Patent Application
20230402046
The present invention relates to audio signal encoding and decoding and, in particular, to audio signal processing using parallel frequency domain and time domain encoder/decoder processors.
The perceptual coding of audio signals for the purpose of data reduction for efficient storage or transmission of these signals is a widely used practice. In particular when lowest bit rates are to be achieved, the employed coding leads to a reduction of audio quality that often is primarily caused by a limitation at the encoder side of the audio signal bandwidth to be transmitted. Here, typically the audio signal is low-pass filtered such that no spectral waveform content remains above a certain pre-determined cut-off frequency.
In contemporary codecs well-known methods exist for the decoder-side signal restoration through audio signal Bandwidth Extension (BWE), e.g. Spectral Band Replication (SBR) that operates in frequency domain or so-called Time Domain Bandwidth Extension (TD-BWE) being is a post-processor in speech coders that operates in time domain.
Additionally, several combined time domain/frequency domain coding concepts exist such as concepts known under the term AMR-WB+ or USAC.
All these combined time domain/coding concepts have in common that the frequency domain coder relies on bandwidth extension technologies which incur a band limitation into the input audio signal and the portion above a cross-over frequency or border frequency is encoded with a low resolution coding concept and synthesized on the decoder-side. Hence, such concepts mainly rely on a pre-processor technology on the encoder side and a corresponding post-processing functionality on the decoder-side.
Typically, the time domain encoder is selected for useful signals to be encoded in the time domain such as speech signals and the frequency domain encoder is selected for non-speech signals, music signals, etc. However, specifically for non-speech signals having prominent harmonics in the high frequency band, the known frequency domain encoders have a reduced accuracy and, therefore, a reduced audio quality due to the fact that such prominent harmonics can only be separately parametrically encoded or are eliminated at all in the encoding/decoding process.
Furthermore, concepts exist in which the time domain encoding/decoding branch additionally relies on the bandwidth extension which also parametrically encodes an upper frequency range while a lower frequency range is typically encoded using an ACELP or any other CELP related coder, for example a speech coder. This bandwidth extension functionality increases the bitrate efficiency but, on the other hand, introduces further inflexibility due to the fact that both encoding branches, i.e., the frequency domain encoding branch and the time domain encoding branch are band limited due to the bandwidth extension procedure or spectral band replication procedure operating above a certain crossover frequency substantially lower than the maximum frequency included in the input audio signal
Relevant Topics in the State-of-Art Comprise
The following papers and patents describe methods that are considered to constitute known technology for the application:
In MPEG-D USAC, a switchable core coder is described. However, in USAC, the band-limited core is restricted to transmit a low-pass filtered signal. Therefore, certain music signals that contain prominent high frequency content e.g. full-band sweeps, triangle sounds, etc. cannot be reproduced faithfully.
According to an embodiment, an audio encoder for encoding an audio signal may have: a first encoding processor for encoding a first audio signal portion in a frequency domain, wherein the first encoding processor has: a time frequency converter for converting the first audio signal portion into a frequency domain representation having spectral lines up to a maximum frequency of the first audio signal portion; an analyzer for analyzing the frequency domain representation up to the maximum frequency to determine first spectral portions to be encoded with a first spectral resolution and second spectral portions to be encoded with a second spectral resolution, the second spectral resolution being lower than the first spectral resolution, wherein the analyzer is configured to determine a first spectral portion from the first spectral portions, the first spectral portion being placed, with respect to frequency, between two second spectral portions from the second spectral portions; a spectral encoder for encoding the first spectral portions with the first spectral resolution and for encoding the second spectral portions with the second spectral resolution, wherein the spectral encoder has a parametric coder for calculating spectral envelope information having the second spectral resolution from the second spectral portions; a second encoding processor for encoding a second different audio signal portion in the time domain, wherein the second encoding processor has: a sampling rate converter for converting the second audio signal portion to a lower sampling rate representation, the lower sampling rate being lower than a sampling rate of the audio signal, wherein the lower sampling rate representation does not include the high band of the input signal; a time domain low band encoder for time domain encoding the lower sampling rate representation; and a time domain bandwidth extension encoder for parametrically encoding the high band; a controller configured for analyzing the audio signal and for determining, which portion of the audio signal is the first audio signal portion encoded in the frequency domain and which portion of the audio signal is the second audio signal portion encoded in the time domain; and an encoded signal former for forming an encoded audio signal having a first encoded signal portion for the first audio signal portion and a second encoded signal portion for the second audio signal portion.
According to another embodiment, an audio decoder for decoding an encoded audio signal may have: a first decoding processor for decoding a first encoded audio signal portion in a frequency domain, the first decoding processor having: a spectral decoder for decoding first spectral portions with a high spectral resolution and for synthesizing second spectral portions using a parametric representation of the second spectral portions and at least a decoded first spectral portion to obtain a decoded spectral representation, wherein the spectral decoder is configured to generate the first decoded representation so that a first spectral portion is placed with respect to frequency between two second spectral portions; and a frequency-time converter for converting the decoded spectral representation into a time domain to obtain a decoded first audio signal portion; a second decoding processor for decoding a second encoded audio signal portion in the time domain to obtain a decoded second audio signal portion, wherein the second decoding processor has: a time domain low band decoder for decoding a low band time domain signal; an upsampler for upsampling the low band time domain signal; a time domain bandwidth extension decoder for synthesizing a high band of a time domain output signal; and a mixer for mixing a synthesized high band of the time domain signal and an upsampled low band time domain signal; and a combiner for combining the decoded first spectral portion and the decoded second spectral portion to obtain a decoded audio signal.
According to still another embodiment, a method of encoding an audio signal may have the steps of: first encoding a first audio signal portion in a frequency domain, wherein the first encoding has: converting the first audio signal portion into a frequency domain representation having spectral lines up to a maximum frequency of the first audio signal portion; analyzing the frequency domain representation up to the maximum frequency to determine first spectral portions to be encoded with a first spectral resolution and second spectral portions to be encoded with a second spectral resolution, the second spectral resolution being lower than the first spectral resolution, wherein the analyzing determines a first spectral portion from the first spectral portions, the first spectral portion being placed, with respect to frequency, between two second spectral portions from the second spectral portions; encoding the first spectral portions with the first spectral resolution and for encoding the second spectral portions with the second spectral resolution, wherein the encoding the second spectral portion has calculating, from the second spectral portions, spectral envelope information having the second spectral resolution; second encoding a second different audio signal portion in the time domain wherein the second encoding has: converting the second audio signal portion to a lower sampling rate representation, the lower sampling rate being lower than a sampling rate of the audio signal, wherein the lower sampling rate representation does not include the high band of the input signal; time domain encoding the lower sampling rate representation; and parametrically encoding the high band; analyzing the audio signal and determining, which portion of the audio signal is the first audio signal portion encoded in the frequency domain and which portion of the audio signal is the second audio signal portion encoded in the time domain; and forming an encoded audio signal having a first encoded signal portion for the first audio signal portion and a second encoded signal portion for the second audio signal portion.
According to another embodiment, a method of decoding an encoded audio signal may have the steps of: first decoding a first encoded audio signal portion in a frequency domain, the first decoding having: decoding first spectral portions with a high spectral resolution and synthesizing second spectral portions using a parametric representation of the second spectral portions and at least a decoded first spectral portion to obtain a decoded spectral representation, wherein decoding has generating the first decoded representation so that a first spectral portion is placed with respect to frequency between two second spectral portions; and converting the decoded spectral representation into a time domain to obtain a decoded first audio signal portion; second decoding a second encoded audio signal portion in the time domain to obtain a decoded second audio signal portion, wherein the second decoding has: decoding a low band time domain signal; upsampling the low band time domain signal; synthesizing a high band of a time domain output signal; and mixing a synthesized high band of the time domain signal and an upsampled low band time domain signal; and combining the decoded first spectral portion and the decoded second spectral portion to obtain a decoded audio signal.
