DECODING DEVICE AND METHOD, AND, COMPUTER-READABLE STORAGE MEANS

专利摘要:
device and method of encoding, device and method of decoding, and, program. the present invention relates to an encoding device and method, a decoding device and method, and a program with which a music signal can be reproduced with higher sound quality by expanding the frequency band. a bandpass filter divides an input signal into a plurality of subband signals. a resource quantity calculation circuit calculates the resource quantities using the plurality of subband signals obtained by said division and/or the input signal. a high-range subband energy inference circuit calculates an inferred value for the high-range subband energy by using the calculated resource quantities. a high-range signal generating circuit generates a high-range signal component by using the plurality of subband signals obtained by the division performed by the bandpass filter and the inferred value for the calculated high-range subband energy by the high-range subband energy inference circuit. a frequency band expansion device expands the frequency band of the input signal by using the high-range signal component generated by the high-range signal generating circuit. the present invention can be applied, for example, in a frequency band expansion device, in an encoding device, in a decoding device and the like.
公开号:BR112013008490B1
申请号:R112013008490-1
申请日:2011-10-05
公开日:2021-06-22
发明作者:Yuki Yamamoto；Toru Chinen
申请人:Sony Corporation；
IPC主号:

专利说明:

FIELD OF THE INVENTION
[001] The present invention relates to an encoding device and method, a decoding device and method and a program, and specifically concerns an encoding device and method, a decoding device and method and the a program that enables music signals to be reproduced with high sound quality by expanding a frequency band. FUNDAMENTALS OF THE INVENTION
[002] In recent years, music distribution service for distributing music data via the Internet or the like has become widespread. With this music distribution service, encoded data obtained by encoding music signals is distributed as music data. As a music signal encoding technique, the trend has become an encoding technique where a bit rate is lowered, yet suppressing the archival capacity of the encoded data so as not to take time at the time of transfer.
[003] Such music signal encoding techniques are roughly divided into an encoding technique such as MP3 (MPEG (Moving Image Specialists Group) Audio Layer 3) (International Standards ISO / IEC 11172-3) and congeners, and an encoding technique such as HE-AAC (High Efficiency MPEG4 AAC) (International Standards ISO / IEC 14496-3) and so on.
[004] With the encoding technique, represented by MP3, of music signals, signal components in a high frequency band (hereinafter, referred to as high frequency) equal to or greater than about 15 kHz, hardly perceived by the human ear , are erased and signal components in the remaining low frequency band (hereinafter, referred to as low frequency) are encoded. Such an encoding technique will be referred to as high frequency deletion encoding technique. With this high frequency deletion encoding technique, the archival capacity of encoded data can be suppressed. However, high-frequency sound may be slightly perceived by the human ear and, therefore, at the time of sound generation and transmission of music signals after decoding obtained by decoding the encoded data, there may be deterioration in sound quality, such as loss of the sense of presence that the original sound has, or the sound may appear to be muffled.
[005] On the other hand, with the encoding technique represented by HE-AAC, characteristic information is extracted from the high frequency signal components and encoded together with the low frequency signal components. In the following, such an encoding technique will be referred to as a high-frequency characteristic encoding technique. With this high-frequency characteristic encoding technique, only characteristic information of the components of the high-frequency signal is encoded as information regarding the components of the high-frequency signal and, in this way, encoding efficiency can be increased, still suppressing the deterioration in sound quality.
[006] With the decoding of encoded data encoded by this high-frequency characteristic encoding technique, low-frequency signal components and characteristic information are decoded, and high-frequency signal components are generated from the signal components frequency and feature information after decoding. Thus, a technique for expanding the frequency band of the low-frequency signal components by generating the high-frequency signal components from the low-frequency signal components will be referred to below as a band-expanding technique.
