a stereo signal using a linear downmix of the left and right channels of the stereo signal and sets of spatial parameters that describe the stereo image. Parametric stereo typically employs three types of spatial parameters: (1) interchannel intensity differences (IID), which describe the intensity differences between channels; (2) interchannel phase differences (IPD), which describe the phase differences between channels; and (3) interchannel coherence (ICC), which describes the coherence (or similarity) between channels. Coherence can be measured as the maximum cross-correlation as a function of time or phase. These three parameters generally provide the ability to reconstruct a stereo image in high quality. However, the IPD parameters specify only the relative phase differences between the channels of the input stereo signal and do not indicate the distribution of these phase differences across the left and right channels. Therefore, a fourth type of parameter may be additionally used, describing the overall phase shift or overall phase difference (OPD). In the stereo reconstruction process, successive window segments of the received downmix signal, s[n], and the decorrelated version of the received downmix, d[n], are processed together with the spatial parameters to generate left (l _k (n)) and right (r _k (n)) reconstructed signals according to:

l _k (n)=H ₁₁ (k, n)s _k (n)+H ₂₁ (k, n)d _k (n)

r _k (n)=H ₁₂ (k, n)s _k (n)+H ₂₂ (k, n)d _k (n)

where H ₁₁ , H ₁₂ , H ₂₁ and H ₂₂ are determined by the stereo parameters. The signals l _k (n) and r _k (n) are finally transformed back to the time domain by means of the time-frequency transform.

The control data generation subsystem 401 according to Fig. 5 is connected and configured to detect at least one property of the encoded audio bitstream to be decoded and to generate eSBR control data (which may be or include any type of eSBR metadata included in the encoded audio bitstreams according to other embodiments of the invention) in response to at least one result of the detection step. The eSBR control data is provided to the cascade 203 to trigger the application of individual eSBR tools or combinations of eSBR tools after detecting a specific property (or combinations of properties) of the bitstream and/or to control the application of such eSBR tools. For example, to control the execution of eSBR processing using harmonic transposition, some embodiments of the control data generation subsystem 401 will include: a music detector (e.g., a simplified version of a traditional music detector) for setting the sbrPatchingMode[ch] parameter (and providing the set parameter to the cascade 203) in response to detecting that the bitstream indicates music or does not indicate music; a transient detector for setting the sbrOversamplingFlag[ch] parameter (and providing the set parameter to the cascade 203) in response to detecting the presence or absence of transients in the audio content indicated by the bitstream; and/or a pitch detector for setting the sbrPitchInBinsFlag[ch] and sbrPitchInBins[ch] parameters (and providing the set parameters to the cascade 203) in response to detecting the pitch of the audio content indicated by the bitstream. Other aspects of the invention are methods for decoding an audio bitstream performed by any embodiment of the inventive decoder described in this paragraph and the preceding paragraph.

Aspects of the invention include a method for encoding or decoding of the type that any embodiment of the inventive APU, system, or device is configured (e.g., programmed) to perform. Other aspects of the invention include a system or device configured (e.g., programmed) to perform any embodiment of the inventive method, and a computer-readable medium (e.g., a disk) that stores code (e.g., in a non-transitory manner) for implementing any embodiment of the inventive method or its steps. For example, the inventive system may be or include a programmable general-purpose processor, a digital signal processor, or a microprocessor programmed with software or firmware and/or otherwise configured to perform any of a variety of operations on data, including an embodiment of the inventive method or its steps. Such a general-purpose processor may be or include a computer system including an input device, a memory, and a processing circuit programmed (and/or otherwise configured) to perform an embodiment of the inventive method (or its steps) in response to data provided thereto.

Embodiments of the present invention may be implemented in hardware, firmware, or software, or a combination of both (e.g., as a programmable logic array). Unless otherwise specified, algorithms or processes included as part of the invention do not necessarily relate to any particular computer or other device. In particular, various general-purpose machines may be used with programs written in accordance with the principles set forth herein, or it may be more convenient to create a more specialized device (e.g., integrated circuits) to perform the required method steps. Thus, the invention can be implemented in one or more computer programs running on one or more programmable computer systems (for example, an implementation of any of the elements according to Fig. 1 or the encoder 100 according to Fig. 2 (or an element thereof) or the decoder 200 according to Fig. 3 (or an element thereof) or the decoder 210 according to Fig. 4 (or an element thereof) or the decoder 400 according to Fig. 5 (or an element thereof)), each of which comprises at least one processor, at least one data storage system (including volatile or non-volatile memory and/or storage elements), at least one input device or port and at least one output device or port. The program code is applied to the input data to perform the functions described herein and to generate output information. The output information is supplied to one or more output devices in a known manner.