Another embodiment may have a non-transitory digital storage medium having stored thereon a computer program for performing a method of encoding an audio signal, having: first encoding a first audio signal portion in a frequency domain, wherein the first encoding has: converting the first audio signal portion into a frequency domain representation having spectral lines up to a maximum frequency of the first audio signal portion; analyzing the frequency domain representation up to the maximum frequency to determine first spectral portions to be encoded with a first spectral resolution and second spectral portions to be encoded with a second spectral resolution, the second spectral resolution being lower than the first spectral resolution, wherein the analyzing determines a first spectral portion from the first spectral portions, the first spectral portion being placed, with respect to frequency, between two second spectral portions from the second spectral portions; encoding the first spectral portions with the first spectral resolution and for encoding the second spectral portions with the second spectral resolution, wherein the encoding the second spectral portion has calculating, from the second spectral portions, spectral envelope information having the second spectral resolution; second encoding a second different audio signal portion in the time domain wherein the second encoding has: converting the second audio signal portion to a lower sampling rate representation, the lower sampling rate being lower than a sampling rate of the audio signal, wherein the lower sampling rate representation does not include the high band of the input signal; time domain encoding the lower sampling rate representation; and parametrically encoding the high band; analyzing the audio signal and determining, which portion of the audio signal is the first audio signal portion encoded in the frequency domain and which portion of the audio signal is the second audio signal portion encoded in the time domain; and forming an encoded audio signal having a first encoded signal portion for the first audio signal portion and a second encoded signal portion for the second audio signal portion, when said computer program is run by a computer.
Still another embodiment may have a non-transitory digital storage medium having stored thereon a computer program for performing a method of decoding an encoded audio signal, having: first decoding a first encoded audio signal portion in a frequency domain, the first decoding having: decoding first spectral portions with a high spectral resolution and synthesizing second spectral portions using a parametric representation of the second spectral portions and at least a decoded first spectral portion to obtain a decoded spectral representation, wherein decoding has generating the first decoded representation so that a first spectral portion is placed with respect to frequency between two second spectral portions; and converting the decoded spectral representation into a time domain to obtain a decoded first audio signal portion; second decoding a second encoded audio signal portion in the time domain to obtain a decoded second audio signal portion, wherein the second decoding has: decoding a low band time domain signal; upsampling the low band time domain signal; synthesizing a high band of a time domain output signal; and mixing a synthesized high band of the time domain signal and an upsampled low band time domain signal; and combining the decoded first spectral portion and the decoded second spectral portion to obtain a decoded audio signal, when said computer program is run by a computer.
The present invention is based on the finding that a time domain encoding/decoding processor can be combined with a frequency domain encoding/decoding processor having a gap filling functionality but this gap filling functionality for filling spectral holes is operated over the whole band of the audio signal or at least above a certain gap filling frequency. Importantly, the frequency domain encoding/decoding processor is particularly in the position to perform accurate or wave form or spectral value encoding/decoding up to the maximum frequency and not only until a crossover frequency. Furthermore, the full-band capability of the frequency domain encoder for encoding with the high resolution allows an integration of the gap filling functionality into the frequency domain encoder.
Hence, in accordance with the present invention by using the full-band spectral encoder/decoder processor, the problems related to the separation of the bandwidth extension on the one hand and the core coding on the other hand can be addressed and overcome by performing the bandwidth extension in the same spectral domain in which the core decoder operates. Therefore, a full rate core decoder is provided which encodes and decodes the full audio signal range. This does not require the need for a downsampler on the encoder side and an upsampler on the decoder side. Instead, the whole processing is performed in the full sampling rate or full-bandwidth domain. In order to obtain a high coding gain, the audio signal is analyzed in order to find a first set of first spectral portions which has to be encoded with a high resolution, where this first set of first spectral portions may include, in an embodiment, tonal portions of the audio signal. On the other hand, non-tonal or noisy components in the audio signal constituting a second set of second spectral portions are parametrically encoded with low spectral resolution. The encoded audio signal then only necessitates the first set of first spectral portions encoded in a waveform-preserving manner with a high spectral resolution and, additionally, the second set of second spectral portions encoded parametrically with a low resolution using frequency “tiles” sourced from the first set. On the decoder side, the core decoder, which is a full-band decoder, reconstructs the first set of first spectral portions in a waveform-preserving manner, i.e., without any knowledge that there is any additional frequency regeneration. However, the so generated spectrum has a lot of spectral gaps. These gaps are subsequently filled with the inventive Intelligent Gap Filling (IGF) technology by using a frequency regeneration applying parametric data on the one hand and using a source spectral range, i.e., first spectral portions reconstructed by the full rate audio decoder on the other hand.
In further embodiments, spectral portions, which are reconstructed by noise filling only rather than bandwidth replication or frequency tile filling, constitute a third set of third spectral portions. Due to the fact that the coding concept operates in a single domain for the core coding/decoding on the one hand and the frequency regeneration on the other hand, the IGF is not only restricted to fill up a higher frequency range but can fill up lower frequency ranges, either by noise filling without frequency regeneration or by frequency regeneration using a frequency tile at a different frequency range.
Furthermore, it is emphasized that an information on spectral energies, an information on individual energies or an individual energy information, an information on a survive energy or a survive energy information, an information a tile energy or a tile energy information, or an information on a missing energy or a missing energy information may comprise not only an energy value, but also an (e.g. absolute) amplitude value, a level value or any other value, from which a final energy value can be derived. Hence, the information on an energy may e.g. comprise the energy value itself, and/or a value of a level and/or of an amplitude and/or of an absolute amplitude.
A further aspect is based on the finding that the correlation situation is not only important for the source range but is also important for the target range. Furthermore, the present invention acknowledges the situation that different correlation situations can occur in the source range and the target range. When, for example, a speech signal with high frequency noise is considered, the situation can be that the low frequency band comprising the speech signal with a small number of overtones is highly correlated in the left channel and the right channel, when the speaker is placed in the middle. The high frequency portion, however, can be strongly uncorrelated due to the fact that there might be a different high frequency noise on the left side compared to another high frequency noise or no high frequency noise on the right side. Thus, when a straightforward gap filling operation would be performed that ignores this situation, then the high frequency portion would be correlated as well, and this might generate serious spatial segregation artifacts in the reconstructed signal. In order to address this issue, parametric data for a reconstruction band or, generally, for the second set of second spectral portions which have to be reconstructed using a first set of first spectral portions is calculated to identify either a first or a second different two-channel representation for the second spectral portion or, stated differently, for the reconstruction band. On the encoder side, a two-channel identification is, therefore calculated for the second spectral portions, i.e., for the portions, for which, additionally, energy information for reconstruction bands is calculated. A frequency regenerator on the decoder side then regenerates a second spectral portion depending on a first portion of the first set of first spectral portions, i.e., the source range and parametric data for the second portion such as spectral envelope energy information or any other spectral envelope data and, additionally, dependent on the two-channel identification for the second portion, i.e., for this reconstruction band under reconsideration.
The two-channel identification is advantageously transmitted as a flag for each reconstruction band and this data is transmitted from an encoder to a decoder and the decoder then decodes the core signal as indicated by advantageously calculated flags for the core bands. Then, in an implementation, the core signal is stored in both stereo representations (e.g. left/right and mid/side) and, for the IGF frequency tile filling, the source tile representation is chosen to fit the target tile representation as indicated by the two-channel identification flags for the intelligent gap filling or reconstruction bands, i.e., for the target range.
It is emphasized that this procedure not only works for stereo signals, i.e., for a left channel and the right channel but also operates for multi-channel signals. In the case of multi-channel signals, several pairs of different channels can be processed in that way such as a left and a right channel as a first pair, a left surround channel and a right surround as the second pair and a center channel and an LFE channel as the third pair. Other pairings can be determined for higher output channel formats such as 7.1, 11.1 and so on.
A further aspect is based on the finding that the audio quality of the reconstructed signal can be improved through IGF since the whole spectrum is accessible to the core encoder so that, for example, perceptually important tonal portions in a high spectral range can still be encoded by the core coder rather than parametric substitution. Additionally, a gap filling operation using frequency tiles from a first set of first spectral portions which is, for example, a set of tonal portions typically from a lower frequency range, but also from a higher frequency range if available, is performed. For the spectral envelope adjustment on the decoder side, however, the spectral portions from the first set of spectral portions located in the reconstruction band are not further post-processed by e.g. the spectral envelope adjustment. Only the remaining spectral values in the reconstruction band which do not originate from the core decoder are to be envelope adjusted using envelope information. Advantageously, the envelope information is a full-band envelope information accounting for the energy of the first set of first spectral portions in the reconstruction band and the second set of second spectral portions in the same reconstruction band, where the latter spectral values in the second set of second spectral portions are indicated to be zero and are, therefore, not encoded by the core encoder, but are parametrically coded with low resolution energy information.