[007] As an application of the band expansion technique, there is post-processing after decoding the encoded data by the aforementioned high frequency deletion encoding technique. With this post-processing, high-frequency signal components lost by encoding are generated from the low-frequency signal components after decoding, thereby expanding the frequency band of the low-frequency signal components (see PTL 1) . Note that the frequency band expansion technique according to PTL 1 will be referred to below as the band expansion technique according to PTL 1.
[008] With the band expansion technique according to PTL 1, a device takes components of the low-frequency signal after decoding as an input signal, estimates high-frequency energy spectrum (hereinafter, referred to as frequency enveloping as appropriate) from the energy spectrum of the input signals and generates high-frequency signal components with high-frequency frequency enveloping from the low-frequency signal components.
[009] Figure 1 illustrates an example of the low-frequency energy spectrum after decoding, which serves as the input signal, and the estimated high-frequency frequency enveloping.
[0010] In figure 1, the vertical geometric axis indicates energy by a logarithm and the horizontal geometric axis indicates frequencies.
[0011] The device determines the low-frequency edge band of the high-frequency signal components (hereinafter referred to as the expansion start band) from the type information of an encoding method with respect to the input signal, the sample rate, the bit rate, and so on (hereafter referred to as overhead). Next, the device splits the input signal that serves as low frequency signal components into multiple subband signals. The device averages each group over a temporal direction of the energy (hereinafter referred to as the group energy) of each of the multiple subband signals following the division, i.e., the multiple subband signals on the lower frequency side with respect to the expansion start band (hereinafter, referred to simply as the low frequency side). As illustrated in Figure 1, the device takes a point with the average energy of the group of each of the multiple subband signals on the low frequency side as energy and also the frequency of the lower end of the start band of expansion as frequency as origin. The device performs estimation with a primary straight line with a predetermined slope that passes through the origin of the latter as a frequency envelope on the side of the frequency higher than the expansion start band (hereafter referred to simply as the high frequency side). Note that a position in relation to the source energy direction can be adjusted by a user. The device generates each of the multiple subband signals on the high frequency side from the multiple subband signals on the low frequency side to obtain the estimated frequency envelope on the high frequency side. The device adds the multiple subband signals generated on the high frequency side to obtain high frequency signal components, further adds the low frequency signal components and transmits them. Thus, music signals after frequency band expansion approach the original music signals. In this way, music signals with high sound quality can be reproduced.
[0012] The aforementioned band expansion technique according to PTL 1 has a feature that, in relation to the various coding techniques with high frequency deletion and encoded data with various bit rates, the frequency band with respect to the signals of music after decoding the encoded data from these can be expanded. CITATION LITERATURE PTL PATENT 1: Unexamined Japanese Patent Application Publication 2008-139844 SUMMARY OF THE INVENTION TECHNICAL ISSUES
[0013] However, with the band expansion technique according to PTL 1, there is room for improvement, in which the estimated frequency envelope on the high frequency side becomes a primary straight line with predetermined slope, that is, in that the shape of the frequency envelope is fixed.
[0014] Specifically, the energy spectra of music signals have various shapes, and there may be many cases to greatly deviate from the frequency envelope on the high frequency side estimated by the band expansion technique according to PTL 1, depending on the types of music signs.
[0015] Figure 2 illustrates an example of the original energy spectrum of a music signal of an attack nature (attack music signal) that accompanies rapid temporal change, such as a strong drum beat.
[0016] Note that Figure 2 also illustrates the frequency envelope on the high-frequency side estimated by the band expansion technique according to PTL 1 from the signal components on the low-frequency side of a music signal with attack that serves as an input signal.
[0017] As illustrated in Figure 2, the original energy spectrum on the high frequency side of the music signal with attack is generally flat.
[0018] On the other hand, the estimated frequency envelope on the high frequency side has a predetermined negative slope and, in this way, even when the energy adjustment at the origin approaches the original energy spectrum, as the frequency increases, the difference with the original energy spectrum increases.