Each such program may be implemented in any desired computer language (including machine, assembly, or high-level procedural, logical, or object-oriented programming languages) for communicating with the computer system. In any case, the language may be a compiled or interpreted language.

For example, when implemented by sequences of computer software instructions, the various functions and steps of embodiments of the invention may be implemented by sequences of multi-threaded software instructions running in suitable digital signal processing hardware, in which case the various devices, steps and functions of the embodiments may correspond to portions of the software instructions.

Each such computer program is preferably stored on or loaded into a storage medium or device (e.g., solid-state memory or media, or magnetic or optical media) readable by a general-purpose or specialized programmable computer, for configuring and operating the computer when the medium or storage device is read by the computer system to perform the procedures described herein. The inventive system may also be implemented as a computer-readable storage medium configured with (i.e., storing) the computer program, where the storage medium, so configured, causes the computer system to operate in a specific and predetermined manner to perform the functions described herein.

A number of embodiments of the invention have been described. However, it will be understood that various modifications can be made without departing from the spirit and scope of the invention. Numerous modifications and variations of the present invention are possible in light of the above. For example, to facilitate efficient implementations, phase shifts can be used in combination with complex QMF analysis and synthesis filter banks. The analysis filter bank is responsible for filtering the time-domain low-pass band signal generated by the core decoder into a plurality of sub-bands (e.g., QMF sub-bands). The synthesis filter bank is responsible for combining the regenerated high band formed by the selected HFR method (as indicated by the received sbrPatchingMode parameter) with the decoded low band to form a wideband output audio signal. A given filter bank implementation operating in a particular sampling frequency mode, such as normal double-rate operation or SBR downsampling mode, should not, however, have phase shifts that depend on the bitstream. The QMF banks used in SBR are a complex-exponential extension of the cosine-modulated filter bank theory. It can be shown that the limitations of sampling noise compensation become obsolete when extending the cosine-modulated filter bank with complex-exponential modulation. Thus, for SBR QMF banks, both the analysis filters, h _k (n), and the synthesis filters, f _k (n), can be defined by:

(1)

where _p0 (n) is a real-valued symmetric or asymmetric prototype filter (usually a low-pass prototype filter), M denotes the number of channels, and N is the order of the prototype filter. The number of channels used in the analysis filter bank may be different from the number of channels used in the synthesis filter bank. For example, the analysis filter bank may have 32 channels, and the synthesis filter bank may have 64 channels. When using the synthesis filter bank in downsampling mode, the synthesis filter bank may have only 32 channels. Since the subband samples from the filter bank are complex-valued, an additional possible channel-dependent phase shift stage may be added to the analysis filter bank. These additional phase shifts must be compensated before the synthesis filter bank. While the phase shift terms can in principle have arbitrary values without disturbing the operation of the QMF analysis/synthesis chain, they may also be limited to certain values for consistency verification. The SBR signal will be affected by the choice of phase coefficients, while the low-pass signal coming from the base decoder will not be affected. The quality of the audio output signal is not affected.

The prototype filter coefficients, p ₀ (n), can be defined by the length, L, 640, as shown in Table 4 below.