It has been found that absolute energy values, either normalized with respect to the bandwidth of the corresponding band or not normalized, are useful and very efficient in an application on the decoder side. This especially applies when gain factors have to be calculated based on a residual energy in the reconstruction band, the missing energy in the reconstruction band and frequency tile information in the reconstruction band.
Furthermore, it is of advantage that the encoded bitstream not only covers energy information for the reconstruction bands but, additionally, scale factors for scale factor bands extending up to the maximum frequency. This ensures that for each reconstruction band, for which a certain tonal portion, i.e., a first spectral portion is available, this first set of first spectral portion can actually be decoded with the right amplitude. Furthermore, in addition to the scale factor for each reconstruction band, an energy for this reconstruction band is generated in an encoder and transmitted to a decoder. Furthermore, it is of advantage that the reconstruction bands coincide with the scale factor bands or in case of energy grouping, at least the borders of a reconstruction band coincide with borders of scale factor bands.
A further aspect is based on the finding that certain impairments in audio quality can be remedied by applying a signal adaptive frequency tile filling scheme. To this end, an analysis on the encoder-side is performed in order to find out the best matching source region candidate for a certain target region. A matching information identifying for a target region a certain source region together with optionally some additional information is generated and transmitted as side information to the decoder. The decoder then applies a frequency tile filling operation using the matching information. To this end, the decoder reads the matching information from the transmitted data stream or data file and accesses the source region identified for a certain reconstruction band and, if indicated in the matching information, additionally performs some processing of this source region data to generate raw spectral data for the reconstruction band. Then, this result of the frequency tile filling operation, i.e., the raw spectral data for the reconstruction band, is shaped using spectral envelope information in order to finally obtain a reconstruction band that comprises the first spectral portions such as tonal portions as well. These tonal portions, however, are not generated by the adaptive tile filling scheme, but these first spectral portions are output by the audio decoder or core decoder directly.
The adaptive spectral tile selection scheme may operate with a low granularity. In this implementation, a source region is subdivided into typically overlapping source regions and the target region or the reconstruction bands are given by non-overlapping frequency target regions. Then, similarities between each source region and each target region are determined on the encoder-side and the best matching pair of a source region and the target region are identified by the matching information and, on the decoder-side, the source region identified in the matching information is used for generating the raw spectral data for the reconstruction band.
For the purpose of obtaining a higher granularity, each source region is allowed to shift in order to obtain a certain lag where the similarities are maximum. This lag can be as fine as a frequency bin and allows an even better matching between a source region and the target region.
Furthermore, in addition of only identifying a best matching pair, this correlation lag can also be transmitted within the matching information and, additionally, even a sign can be transmitted. When the sign is determined to be negative on the encoder-side, then a corresponding sign flag is also transmitted within the matching information and, on the decoder-side, the source region spectral values are multiplied by “−1” or, in a complex representation, are “rotated” by 180 degrees.
A further implementation of this invention applies a tile whitening operation. Whitening of a spectrum removes the coarse spectral envelope information and emphasizes the spectral fine structure which is of foremost interest for evaluating tile similarity. Therefore, a frequency tile on the one hand and/or the source signal on the other hand are whitened before calculating a cross correlation measure. When only the tile is whitened using a predefined procedure, a whitening flag is transmitted indicating to the decoder that the same predefined whitening process shall be applied to the frequency tile within IGF.
Regarding the tile selection, it is of advantage to use the lag of the correlation to spectrally shift the regenerated spectrum by an integer number of transform bins. Depending on the underlying transform, the spectral shifting may necessitate addition corrections. In case of odd lags, the tile is additionally modulated through multiplication by an alternating temporal sequence of −1/1 to compensate for the frequency-reversed representation of every other band within the MDCT. Furthermore, the sign of the correlation result is applied when generating the frequency tile.
Furthermore, it is of advantage to use tile pruning and stabilization in order to make sure that artifacts created by fast changing source regions for the same reconstruction region or target region are avoided. To this end, a similarity analysis among the different identified source regions is performed and when a source tile is similar to other source tiles with a similarity above a threshold, then this source tile can be dropped from the set of potential source tiles since it is highly correlated with other source tiles. Furthermore, as a kind of tile selection stabilization, it is of advantage to keep the tile order from the previous frame if none of the source tiles in the current frame correlate (better than a given threshold) with the target tiles in the current frame.
A further aspect is based on the finding that an improved quality and reduced bitrate specifically for signals comprising transient portions as they occur very often in audio signals is obtained by combining the Temporal Noise Shaping (TNS) or Temporal Tile Shaping (TTS) technology with high frequency reconstruction. The TNS/TTS processing on the encoder-side being implemented by a prediction over frequency reconstructs the time envelope of the audio signal. Depending on the implementation, i.e., when the temporal noise shaping filter is determined within a frequency range not only covering the source frequency range but also the target frequency range to be reconstructed in a frequency regeneration decoder, the temporal envelope is not only applied to the core audio signal up to a gap filling start frequency, but the temporal envelope is also applied to the spectral ranges of reconstructed second spectral portions. Thus, pre-echoes or post-echoes that would occur without temporal tile shaping are reduced or eliminated. This is accomplished by applying an inverse prediction over frequency not only within the core frequency range up to a certain gap filling start frequency but also within a frequency range above the core frequency range. To this end, the frequency regeneration or frequency tile generation is performed on the decoder-side before applying a prediction over frequency. However, the prediction over frequency can either be applied before or subsequent to spectral envelope shaping depending on whether the energy information calculation has been performed on the spectral residual values subsequent to filtering or to the (full) spectral values before envelope shaping.
The TTS processing over one or more frequency tiles additionally establishes a continuity of correlation between the source range and the reconstruction range or in two adjacent reconstruction ranges or frequency tiles.
In an implementation, it is of advantage to use complex TNS/TTS filtering. Thereby, the (temporal) aliasing artifacts of a critically sampled real representation, like MDCT, are avoided. A complex TNS filter can be calculated on the encoder-side by applying not only a modified discrete cosine transform but also a modified discrete sine transform in addition to obtain a complex modified transform. Nevertheless, only the modified discrete cosine transform values, i.e., the real part of the complex transform is transmitted. On the decoder-side, however, it is possible to estimate the imaginary part of the transform using MDCT spectra of preceding or subsequent frames so that, on the decoder-side, the complex filter can be again applied in the inverse prediction over frequency and, specifically, the prediction over the border between the source range and the reconstruction range and also over the border between frequency-adjacent frequency tiles within the reconstruction range.
The inventive audio coding system efficiently codes arbitrary audio signals at a wide range of bitrates. Whereas, for high bitrates, the inventive system converges to transparency, for low bitrates perceptual annoyance is minimized. Therefore, the main share of available bitrate is used to waveform code just the perceptually most relevant structure of the signal in the encoder, and the resulting spectral gaps are filled in the decoder with signal content that roughly approximates the original spectrum. A very limited bit budget is consumed to control the parameter driven so-called spectral Intelligent Gap Filling (IGF) by dedicated side information transmitted from the encoder to the decoder.
In further embodiments, the time domain encoding/decoding processor relies on a lower sampling rate and the corresponding bandwidth extension functionality.
In further embodiments, a cross-processor is provided in order to initialize the time domain encoder/decoder with initialization data derived from the currently processed frequency domain encoder/decoder signal This allows that when the currently processed audio signal portion is processed by the frequency domain encoder, the parallel time domain encoder is initialized so that when a switch from the frequency domain encoder to a time domain encoder takes place, this time domain encoder can start processing since all the initialization data relating to earlier signals are already there due to the cross-processor. This cross-processor may be applied on the encoder-side and, additionally, on the decoder-side and may use a frequency-time transform which additionally performs a very efficient downsampling from the higher output or input sampling rate into the lower time domain core coder sampling rate by only selecting a certain low band portion of the domain signal together with a certain reduced transform size. Thus, a sample rate conversion from the high sampling rate to the low sampling rate is very efficiently performed and this signal obtained by the transform with the reduced transform size can then be used for initializing the time domain encoder/decoder so that the time domain encoder/decoder is ready to immediately perform time domain encoding when this situation is signaled by a controller and the immediately preceding audio signal portion was encoded in the frequency domain.