Table 4

n p ₀ (n) n p ₀ (n) n p ₀ (n) 0 0,0000000000 214 0.0019765601 428 0.0117623832 1 -0.0005525286 215 -0.0032086896 429 0.0163701258 2 -0.0005617692 216 -0.0085711749 430 0.0207997072 3 -0.0004947518 217 -0.0141288827 431 0.0250307561 4 -0.0004875227 218 -0.0198834129 432 0.0290824006 5 -0.0004893791 219 -0.0258227288 433 0.0329583930 6 -0.0005040714 220 -0.0319531274 434 0.0366418116 7 -0.0005226564 221 -0.0382776572 435 0.0401458278 8 -0.0005466565 222 -0.0447806821 436 0.0434768782 9 -0.0005677802 223 -0.0514804176 437 0.0466303305 10 -0.0005870930 224 -0.0583705326 438 0.0495978676 11 -0.0006132747 225 -0.0654409853 439 0.0524093821 12 -0.0006312493 226 -0.0726943300 440 0.0550460034 13 -0.0006540333 227 -0.0801372934 441 0.0575152691 14 -0.0006777690 228 -0.0877547536 442 0.0598166570 15 -0.0006941614 229 -0.0955533352 443 0.0619602779 16 -0.0007157736 230 -0.1035329531 444 0.0639444805 17 -0.0007255043 231 -0.1116826931 445 0.0657690668 18 -0.0007440941 232 -0.1200077984 446 0.0674525021 19 -0.0007490598 233 -0.1285002850 447 0.0689664013 20 -0.0007681371 234 -0.1371551761 448 0.0703533073 21 -0.0007724848 235 -0.1459766491 449 0.0715826364 22 -0.0007834332 236 -0.1549607071 450 0.0726774642 23 -0.0007779869 237 -0.1640958855 451 0.0736406005 24 -0.0007803664 238 -0.1733808172 452 0.0744664394 25 -0.0007801449 239 -0.1828172548 453 0.0751576255 26 -0.0007757977 240 -0.1923966745 454 0.0757305756 27 -0.0007630793 241 -0,2021250176 455 0.0761748321 28 -0.0007530001 242 -0.2119735853 456 0.0765050718 29 -0.0007319357 243 -0.2219652696 457 0.0767204924 30 -0.0007215391 244 -0.2320690870 458 0.0768230011 31 -0.0006917937 245 -0.2423016884 459 0.0768173975 32 -0.0006650415 246 -0.2526480309 460 0.0767093490 33 -0.0006341594 247 -0.2631053299 461 0.0764992170 34 -0.0005946118 248 -0.2736634040 462 0.0761992479 35 -0.0005564576 249 -0.2843214189 463 0.0758008358 36 -0.0005145572 250 -0.2950716717 464 0.0753137336 37 -0.0004606325 251 -0.3059098575 465 0.0747452558 38 -0.0004095121 252 -0.3168278913 466 0.0741003642 39 -0.0003501175 253 -0.3278113727 467 0.0733620255 40 -0.0002896981 254 -0.3388722693 468 0.0725682583 41 -0.0002098337 255 -0.3499914122 469 0.0717002673 42 -0.0001446380 256 0.3611589903 470 0.0707628710 43 -0.0000617334 257 0.3723795546 471 0.0697630244 44 0.0000134949 258 0.3836350013 472 0.0687043828 45 0.0001094383 259 0.3949211761 473 0.0676075985 46 0.0002043017 260 0.4062317676 474 0.0664367512 47 0.0002949531 261 0.4175696896 475 0.0652247106 48 0.0004026540 262 0.4289119920 476 0.0639715898 49 0.0005107388 263 0.4402553754 477 0.0626857808 50 0.0006239376 264 0.4515996535 478 0.0613455171 51 0.0007458025 265 0.4629308085 479 0.0599837480 52 0.0008608443 266 0.4742453214 480 0.0585915683 53 0.0009885988 267 0.4855253091 481 0.0571616450 54 0.0011250155 268 0.4967708254 482 0.0557173648 55 0.0012577884 269 0.5079817500 483 0.0542452768 56 0.0013902494 270 0.5191234970 484 0.0527630746 57 0.0015443219 271 0.5302240895 485 0.0512556155 58 0.0016868083 272 0.5412553448 486 0.0497385755 59 0.0018348265 273 0.5522051258 487 0.0482165720 60 0.0019841140 274 0.5630789140 488 0.0466843027 61 0.0021461583 275 0.5738524131 489 0.0451488405 62 0.0023017254 276 0.5845403235 490 0.0436097542 63 0.0024625616 277 0.5951123086 491 0.0420649094 64 0.0026201758 278 0.6055783538 492 0.0405349170 65 0.0027870464 279 0.6159109932 493 0.0390053679 66 0.0029469447 280 0.6261242695 494 0.0374812850 67 0.0031125420 281 0.6361980107 495 0.0359697560 68 0.0032739613 282 0.6461269695 496 0.0344620948 69 0.0034418874 283 0.6559016302 497 0.0329754081 70 0.0036008268 284 0.6655139880 498 0.0315017608 71 