Hence, embodiments of the present invention allow a seamless switching of a perceptual audio coder comprising spectral gap filling and a time domain encoder with or without bandwidth extension.
Hence, the present invention relies on methods that are not restricted to removing the high frequency content above a cut-off frequency in the frequency domain encoder from the audio signal but rather signal-adaptively removes spectral band-pass regions leaving spectral gaps in the encoder and subsequently reconstructs these spectral gaps in the decoder. Advantageously, an integrated solution such as intelligent gap filling is used that efficiently combines full-bandwidth audio coding and spectral gap filling particularly in the MDCT transform domain.
Hence, the present invention provides an improved concept for combining speech coding and a subsequent time domain bandwidth extension with a full-band wave form decoding comprising spectral gap filling into a switchable perceptual encoder/decoder.
Hence, in contrast to already existing methods, the new concept utilizes full-band audio signal wave form coding in the transform domain coder and at the same time allows a seamless switching to a speech coder advantageously followed by a time domain bandwidth extension.
Further embodiments of the present invention avoid the explained problems that occur due to a fixed band limitation. The concept enables the switchable combination of a full-band wave form coder in the frequency domain equipped with a spectral gap filling and a lower sampling rate speech coder and a time domain bandwidth extension. Such a coder is capable of wave form coding the aforementioned problematic signals providing full audio bandwidth up to the Nyquist frequency of the audio input signal. Nevertheless, seamless switching between both coding strategies is guaranteed particularly by the embodiments having the cross-processor. For this seamless switching, the cross-processor represents a cross connection at both encoder and decoder between the full-band capable full-rate (input sampling rate) frequency domain encoder and the low-rate ACELP coder having a lower sampling rate to properly initialize the ACELP parameters and buffers particularly within the adaptive codebook, the LPC filter or the resampling stage, when switching from the frequency domain coder such as TCX to the time domain encoder such as ACELP.
Embodiments of the present invention will subsequently be discussed with respect to the accompanying drawings in which:
The audio encoder of
Hence, the controller 620 makes sure that for a single audio signal portion only a time domain representation or a frequency domain representation is in the encoded signal. This can be accomplished by the controller 620 in several ways. One way would be that, for one and the same audio signal portion, both representations arrive at block 630 and the controller 620 controls the encoded signal former 630 to only introduce one of both representations into the encoded signal. Alternatively, however, the controller 620 can control an input into the first encoding processor and an input into the second encoding processor so that, based on the analysis of the corresponding signal portion, only one of both blocks 600 or 610 is activated to actually perform the full encoding operation and the other block is deactivated.
This deactivation can be a deactivation or, as illustrated with respect to, for example,
In the further specific implementation of the second encoding processor operating in the time domain, the second encoding processor comprises a downsampler 900 or sampling rate converter for converting the audio signal portion into a representation with a lower sampling rate, wherein the lower sampling rate is lower than a sampling rate at the input into the first encoding processor. This is illustrated in
In a further embodiment of the present invention the audio encoder additionally comprises, although not illustrated in
Furthermore, the preprocessor additionally comprises an entropy coder for generating an encoded version of the quantized prediction coefficients. It is important to note that the encoded signal former 630 or the specific implementation, i.e., the bit stream multiplexor 613 makes sure that the encoded version of the quantized prediction coefficients is included into the encoded audio signal 632. Advantageously, the LPC coefficients are not directly quantized but are converted into an ISF, for example, or any other representation better suited for quantization. This conversion may be performed either by the determined LPC coefficients block 1002 or is performed within the block 1010 for quantizing the LPC coefficients.
Furthermore, the preprocessor may comprise a resampler 1004 for resampling an audio input signal at an input sampling rate into a lower sampling rate for the time domain encoder. When the time domain encoder is an ACELP encoder having a certain ACELP sampling rate then the down sampling is performed to advantageously either 12.8 kHz or 16 kHz. The input sampling rate can be any of a particular number of sampling rates such as 32 kHz or an even higher sampling rate. On the other hand, the sampling rate of the time domain encoder will be predetermined by certain restrictions and the resampler 1004 performs this resampling and outputs the lower sampling rate representation of the input signal. Hence, the resampler 1004 can perform a similar functionality and can even be one and the same element as the downsampler 900 illustrated in the context of
Furthermore, it is of advantage to apply a pre-emphasis in the pre-emphasis block 1005 in
Furthermore, the preprocessor may additionally comprise a TCX-LTP parameter extraction for controlling an LTP post filter illustrated at 1420 in
As illustrated, the result of block 1006 is input into the encoded signal, i.e., is in the embodiment of
Hence, to summarize, common to both paths is a preprocessing operation 1000 in which commonly used signal processing operations are performed. These comprise a resampling to an ACELP sampling rate (12.8 or 16 kHz) for one parallel path and this resampling is performed. Furthermore, a TCX LTP parameter extraction illustrated at block 1006 is performed and, additionally, a pre-emphasis and a determination of LPC coefficients is performed. As outlined, the pre-emphasis compensates for the spectral tilt and, therefore, makes the calculation of LPC parameters at a given LPC order more efficient.
Subsequently, reference is made to
Based on this audio signal portion, the controller 620 addresses a frequency domain encoder simulator 621 and a time domain encoder simulator 622 in order to calculate for each encoder possibility an estimated signal to noise ratio. Subsequently, the selector 623 selects the encoder which has provided the better signal to noise ratio, naturally under the consideration of a predefined bit rate. The selector then identifies the corresponding encoder via the control output. When it is determined that the audio signal portion under consideration is to be encoded using the frequency domain encoder, the time domain encoder is set into an initialization state or in other embodiments not requiring a very instant switching in a completely deactivated state. However, when it is determined that the audio signal portion under consideration is to be encoded by the time domain encoder, the frequency domain encoder is then deactivated.
Subsequently, an implementation of the controller illustrated in
In case the TCX branch is chosen, a TCX decoder is run in each frame which outputs a signal at the ACELP sampling rate. This is used to update the memories used for the ACELT encoding path (LPC residual, Mem w0, Memory deemphasis), to enable instant switching from TCX to ACELP. The memory update is performed in each TCX path.
Alternatively, a full analysis by synthesis process can performed, i.e., both encoder simulators 621, 622 implement the actual encoding operations and the results are compared by the selector 623. Alternatively, again, a complete feed forward calculation can be done by performing a signal analysis. For example, when it is determined that the signal is a speech signal by a signal classifier the time domain encoder is selected and when it is determined that the signal is a music signal then the frequency domain encoder is selected. Other procedures in order to distinguish between both encoders based on a signal analysis of the audio signal portion under consideration can also be applied.
Advantageously, the audio encoder additionally comprises a cross-processor 700 illustrated in
Hence, the time domain encoder 610 is configured to be initialized by the initialization data in order to encode an audio signal portion following an earlier audio signal portion encoded by the frequency domain encoder 600 in an efficient manner.
In particular, the cross-processor comprises a time converter for converting a frequency domain representation into a time domain representation which can be forwarded to the time domain encoder directly or after some further processing. This converter is illustrated in
The ratio of the time domain coder sampling rate or ACELP sampling rate and the frequency domain coder sampling rate or input sampling rate can be calculated and is a downsampling factor DS illustrated in
This low frequency portion of the full-band spectrum is input into a small size transform and foldout block 720, as illustrated in
Thus, a very efficient downsampling operation can be applied since the downsampling is included in the IMDCT implementation. In this context, it is emphasized that the block 702 can be implemented by an IMDCT but can also be implemented by any other transform or filterbank implementation which can be suitably sized in the actual transform kernel and other transform related operations.
In a further embodiment illustrated in
Furthermore, the frequency domain encoder may comprise a noise shaping block 606a. The noise shaping block 606a is controlled by quantized LPC coefficients as generated by block 1010. The quantized LPC coefficients used for noise shaping 606a perform a spectral shaping of the high resolution spectral values or spectral lines directly encoded (rather than parametrically encoded) and the result of block 606a is similar to the spectrum of a signal subsequent to an LPC filtering stage operating in the time domain such as an LPC analysis filtering block 704 to be described later on. Furthermore, the result of the noise shaping block 606a is then quantized and entropy coded as indicated by block 606b. The result of block 606b corresponds to the encoded first audio signal portion or a frequency domain coded audio signal portion (together with other side information).
The cross-processor 700 comprises a spectral decoder for calculating a decoded version of the first encoded signal portion. In the embodiment of
Furthermore, the cross-processor 17 may comprise in addition or alternatively a weighted prediction coefficient analysis filtering stage 708 for filtering the decoded version and for feeding a filtered decoded version to a codebook determinator 613 indicated as “MMSE” in
The time domain encoder processor 610 comprises, as illustrated in
Furthermore, an ACELP gains/coding stage 612 is provided in series to the innovative codebook stage 614 and the result of this block is input into a codebook determinator 613 indicated as MMSE in
As illustrated, several blocks of the time domain decoder depend on previous signals and these blocks are the adaptive codebook block, the codebook determinator 613, the LPC synthesis filtering block 616 and the de-emphasis block 617. These blocks are provided with data from the cross-processor derived from the frequency domain encoding processor data in order to initialize these blocks for the purpose of being ready for an instant switch from the frequency domain encoder to the time domain encoder. As can also be seen from
An embodiment of an audio encoder therefore comprises the following parts:
The audio decoder is described in the following: The waveform decoder part consists of a full-band TCX decoder path with IGF both operating at the input sampling rate of the codec. In parallel, an alternative ACELP decoder path at lower sampling rate exists that is reinforced further downstream by a TD-BWE.
For ACELP initialization when switching from TCX to ACELP, a cross path (consisting of a shared TCX decoder frontend but additionally providing output at the lower sampling rate and some post-processing) exists that performs the inventive ACELP initialization. Sharing the same sampling rate and filter order between TCX and ACELP in the LPCs allows for an easier and more efficient ACELP initialization.
For visualizing the switching, two switches are sketched in 14B. While the second switch downstream chooses between TCX/IGF or ACELP/TD-BWE output, the first switch either pre-updates the buffers in the resampling QMF stage downstream the ACELP path by the output of the cross path or simply passes on the ACELP output.
Subsequently, audio decoder implementations in accordance with aspects of the present invention are discussed in the context of
An audio decoder for decoding an encoded audio signal 1101 comprises a first decoding processor 1120 for decoding a first encoded audio signal portion in a frequency domain. The first decoding processor 1120 comprises a spectral decoder 1122 for decoding first spectral regions with a high spectral resolution and for synthesizing second spectral regions using a parametric representation of the second spectral regions and at least a decoded first spectral region to obtain a decoded spectral representation. The decoded spectral representation is a full-band decoded spectral representation as discussed in the context of
Furthermore, the audio decoder comprises a second decoding processor 1140 for decoding the second encoded audio signal portion in the time domain to obtain a decoded second signal portion. Furthermore, the audio decoder comprises a combiner 1160 for combining the decoded first signal portion and the decoded second signal portion to obtain a decoded audio signal. The decoded signal portions are combined in sequence which is also illustrated in
Advantageously, the second decoding processor 1140 is a time domain bandwidth extension processor and comprises, as illustrated in
Subsequently, an implementation of the upsampler 1210 of
Further processing operations can be performed within the QMF domain in addition or instead of the bandpass filtering 1472. If no processing is performed at all, then the QMF analysis and the QMF synthesis constitute an efficient upsampler 1210.
Subsequently, the construction of the individual elements in
The full-band frequency domain decoder 1120 comprises a first decoding block 1122a for decoding the high resolution spectral coefficients and for additionally performing noise filling in the low band portion as known, for example, from the USAC technology. Furthermore, the full-band decoder comprises an IGF processor 1122b for filling the spectral holes using synthesized spectral values which have been only parametrically and, therefore, encoded with a low resolution on the encoder-side. Then, in block 1122c, an inverse noise shaping is performed and the result is input into a TNS/TTS synthesis block 705 which provides, as a final output, an input to a frequency-time converter 1124, which may be implemented as an inverse modified discrete cosine transform operating at the output, i.e., high sampling rate.
Furthermore, a harmonic or LTP post-filter is used which is controlled by data obtained by the TCX LTP parameter extraction block 1006 in
Several elements in
The time domain decoding processor 1140 may comprise the ACELP or time domain low band decoder 1200 comprising an ACELP decoder stage 1149 for obtaining decoded gains and the innovative codebook information. Additionally, an ACELP adaptive codebook stage 1141 is provided and a subsequent ACELP post-processing stage 1142 and a final synthesis filter such as LPC synthesis filter 1143, which is again controlled by the quantized LPC coefficients 1145 obtained from the bitstream demultiplexer 1100 corresponding to the encoded signal parser 1100 in
In accordance with embodiments of the present invention, the audio decoder additionally comprises the cross-processor 1170 illustrated in
Advantageously, the cross-processor 1170 comprises an additional frequency-time converter 1171 operating at a lower sampling rate than the frequency-time converter of the first decoding processor in order to obtain a further decoded first signal portion in the time domain to be used as the initialization signal or for which any initialization data can be derived. Advantageously, this IMDCT or low sampling rate frequency-time converter is implemented as illustrated in
As illustrated in
Furthermore, the cross-processor may comprise alternatively or in addition to the other mentioned elements an LPC analysis filter 1174 for generating a prediction residual signal from the further decoded first signal portion or a pre-emphasized further decoded first signal portion and for feeding the data into a codebook synthesizer of the second decoding processor and advantageously, into the adaptive codebook stage 1141. Furthermore, the output of the frequency-time converter 1171 with the low sampling rate is also input into the QMF analysis stage 1471 of the upsampler 1210 for the purpose of initialization, i.e., when the currently decoded audio signal portion is delivered by the frequency domain full-band decoder 1120.
The audio decoder is described in the following: The waveform decoder part consists of a full-band TCX decoder path with IGF both operating at the input sampling rate of the codec. In parallel, an alternative ACELP decoder path at lower sampling rate exists that is reinforced further downstream by a TD-BWE.
For ACELP initialization when switching from TCX to ACELP, a cross path (consisting of a shared TCX decoder frontend but additionally providing output at the lower sampling rate and some post-processing) exists that performs the inventive ACELP initialization. Sharing the same sampling rate and filter order between TCX and ACELP in the LPCs allows for an easier and more efficient ACELP initialization.
For visualizing the switching, two switches are sketched in
To summarize, advantageous aspects of the invention which can be used alone or in combination relate to a combination of an ACELP and TD-BWE coder with a full-band capable TCX/IGF technology advantageously associated with using a cross signal.
A further specific feature is a cross signal path for the ACELP initialization to enable seamless switching.
A further aspect is that a short IMDCT is fed with a lower part of high-rate long MDCT coefficients to efficiently implement a sample rate conversion in the cross-path.
A further feature is an efficient realization of the cross-path partly shared with a full-band TCX/IGF in the decoder.
A further feature is the cross signal path for the QMF initialization to enable seamless switching from TCX to ACELP.
An additional feature is a cross-signal path to the QMF allowing compensating the delay gap between ACELP resampled output and a filterbank-TCX/IGF output when switching from ACELP to TCX.
A further aspect is that an LPC is provided for both the TCX and the ACELP coder at the same sampling rate and filter order, although the TCX/IGF encoder/decoder is full-band capable.
Subsequently,
Generally, the time domain decoder comprises an ACELP decoder, a subsequently connected resampler or upsampler and a time domain bandwidth extension functionality. Particularly, the ACELP decoder comprises an ACELP decoding stage for restoring gains and the innovative codebook 1149, an ACELP-adaptive codebook stage 1141, an ACELP post-processor 1142, an LPC synthesis filter 1143 controlled by quantized LPC coefficients from a bitstream demultiplexer or encoded signal parser and the subsequently connected de-emphasis stage 1144. Advantageously, the time domain residual signal being at an ACELP sampling rate is input into a time domain bandwidth extension decoder 1220 which provides a high band at the outputs.