0.0037603922 285 0.6749663190 499 0.0300502657 72 0.0039207432 286 0.6842353293 500 0.0286072173 73 0.0040819753 287 0.6933282376 501 0.0271859429 74 0.0042264269 288 0.7022388719 502 0.0257875847 75 0.0043730719 289 0.7109410426 503 0.0244160992 76 0.0045209852 290 0.7194462634 504 0.0230680169 77 0.0046606460 291 0.7277448900 505 0.0217467550 78 0.0047932560 292 0.7358211758 506 0.0204531793 79 0.0049137603 293 0.7436827863 507 0.0191872431 80 0.0050393022 294 0.7513137456 508 0.0179433381 81 0.0051407353 295 0.7587080760 509 0.0167324712 82 0.0052461166 296 0.7658674865 510 0.0155405553 83 0.0053471681 297 0.7727780881 511 0.0143904666 84 0.0054196775 298 0.7794287519 512 -0.0132718220 85 0.0054876040 299 0.7858353120 513 -0.0121849995 86 0.0055475714 300 0.7919735841 514 -0.0111315548 87 0.0055938023 301 0.7978466413 515 -0.0101150215 88 0.0056220643 302 0.8034485751 516 -0.0091325329 89 0.0056455196 303 0.8087695004 517 -0.0081798233 90 0.0056389199 304 0.8138191270 518 -0.0072615816 91 0.0056266114 305 0.8185776004 519 -0.0063792293 92 0.0055917128 306 0.8230419890 520 -0.0055337211 93 0.0055404363 307 0.8272275347 521 -0.0047222596 94 0.0054753783 308 0.8311038457 522 -0.0039401124 95 0.0053838975 309 0.8346937361 523 -0.0031933778 96 0.0052715758 310 0.8379717337 524 -0.0024826723 97 0.0051382275 311 0.8409541392 525 -0.0018039472 98 0.0049839687 312 0.8436238281 526 -0.0011568135 99 0.0048109469 313 0.8459818469 527 -0.0005464280 100 0.0046039530 314 0.8480315777 528 0.0000276045 101 0.0043801861 315 0.8497805198 529 0.0005832264 102 0.0041251642 316 0.8511971524 530 0.0010902329 103 0.0038456408 317 0.8523047035 531 0.0015784682 104 0.0035401246 318 0.8531020949 532 0.0020274176 105 0.0032091885 319 0.8535720573 533 0.0024508540 106 0.0028446757 320 0.8537385600 534 0.0028446757 107 0.0024508540 321 0.8535720573 535 0.0032091885 108 0.0020274176 322 0.8531020949 536 0.0035401246 109 0.0015784682 323 0.8523047035 537 0.0038456408 110 0.0010902329 324 0.8511971524 538 0.0041251642 111 0.0005832264 325 0.8497805198 539 0.0043801861 112 0.0000276045 326 0.8480315777 540 0.0046039530 113 -0.0005464280 327 0.8459818469 541 0.0048109469 114 -0.0011568135 328 0.8436238281 542 0.0049839687 115 -0.0018039472 329 0.8409541392 543 0.0051382275 116 -0.0024826723 330 0.8379717337 544 0.0052715758 117 -0.0031933778 331 0.8346937361 545 0.0053838975 118 -0.0039401124 332 0.8311038457 546 0.0054753783 119 -0.0047222596 333 0.8272275347 547 0.0055404363 120 -0.0055337211 334 0.8230419890 548 0.0055917128 121 -0.0063792293 335 0.8185776004 549 0.0056266114 122 -0.0072615816 336 0.8138191270 550 0.0056389199 123 -0.0081798233 337 0.8087695004 551 0.0056455196 124 -0.0091325329 338 0.8034485751 552 0.0056220643 125 -0.0101150215 339 0.7978466413 553 0.0055938023 126 -0.0111315548 340 0.7919735841 554 0.0055475714 127 -0.0121849995 341 0.7858353120 555 0.0054876040 128 0.0132718220 342 0.7794287519 556 0.0054196775 129 0.0143904666 343 0.7727780881 557 0.0053471681 130 0.0155405553 344 0.7658674865 558 0.0052461166 131 0.0167324712 345 0.7587080760 559 0.0051407353 132 0.0179433381 346 0.7513137456 560 0.0050393022 133 0.0191872431 347 0.7436827863 561 0.0049137603 134 0.0204531793 348 0.7358211758 562 0.0047932560 135 0.0217467550 349 0.7277448900 563 0.0046606460 136 0.0230680169 350 0.7194462634 564 0.0045209852 137 0.0244160992 351 0.7109410426 565 0.0043730719 138 0.0257875847 352 0.7022388719 566 0.0042264269 139 0.0271859429 353 0.6933282376 567 0.0040819753 140 0.0286072173 354 0.6842353293 568 0.0039207432 141 0.0300502657 355 0.6749663190 569 0.0037603922 142 0.0315017608 356 0.6655139880 570 0.0036008268 143 