In order to upsample the de-emphasis 1144 output, an upsampler comprising the QMF analysis block 1471, and the QMF synthesis block 1473 are provided. Within the filterbank domain defined by blocks 1471 and 1473, a bandpass filter may be applied. Particularly, as has been discussed before, the same functionalities can also be used which have been discussed with respect to the same reference numbers. Furthermore, the time domain bandwidth extension decoder 1220 can be implemented as illustrated in
Subsequently, further details with respect to the frequency domain encoder and decoder being full-band capable are discussed with respect to
Typically, a first spectral portion such as 306 of
The decoder further comprises a frequency regenerator 116 for regenerating a reconstructed second spectral portion having the first spectral resolution using a first spectral portion. The frequency regenerator 116 performs a tile filling operation, i.e., uses a tile or portion of the first set of first spectral portions and copies this first set of first spectral portions into the reconstruction range or reconstruction band having the second spectral portion and typically performs spectral envelope shaping or another operation as indicated by the decoded second representation output by the parametric decoder 114, i.e., by using the information on the second set of second spectral portions. The decoded first set of first spectral portions and the reconstructed second set of spectral portions as indicated at the output of the frequency regenerator 116 on line 117 is input into a spectrum-time converter 118 configured for converting the first decoded representation and the reconstructed second spectral portion into a time representation 119, the time representation having a certain high sampling rate.
The spectral analyzer/tonal mask 226 separates the output of TNS block 222 into the core band and the tonal components corresponding to the first set of first spectral portions 103 and the residual components corresponding to the second set of second spectral portions 105 of
Advantageously, the analysis filterbank 222 is implemented as an MDCT (modified discrete cosine transform filterbank) and the MDCT is used to transform the signal 99 into a time-frequency domain with the modified discrete cosine transform acting as the frequency analysis tool.
The spectral analyzer 226 may apply a tonality mask. This tonality mask estimation stage is used to separate tonal components from the noise-like components in the signal. This allows the core coder 228 to code all tonal components with a psycho-acoustic module. The tonality mask estimation stage can be implemented in numerous different ways and may be implemented similar in its functionality to the sinusoidal track estimation stage used in sine and noise-modeling for speech/audio coding [8, 9] or an HILN model based audio coder described in [10]. Advantageously, an implementation is used which is easy to implement without the need to maintain birth-death trajectories, but any other tonality or noise detector can be used as well.
The IGF module calculates the similarity that exists between a source region and a target region. The target region will be represented by the spectrum from the source region. The measure of similarity between the source and target regions is done using a cross-correlation approach. The target region is split into nTar non-overlapping frequency tiles. For every tile in the target region, nSrc source tiles are created from a fixed start frequency. These source tiles overlap by a factor between 0 and 1, where 0 means 0% overlap and 1 means 100% overlap. Each of these source tiles is correlated with the target tile at various lags to find the source tile that best matches the target tile. The best matching tile number is stored in tileNum[idx_tar], the lag at which it best correlates with the target is stored in xcorr_lag[idx_tar][idx_src] and the sign of the correlation is stored in xcorr_sign[ids_tar][idx_src]. In case the correlation is highly negative, the source tile needs to be multiplied by −1 before the tile filling process at the decoder. The IGF module also takes care of not overwriting the tonal components in the spectrum since the tonal components are preserved using the tonality mask. A band-wise energy parameter is used to store the energy of the target region enabling us to reconstruct the spectrum accurately.
This method has certain advantages over the classical SBR [1] in that the harmonic grid of a multi-tone signal is preserved by the core coder while only the gaps between the sinusoids is filled with the best matching “shaped noise” from the source region. Another advantage of this system compared to ASR (Accurate Spectral Replacement) [2-4] is the absence of a signal synthesis stage which creates the important portions of the signal at the decoder. Instead, this task is taken over by the core coder, enabling the preservation of important components of the spectrum. Another advantage of the proposed system is the continuous scalability that the features offer. Just using tileNum[idx_tar] and xcorr_lag=0, for every tile is called gross granularity matching and can be used for low bitrates while using variable xcorr_lag for every tile enables us to match the target and source spectra better.
In addition, a tile choice stabilization technique is proposed which removes frequency domain artifacts such as trilling and musical noise.
In case of stereo channel pairs an additional joint stereo processing is applied. This is done, because for a certain destination range the signal can a highly correlated panned sound source. In case the source regions chosen for this particular region are not well correlated, although the energies are matched for the destination regions, the spatial image can suffer due to the uncorrelated source regions. The encoder analyses each destination region energy band, typically performing a cross-correlation of the spectral values and if a certain threshold is exceeded, sets a joint flag for this energy band. In the decoder the left and right channel energy bands are treated individually if this joint stereo flag is not set. In case the joint stereo flag is set, both the energies and the patching are performed in the joint stereo domain. The joint stereo information for the IGF regions is signaled similar the joint stereo information for the core coding, including a flag indicating in case of prediction if the direction of the prediction is from downmix to residual or vice versa.
The energies can be calculated from the transmitted energies in the L/R-domain.
midNrg[k]=leftNrg[k]+rightNrg[k];
sidNrg[k]=leftNrg[k]−rightNrg[k];
With k being the frequency index in the transform domain.
Another solution is to calculate and transmit the energies directly in the joint stereo domain for bands where joint stereo is active, so no additional energy transformation is needed at the decoder side.
The source tiles are created according to the Mid/Side-Matrix:
midTile[k]=0.5·(leftTile[k]+rightTile[k])
sideTile [k]=0.5·(leftTile[k]−rightTile[k])
Energy adjustment:
midTile[k]=midTile[k]*midNrg[k];
sideTile[k]=sideTile[k]*sideNrg[k];
Joint stereo->LR transformation:
If no additional prediction parameter is coded:
leftTile[k]=midTile[k]+sideTile[k]
rightTile[k]=midTile[k]−sideTile[k]
If an additional prediction parameter is coded and if the signalled direction is from mid to side:
sideTile[k]=sideTile[k]−predictionCoeff·midTile[k]
leftTile[k]=midTile[k]+sideTile[k]
rightTile[k]=midTile[k]−sideTile[k]
If the signalled direction is from side to mid:
midTile1[k]=midTile[k]−predictionCoeff·sideTile[k]
leftTile[k]=midTile1[k]−sideTile[k]
rightTile[k]=midTile1[k]+sideTile[k]
This processing ensures that from the tiles used for regenerating highly correlated destination regions and panned destination regions, the resulting left and right channels still represent a correlated and panned sound source even if the source regions are not correlated, preserving the stereo image for such regions.
In other words, in the bitstream, joint stereo flags are transmitted that indicate whether L/R or M/S as an example for the general joint stereo coding shall be used. In the decoder, first, the core signal is decoded as indicated by the joint stereo flags for the core bands. Second, the core signal is stored in both L/R and M/S representation. For the IGF tile filling, the source tile representation is chosen to fit the target tile representation as indicated by the joint stereo information for the IGF bands.
Temporal Noise Shaping (TNS) is a standard technique and part of AAC [11-13]. TNS can be considered as an extension of the basic scheme of a perceptual coder, inserting an optional processing step between the filterbank and the quantization stage. The main task of the TNS module is to hide the produced quantization noise in the temporal masking region of transient like signals and thus it leads to a more efficient coding scheme. First, TNS calculates a set of prediction coefficients using “forward prediction” in the transform domain, e.g. MDCT. These coefficients are then used for flattening the temporal envelope of the signal. As the quantization affects the TNS filtered spectrum, also the quantization noise is temporarily flat. By applying the invers TNS filtering on decoder side, the quantization noise is shaped according to the temporal envelope of the TNS filter and therefore the quantization noise gets masked by the transient.
IGF is based on an MDCT representation. For efficient coding, advantageously long blocks of approx. 20 ms have to be used. If the signal within such a long block contains transients, audible pre- and post-echoes occur in the IGF spectral bands due to the tile filling.
This pre-echo effect is reduced by using TNS in the IGF context. Here, TNS is used as a temporal tile shaping (TTS) tool as the spectral regeneration in the decoder is performed on the TNS residual signal. The necessitated TTS prediction coefficients are calculated and applied using the full spectrum on encoder side as usual. The TNS/TTS start and stop frequencies are not affected by the IGF start frequency figfstart of the IGF tool. In comparison to the legacy TNS, the TTS stop frequency is increased to the stop frequency of the IGF tool, which is higher figfstart. On decoder side the TNS/TTS coefficients are applied on the full spectrum again, i.e. the core spectrum plus the regenerated spectrum plus the tonal components from the tonality map (see
In legacy decoders, spectral patching on an audio signal corrupts spectral correlation at the patch borders and thereby impairs the temporal envelope of the audio signal by introducing dispersion. Hence, another benefit of performing the IGF tile filling on the residual signal is that, after application of the shaping filter, tile borders are seamlessly correlated, resulting in a more faithful temporal reproduction of the signal.
In an inventive encoder, the spectrum having undergone TNS/TTS filtering, tonality mask processing and IGF parameter estimation is devoid of any signal above the IGF start frequency except for tonal components. This sparse spectrum is now coded by the core coder using principles of arithmetic coding and predictive coding. These coded components along with the signaling bits form the bitstream of the audio.
Advantageously, the high resolution is defined by a line-wise coding of spectral lines such as MDCT lines, while the second resolution or low resolution is defined by, for example, calculating only a single spectral value per scale factor band, where a scale factor band covers several frequency lines. Thus, the second low resolution is, with respect to its spectral resolution, much lower than the first or high resolution defined by the line-wise coding typically applied by the core encoder such as an AAC or USAC core encoder.
Regarding scale factor or energy calculation, the situation is illustrated in
Particularly, when the core encoder is under a low bitrate condition, an additional noise-filling operation in the core band, i.e., lower in frequency than the IGF start frequency, i.e., in scale factor bands SCB1 to SCB3 can be applied in addition. In noise-filling, there exist several adjacent spectral lines which have been quantized to zero. On the decoder-side, these quantized to zero spectral values are re-synthesized and the re-synthesized spectral values are adjusted in their magnitude using a noise-filling energy such as NF2 illustrated at 308 in
Advantageously, the bands, for which energy information is calculated coincide with the scale factor bands. In other embodiments, an energy information value grouping is applied so that, for example, for scale factor bands 4 and 5, only a single energy information value is transmitted, but even in this embodiment, the borders of the grouped reconstruction bands coincide with borders of the scale factor bands. If different band separations are applied, then certain re-calculations or synchronization calculations may be applied, and this can make sense depending on the certain implementation.
Advantageously, the spectral domain encoder 106 of
In the audio encoder of
Then, at the output of block 422, a quantized spectrum is obtained corresponding to what is illustrated in
The set to zero blocks 410, 418, 422, which are provided alternatively to each other or in parallel are controlled by the spectral analyzer 424. The spectral analyzer may comprise any implementation of a well-known tonality detector or comprises any different kind of detector operative for separating a spectrum into components to be encoded with a high resolution and components to be encoded with a low resolution. Other such algorithms implemented in the spectral analyzer can be a voice activity detector, a noise detector, a speech detector or any other detector deciding, depending on spectral information or associated metadata on the resolution requirements for different spectral portions.
Subsequently, reference is made to
As illustrated at 301 in
Advantageously, an IGF operation, i.e., a frequency tile filling operation using spectral values from other portions can be applied in the complete spectrum. Thus, a spectral tile filling operation can not only be applied in the high band above an IGF start frequency but can also be applied in the low band. Furthermore, the noise-filling without frequency tile filling can also be applied not only below the IGF start frequency but also above the IGF start frequency. It has, however, been found that high quality and high efficient audio encoding can be obtained when the noise-filling operation is limited to the frequency range below the IGF start frequency and when the frequency tile filling operation is restricted to the frequency range above the IGF start frequency as illustrated in
Advantageously, the target tiles (TT) (having frequencies greater than the IGF start frequency) are bound to scale factor band borders of the full rate coder. Source tiles (ST), from which information is taken, i.e., for frequencies lower than the IGF start frequency are not bound by scale factor band borders. The size of the ST should correspond to the size of the associated TT. This is illustrated using the following example. TT[0] has a length of 10 MDCT Bins. This exactly corresponds to the length of two subsequent SCBs (such as 4+6). Then, all possible ST that are to be correlated with TT[0], have a length of 10 bins, too. A second target tile TT[1] being adjacent to TT[0] has a length of 15 bins I (SCB having a length of 7+8). Then, the ST for that have a length of 15 bins rather than 10 bins as for TT[0].
Should the case arise that one cannot find a TT for an ST with the length of the target tile (when e.g. the length of TT is greater than the available source range), then a correlation is not calculated and the source range is copied a number of times into this TT (the copying is done one after the other so that a frequency line for the lowest frequency of the second copy immediately follows—in frequency—the frequency line for the highest frequency of the first copy), until the target tile TT is completely filled up.
Subsequently, reference is made to
Then, the first spectral portion of the reconstruction band such as 307 of
In this context, it is very important to evaluate the high frequency reconstruction accuracy of the present invention compared to HE-AAC. This is explained with respect to scale factor band 7 in
In an implementation, the spectral analyzer is also implemented to calculating similarities between first spectral portions and second spectral portions and to determine, based on the calculated similarities, for a second spectral portion in a reconstruction range a first spectral portion matching with the second spectral portion as far as possible. Then, in this variable source range/destination range implementation, the parametric coder will additionally introduce into the second encoded representation a matching information indicating for each destination range a matching source range. On the decoder-side, this information would then be used by a frequency tile generator 522 of
Furthermore, as illustrated in
As illustrated, the encoder operates without downsampling and the decoder operates without upsampling. In other words, the spectral domain audio coder is configured to generate a spectral representation having a Nyquist frequency defined by the sampling rate of the originally input audio signal.
Furthermore, as illustrated in
As outlined, the spectral domain audio decoder 112 is configured so that a maximum frequency represented by a spectral value in the first decoded representation is equal to a maximum frequency included in the time representation having the sampling rate wherein the spectral value for the maximum frequency in the first set of first spectral portions is zero or different from zero. Anyway, for this maximum frequency in the first set of spectral components a scale factor for the scale factor band exists, which is generated and transmitted irrespective of whether all spectral values in this scale factor band are set to zero or not as discussed in the context of
The invention is, therefore, advantageous that with respect to other parametric techniques to increase compression efficiency, e.g. noise substitution and noise filling (these techniques are exclusively for efficient representation of noise like local signal content) the invention allows an accurate frequency reproduction of tonal components. To date, no state-of-the-art technique addresses the efficient parametric representation of arbitrary signal content by spectral gap filling without the restriction of a fixed a-priory division in low band (LF) and high band (HF).
Embodiments of the inventive system improve the state-of-the-art approaches and thereby provides high compression efficiency, no or only a small perceptual annoyance and full audio bandwidth even for low bitrates.
The general system consists of
A first step towards a more efficient system is to remove the need for transforming spectral data into a second transform domain different from the one of the core coder. As the majority of audio codecs, such as AAC for instance, use the MDCT as basic transform, it is useful to perform the BWE in the MDCT domain also. A second requirement for the BWE system would be the need to preserve the tonal grid whereby even HF tonal components are preserved and the quality of the coded audio is thus superior to the existing systems. To take care of both the above mentioned requirements for a BWE scheme, a new system is proposed called Intelligent Gap Filling (IGF).
Subsequently, further optional features of the full band frequency domain first encoding processor and the full band frequency domain decoding processor incorporating the gap-filling operation, which can be implemented separately or together are discussed and defined.
Particularly, the spectral domain decoder 112 corresponding to block 1122a is configured to output a sequence of decoded frames of spectral values, a decoded frame being the first decoded representation, wherein the frame comprises spectral values for the first set of spectral portions and zero indications for the second spectral portions. The apparatus for decoding furthermore comprises a combiner 208. The spectral values are generated by a frequency regenerator for the second set of second spectral portions, where both, the combiner and the frequency regenerator are included within block 1122b. Thus, by combining the second spectral portions and the first spectral portions a reconstructed spectral frame comprising spectral values for the first set of the first spectral portions and the second set of spectral portions are obtained and the spectrum-time converter 118 corresponding to the IMDCT block 1124 in
As outlined, the spectrum-time converter 118 or 1124 is configured to perform an inverse modified discrete cosine transform 512, 514 and further comprises an overlap-add stage 516 for overlapping and adding subsequent time domain frames.