0.0329754081 357 0.6559016302 571 0.0034418874 144 0.0344620948 358 0.6461269695 572 0.0032739613 145 0.0359697560 359 0.6361980107 573 0.0031125420 146 0.0374812850 360 0.6261242695 574 0.0029469447 147 0.0390053679 361 0.6159109932 575 0.0027870464 148 0.0405349170 362 0.6055783538 576 0.0026201758 149 0.0420649094 363 0.5951123086 577 0.0024625616 150 0.0436097542 364 0.5845403235 578 0.0023017254 151 0.0451488405 365 0.5738524131 579 0.0021461583 152 0.0466843027 366 0.5630789140 580 0.0019841140 153 0.0482165720 367 0.5522051258 581 0.0018348265 154 0.0497385755 368 0.5412553448 582 0.0016868083 155 0.0512556155 369 0.5302240895 583 0.0015443219 156 0.0527630746 370 0.5191234970 584 0.0013902494 157 0.0542452768 371 0.5079817500 585 0.0012577884 158 0.0557173648 372 0.4967708254 586 0.0011250155 159 0.0571616450 373 0.4855253091 587 0.0009885988 160 0.0585915683 374 0.4742453214 588 0.0008608443 161 0.0599837480 375 0.4629308085 589 0.0007458025 162 0.0613455171 376 0.4515996535 590 0.0006239376 163 0.0626857808 377 0.4402553754 591 0.0005107388 164 0.0639715898 378 0.4289119920 592 0.0004026540 165 0.0652247106 379 0.4175696896 593 0.0002949531 166 0.0664367512 380 0.4062317676 594 0.0002043017 167 0.0676075985 381 0.3949211761 595 0.0001094383 168 0.0687043828 382 0.3836350013 596 0.0000134949 169 0.0697630244 383 0.3723795546 597 -0.0000617334 170 0.0707628710 384 -0.3611589903 598 -0.0001446380 171 0.0717002673 385 -0.3499914122 599 -0.0002098337 172 0.0725682583 386 -0.3388722693 600 -0.0002896981 173 0.0733620255 387 -0.3278113727 601 -0.0003501175 174 0.0741003642 388 -0.3168278913 602 -0.0004095121 175 0.0747452558 389 -0.3059098575 603 -0.0004606325 176 0.0753137336 390 -0.2950716717 604 -0.0005145572 177 0.0758008358 391 -0.2843214189 605 -0.0005564576 178 0.0761992479 392 -0.2736634040 606 -0.0005946118 179 0.0764992170 393 -0.2631053299 607 -0.0006341594 180 0.0767093490 394 -0.2526480309 608 -0.0006650415 181 0.0768173975 395 -0.2423016884 609 -0.0006917937 182 0.0768230011 396 -0.2320690870 610 -0.0007215391 183 0.0767204924 397 -0.2219652696 611 -0.0007319357 184 0.0765050718 398 -0.2119735853 612 -0.0007530001 185 0.0761748321 399 -0,2021250176 613 -0.0007630793 186 0.0757305756 400 -0.1923966745 614 -0.0007757977 187 0.0751576255 401 -0.1828172548 615 -0.0007801449 188 0.0744664394 402 -0.1733808172 616 -0.0007803664 189 0.0736406005 403 -0.1640958855 617 -0.0007779869 190 0.0726774642 404 -0.1549607071 618 -0.0007834332 191 0.0715826364 405 -0.1459766491 619 -0.0007724848 192 0.0703533073 406 -0.1371551761 620 -0.0007681371 193 0.0689664013 407 -0.1285002850 621 -0.0007490598 194 0.0674525021 408 -0.1200077984 622 -0.0007440941 195 0.0657690668 409 -0.1116826931 623 -0.0007255043 196 0.0639444805 410 -0.1035329531 624 -0.0007157736 197 0.0619602779 411 -0.0955533352 625 -0.0006941614 198 0.0598166570 412 -0.0877547536 626 -0.0006777690 199 0.0575152691 413 -0.0801372934 627 -0.0006540333 200 0.0550460034 414 -0.0726943300 628 -0.0006312493 201 0.0524093821 415 -0.0654409853 629 -0.0006132747 202 0.0495978676 416 -0.0583705326 630 -0.0005870930 203 0.0466303305 417 -0.0514804176 631 -0.0005677802 204 0.0434768782 418 -0.0447806821 632 -0.0005466565 205 0.0401458278 419 -0.0382776572 633 -0.0005226564 206 0.0366418116 420 -0.0319531274 634 -0.0005040714 207 0.0329583930 421 -0.0258227288 635 -0.0004893791 208 0.0290824006 422 -0.0198834129 636 -0.0004875227 209 0.0250307561 423 -0.0141288827 637 -0.0004947518 210 0.0207997072 424 -0.0085711749 638 -0.0005617692 211 0.0163701258 425 -0.0032086896 639 -0.0005525280 212 0.0117623832 426 0.0019765601 213 0.0069636862 427 0.0069636862

The prototype filter, p ₀ (n), can also be derived from Table 4 by one or more mathematical operations such as rounding, downsampling, interpolation, and decimation.

Although the setting of the SBR-related control information is typically independent of the transposition details (as discussed previously), in some embodiments, certain control data elements may be simultaneously transmitted in the eSBR extension container (bs_extension_id==EXTENSION_ID_ESBR) to improve the quality of the regenerated signal. Some of the simultaneously transmitted elements may include noise level data (e.g., scaling factors of the noise floor and a parameter indicating the direction, frequency or time, of delta encoding for each noise level), inverse filtering data (e.g., a parameter indicating the inverse filtering mode selected from no inverse filtering, low inverse filtering, intermediate inverse filtering, and high inverse filtering), and missing harmonics data (e.g., a parameter indicating whether a sine wave should be added to a specific frequency range of the regenerated high band). All of these elements rely on the synthesized emulation of the decoder transposer performed in the encoder, and therefore, when properly configured for the selected transposer, can improve the quality of the regenerated signal.

Specifically, in some embodiments, the missing harmonics data and the inverse filtering control data are transmitted in an eSBR extension container (along with other bitstream parameters of Table 3) and are configured for the eSBR harmonic transposer. The additional bit rate required to transmit these two classes of metadata for the eSBR harmonic transposer is relatively low. Therefore, sending the configured missing harmonics data and/or the inverse filtering control data in an eSBR extension container will improve the quality of the audio generated by the transposer with only a minimal impact on the bit rate. To ensure backward compatibility with legacy decoders, the parameters configured for the SBR spectral transfer operation may also be sent in the bitstream as part of the SBR control data using either implicit or explicit signaling.

It is to be understood that within the scope of the appended claims, the invention may be practiced otherwise than as specifically described herein. Any reference signs contained in the following claims are for illustrative purposes only and are not to be used to interpret or limit the claims in any way. Various aspects of the present disclosure will be apparent from the following numbered exemplary embodiments (EEE):