Particularly, the spectral domain audio decoder 1122a is configured to generate the first decoded representation so that the first decoded representation has a Nyquist frequency defining a sampling rate being equal to a sampling rate of the time representation generated by the spectrum-time converter 1124.
Furthermore, the decoder 1112 or 1122a is configured to generate the first decoded representation so that a first spectral portion 306 is placed with respect to frequency between two second spectral portions 307a, 307b.
In a further embodiment, a maximum frequency represented by a spectral value for the maximum frequency in the first decoded representation is equal to a maximum frequency included in the time representation generated by the spectrum-time converter, wherein the spectral value for the maximum frequency in the first representation is zero or different from zero.
Furthermore, as illustrated in
Furthermore, the spectral domain audio decoder 112 is configured to generate the first decoded representation having the first spectral portions with the frequency values being greater than the frequency being equal to a frequency in the middle of the frequency range covered by the time representation output by the spectrum-time converter 118 or 1124.
Furthermore, the spectral analyzer or full-band analyzer 604 is configured to analyze the representation generated by the time-frequency converter 602 for determining a first set of first spectral portions to be encoded with the first high spectral resolution and the different second set of second spectral portions to be encoded with a second spectral resolution which is lower than the first spectral resolution and, by means of the spectral analyzer, a first spectral portion 306 is determined, with respect to frequency, between two second spectral portions in
Particularly, the spectral analyzer is configured for analyzing the spectral representation up to a maximum analysis frequency being at least one quarter of a sampling frequency of the audio signal.
Particularly, the spectral domain audio encoder is configured to process a sequence of frames of spectral values for a quantization and entropy coding, wherein, in a frame, spectral values of the second set of second portions are set to zero, or wherein, in the frame, spectral values of the first set of first spectral portions and the second set of the second spectral portions are present and wherein, during subsequent processing, spectral values in the second set of spectral portions are set to zero as exemplarily illustrated at 410, 418, 422.
The spectral domain audio encoder is configured to generate a spectral representation having a Nyquist frequency defined by the sampling rate of the audio input signal or the first portion of the audio signal processed by the first encoding processor operating in the frequency domain.
The spectral domain audio encoder 606 is furthermore configured to provide the first encoded representation so that, for a frame of a sampled audio signal, the encoded representation comprises the first set of first spectral portions and the second set of second spectral portions, wherein the spectral values in the second set of spectral portions are encoded as zero or noise values.
The full band analyzer 604 or 102 is configured to analyze the spectral representation starting with the gap-filing start frequency 209 and ending with a maximum frequency fmax represented by a maximum frequency included in the spectral representation and a spectral portion extending from a minimum frequency up to the gap-filling start frequency 309 belongs to the first set of first spectral portions.
Particularly, the analyzer is configured to apply a tonal mask processing at least of a portion of the spectral representation so that tonal components and non-tonal components are separated from each other, wherein the first set of the first spectral portions comprises the tonal components and wherein the second set of the second spectral portions comprises the non-tonal components.
Although the present invention has been described in the context of block diagrams where the blocks represent actual or logical hardware components, the present invention can also be implemented by a computer-implemented method. In the latter case, the blocks represent corresponding method steps where these steps stand for the functionalities performed by corresponding logical or physical hardware blocks.
Although some aspects have been described in the context of an apparatus, it is clear that these aspects also represent a description of the corresponding method, where a block or device corresponds to a method step or a feature of a method step. Analogously, aspects described in the context of a method step also represent a description of a corresponding block or item or feature of a corresponding apparatus. Some or all of the method steps may be executed by (or using) a hardware apparatus, like for example, a microprocessor, a programmable computer or an electronic circuit. In some embodiments, some one or more of the most important method steps may be executed by such an apparatus.
The inventive transmitted or encoded signal can be stored on a digital storage medium or can be transmitted on a transmission medium such as a wireless transmission medium or a wired transmission medium such as the Internet.
Depending on certain implementation requirements, embodiments of the invention can be implemented in hardware or in software. The implementation can be performed using a digital storage medium, for example a floppy disc, a DVD, a Blu-Ray, a CD, a ROM, a PROM, and EPROM, an EEPROM or a FLASH memory, having electronically readable control signals stored thereon, which cooperate (or are capable of cooperating) with a programmable computer system such that the respective method is performed. Therefore, the digital storage medium may be computer readable.
Some embodiments according to the invention comprise a data carrier having electronically readable control signals, which are capable of cooperating with a programmable computer system, such that one of the methods described herein is performed.
Generally, embodiments of the present invention can be implemented as a computer program product with a program code, the program code being operative for performing one of the methods when the computer program product runs on a computer. The program code may, for example, be stored on a machine readable carrier.
Other embodiments comprise the computer program for performing one of the methods described herein, stored on a machine readable carrier.
In other words, an embodiment of the inventive method is, therefore, a computer program having a program code for performing one of the methods described herein, when the computer program runs on a computer.
A further embodiment of the inventive method is, therefore, a data carrier (or a non-transitory storage medium such as a digital storage medium, or a computer-readable medium) comprising, recorded thereon, the computer program for performing one of the methods described herein. The data carrier, the digital storage medium or the recorded medium are typically tangible and/or non-transitory.
A further embodiment of the invention method is, therefore, a data stream or a sequence of signals representing the computer program for performing one of the methods described herein. The data stream or the sequence of signals may, for example, be configured to be transferred via a data communication connection, for example, via the internet.
A further embodiment comprises a processing means, for example, a computer or a programmable logic device, configured to, or adapted to, perform one of the methods described herein.
A further embodiment comprises a computer having installed thereon the computer program for performing one of the methods described herein.
A further embodiment according to the invention comprises an apparatus or a system configured to transfer (for example, electronically or optically) a computer program for performing one of the methods described herein to a receiver. The receiver may, for example, be a computer, a mobile device, a memory device or the like. The apparatus or system may, for example, comprise a file server for transferring the computer program to the receiver.
In some embodiments, a programmable logic device (for example, a field programmable gate array) may be used to perform some or all of the functionalities of the methods described herein. In some embodiments, a field programmable gate array may cooperate with a microprocessor in order to perform one of the methods described herein. Generally, the methods may be performed by any hardware apparatus.
While this invention has been described in terms of several embodiments, there are alterations, permutations, and equivalents which will be apparent to others skilled in the art and which fall within the scope of this invention. It should also be noted that there are many alternative ways of implementing the methods and compositions of the present invention. It is therefore intended that the following appended claims be interpreted as including all such alterations, permutations, and equivalents as fall within the true spirit and scope of the present invention.
| Number | Date | Country | Kind |
|---|---|---|---|
| 14178817.4 | Jul 2014 | EP | regional |
This application is a continuation of co-pending U.S. application Ser. No. 18/158,035 filed Jan. 23, 2023 which is a continuation of co-pending U.S. application Ser. No. 17/336,132 filed Jun. 1, 2021 which is a continuation of co-pending U.S. application Ser. No. 16/286,397 filed Feb. 26, 2019 which is a continuation of co-pending U.S. application Ser. No. 15/414,427 filed Jan. 24, 2017 which is now U.S. Pat. No. 10,332,535 issued Jun. 25 2019; which is a continuation of International Application No. PCT/EP2015/067003, filed Jul. 24, 2015 and additionally claims priority from European Application No. 14178817.4, filed Jul. 28, 2014, of which all the foregoing applications and patents are incorporated herein by reference in their entirety.
| Number | Date | Country | |
|---|---|---|---|
| Parent | 18158035 | Jan 2023 | US |
| Child | 18455738 | US | |
| Parent | 17336132 | Jun 2021 | US |
| Child | 18158035 | US | |
| Parent | 16286397 | Feb 2019 | US |
| Child | 17336132 | US | |
| Parent | 15414427 | Jan 2017 | US |
| Child | 16286397 | US |