EEE1. A method for performing high-frequency restoration of an audio signal, the method comprising:

receiving an encoded audio bitstream, wherein the encoded audio bitstream includes audio data representing a portion of a low frequency band of an audio signal and high frequency reconstruction metadata;

decoding the audio data to generate a decoded audio signal in the low frequency band;

extracting high frequency restoration metadata from an encoded audio bitstream, the high frequency restoration metadata including operating parameters for a high frequency restoration process, the operating parameters including a patch insertion mode parameter located in an extension container of the encoded audio bitstream, wherein a first value of the patch insertion mode parameter indicates a spectral transposition, and a second value of the patch insertion mode parameter indicates a harmonic transposition by frequency extension of a phase vocoder;

low-pass filtering the decoded audio signal to generate a low-pass filtered audio signal;

regenerating a portion of the highband of the audio signal using the filtered lowband audio signal and the highband reconstruction metadata, wherein the regeneration includes a spectral transposition if the patch insertion mode parameter has a first value, and the regeneration includes a harmonic transposition by frequency extension of a phase vocoder if the patch insertion mode parameter has a second value; and

combining a filtered low-pass audio signal with a regenerated portion of the high-pass band to form a wideband audio signal.

EEE2. The method according to EEE 1, wherein the extension container includes reverse filtering control data to be used when the patch insertion mode parameter is equal to the second value.

EEE3. The method according to any one of EEE 1-2, wherein the expansion container further includes control data of missing harmonics to be used when the patch insertion mode parameter is equal to the second value.

EEE4. The method according to any preceding EEE, wherein the encoded audio bitstream further includes a fill element with an identifier indicating the start of the fill element and fill data after the identifier, wherein the fill data includes an extension container.

EEE5. The method according to EEE 4, wherein the identifier is a three-bit unsigned integer with the most significant bit transmitted first and has the value 0x6.

EEE6. The method of EEE 4 or EEE 5, wherein the fill data includes an extension payload, the extension payload includes spectral range replication extension data, and the extension payload is identified by a first four-bit unsigned integer with the most significant bit transmitted first and having a value of ‘1101’ or ‘1110’, and, optionally,

wherein these spectral range replication expansion data include:

optional spectral range replication header,

spectral range replication data after the header, and

a spectral range replication extension element after the spectral range replication data, and wherein the label is included in the spectral range replication extension element.

EEE7. The method of any one of EEE 1-6, wherein the high frequency reconstruction metadata includes envelope scaling factors, noise floor scaling factors, time/frequency grid information, or a parameter indicating a transition frequency.

EEE8. The method of any one of EEE 1-7, wherein the filtering is performed by an analysis filter bank that includes analysis filters, h _k (n), that are modulated versions of a prototype filter, p ₀ (n), according to:

where p ₀ (n) is a real-valued symmetric or asymmetric prototype filter, M is the number of channels in the analysis filter bank, and N is the order of the prototype filter.

EEE9. The method of EEE 8, wherein the prototype filter, p ₀ (n), is derived from the coefficients of Table 4 in this document.

EEE10. The method of EEE 8, wherein the prototype filter, p ₀ (n), is derived from the coefficients of Table 4 herein by one or more mathematical operations selected from the group consisting of rounding, subsampling, interpolation, or decimation.

EEE11. The method of any one of EEE 1-10, wherein the phase shift is added to the filtered audio signal in the low-pass band after filtering and compensated before combining to reduce the complexity of the method.

EEE12. The method according to any preceding EEE, wherein the extension container further includes a label indicating whether additional pre-processing is used to avoid interruptions in the shape of the spectral envelope of the high-frequency band portion when the patch insertion mode parameter is equal to a first value, wherein the first label value enables the additional pre-processing, and the second label value disables the additional pre-processing.

EEE13. The method of EEE 12, wherein the additional pre-processing comprises calculating a pre-gain curve using a linear prediction filter coefficient.

EEE14. The method according to any one of EEE 1-13, wherein the extension container is a backward compatible extension container.

EEE15. The method according to any one of EEE 1-14, wherein the encoded audio stream is encoded in accordance with a format, and wherein the extension container is an extension container that is defined in at least one legacy version of said format.

EEE16. A non-transitory computer-readable medium containing instructions that, when executed by a processor, perform the method of any one of EEE 1-15.

EEE17. An audio processing module for performing high-frequency restoration of an audio signal, wherein the audio processing module is configured to perform the method according to any one of EEE 1-15.

Claims (22)

1. A method for performing high-frequency restoration of an audio signal, comprising: receiving an encoded audio bitstream, wherein the encoded audio bitstream includes audio data representing a portion of a low frequency band of an audio signal and high frequency reconstruction metadata; decoding the audio data to generate a decoded audio signal in the low frequency band; extracting high frequency restoration metadata from an encoded audio bitstream, wherein the high frequency restoration metadata includes operating parameters for a high frequency restoration process, the operating parameters including a patch insertion mode parameter located in a backward compatible expansion container of the encoded audio bitstream, wherein a first value of the patch insertion mode parameter indicates a spectral transposition and a second value of the patch insertion mode parameter indicates a harmonic transposition by frequency extension of a phase vocoder, wherein the encoded audio bitstream further includes a filler element with an identifier indicating the start of the filler element and filler data following the identifier, wherein the filler data includes a backward compatible expansion container and wherein the identifier is a three-bit unsigned integer with the most significant bit transmitted first, having a value of 0x6; low-band filtering the decoded audio signal to generate a low-band filtered audio signal, the filtering being performed using an analysis filter bank that includes analysis filters, h _k (n), that are modulated versions of a prototype filter, p ₀ (n), according to , , where p ₀ (n) is a real-valued symmetric or asymmetric prototype filter, M is the number of channels in the analysis filter bank, and N is the order of the prototype filter; and regenerating a portion of the high-band audio signal using the filtered low-band audio signal and high-frequency restoration metadata, wherein the regeneration includes spectral transposition if the patch insertion mode parameter has a first value, and the regeneration includes harmonic transposition by frequency expansion of a phase vocoder if the patch insertion mode parameter has a second value. 2. The method according to claim 1, wherein the backward compatible extension container includes reverse filtering control data to be used when the patch insertion mode parameter is equal to the second value. 3. The method according to claim 1, wherein the backward compatible expansion container further includes control data of missing harmonics to be used when the patch insertion mode parameter is equal to the second value. 4. The method according to claim 1, wherein the phase shift is added to the filtered audio signal in the low-frequency band after filtering and compensated before combining in order to reduce the complexity of the method. 5. The method of claim 1, wherein the backward compatible extension container further includes a label indicating whether additional pre-processing is used to avoid interruptions in the shape of the spectral envelope of the high frequency band portion when the patch insertion mode parameter is equal to the first value, wherein the first label value enables additional pre-processing, and the second label value disables additional pre-processing. 6. The method of claim 5, wherein the additional pre-processing comprises calculating a pre-gain curve using a linear prediction filter coefficient. 7. A non-transitory computer-readable medium containing instructions that, when executed by a processor, perform the method of claim 1. 8. An audio processing module for performing high-frequency restoration of an audio signal, wherein the audio processing module comprises: an input interface for receiving an encoded audio bitstream, wherein the encoded audio bitstream includes audio data representing a portion of a low frequency band of an audio signal and high frequency reconstruction metadata; a basic audio decoder for decoding audio data to generate a decoded audio signal in the low-frequency band; a deformatting unit for extracting high-frequency restoration metadata from an encoded audio bitstream, wherein the high-frequency restoration metadata includes operating parameters for a high-frequency restoration process, wherein the operating parameters include a fill element with an identifier indicating the start of the fill element and fill data following the identifier, wherein the fill data includes a backward-compatible extension container including a patch insertion mode parameter, wherein a first value of the patch insertion mode parameter indicates a spectral transfer and a second value of the patch insertion mode parameter indicates a harmonic transposition by means of frequency extension of a phase vocoder, and wherein the identifier is a three-bit unsigned integer with the most significant bit transmitted first, having a value of 0×6; an analysis filter bank for low-band filtering the decoded audio signal to generate a low-band filtered audio signal, the filtering being performed using an analysis filter bank that includes analysis filters, h _k (n), that are modulated versions of a prototype filter, p ₀ (n), according to , , where p ₀ (n) is a real-valued symmetric or asymmetric prototype filter, M is the number of channels in the analysis filter bank, and N is the order of the prototype filter; and a high-frequency regenerator for regenerating a portion of a high-frequency band of an audio signal using the filtered low-frequency band audio signal and high-frequency regeneration metadata, wherein the regeneration includes spectral transposition if the patch insertion mode parameter has a first value, and the regeneration includes harmonic transposition by frequency extension of a phase vocoder if the patch insertion mode parameter has a second value.