[0019] Thus, with the band expansion technique according to PTL 1, according to the estimated frequency envelope on the high frequency side, the original frequency envelope on the high frequency side cannot be reproduced with high accuracy. As a result, at the time of generation and transmission of sound from a music signal after the expansion of the frequency band, sound clarity was lost compared to the original sound in relation to audibility.
[0020] Also, with the aforementioned high-frequency characteristic encoding technique, such as HE-AAC or the like, although frequency enveloping on the high-frequency side is employed as characteristic information of the high-frequency signal components to be encoded , the decoding side is required to reproduce the frequency envelope on the high frequency side with high accuracy.
[0021] The present invention was made in light of such situations, and enables music signals to be reproduced with high sound quality by expanding the frequency band. SOLUTION OF THE PROBLEM
[0022] An encoding device according to a first aspect of the present invention includes: subband dividing means configured to divide an input signal into multiple subbands and to generate a low frequency subband signal, made up of multiple subbands on the low frequency side, and a high frequency subband signal made up of multiple subbands on the high frequency side; resource amount calculating means configured to calculate resource amount representing resources of the input signal based on at least any one of the low frequency subband signal and the input signal; standardization means configured to subject the amount of resource to standardization; pseudo-high-frequency sub-band energy calculation means configured to calculate the pseudo-high-frequency sub-band energy, which is an estimated value of the energy of the high-frequency sub-band signal, based on the smoothed resource amount and at a predetermined coefficient; selection means configured to calculate the high frequency subband energy, which is the energy of the high frequency subband signal, from the high frequency subband signal, and to compare the subband energy. high frequency band and the pseudo high frequency subband energy to select any one of multiple coefficients; high-frequency encoding means configured to encode coefficient information to obtain the selected coefficient and smoothing information with respect to the smoothing to generate high-frequency encoded data; low-frequency encoding means configured to encode a low-frequency signal, which is a low-frequency signal of the input signal, to generate low-frequency encoded data; and multiplexing means configured to multiplex the low frequency encoded data and the high frequency encoded data to obtain an output code sequence.
[0023] The smoothing means may subject the resource amount to smoothing by performing weighted average for the resource amount of a predetermined number of continuous frames of the input signal.
[0024] The smoothing information can be information that indicates at least one of the numerous frames used for the weighted average or the weight used for the weighted average.
[0025] The encoding device may include parameter determining means configured to determine at least one of the numerous frames used for the weighted average or the weight used for the weighted average based on the high frequency subband signal.
[0026] The coefficient can be generated by learning with the amount of resource and the high frequency subband energy obtained from a wideband supervisory signal as an explanatory variable and an explained variable.
[0027] The wideband supervisor signal may be a signal obtained by encoding a predetermined signal according to an encoding method and an encoding algorithm and by decoding the encoded predetermined signal; with the coefficient being generated by learning using the broadband supervisor signal for each of multiple different encoding methods and encoding algorithms.
[0028] An encoding method or program according to the first aspect of the present invention includes the steps of: dividing an input signal into multiple subbands and generating a low frequency subband signal consisting of multiple subbands. bands on the low frequency side, and a high frequency subband signal consisting of multiple subbands on the high frequency side; calculating the resource amount representing resources of the input signal based on at least any one of the low frequency subband signal and the input signal; subject the amount of resource to standardization; calculate the pseudo-high-frequency subband energy, which is an estimated value of the energy of the high-frequency subband signal, based on the smoothed resource amount and a predetermined coefficient; calculate the high frequency subband energy, which is the energy of the high frequency subband signal, from the high frequency subband signal, and compare the high frequency subband energy and the pseudohigh-frequency subband energy to select any one of multiple coefficients; encoding coefficient information to obtain the selected coefficient and smoothing information regarding the smoothing to generate high frequency encoded data; encoding a low-frequency signal, which is a low-frequency signal of the input signal, to generate low-frequency encoded data; and multiplexing the low frequency encoded data and the high frequency encoded data to obtain an output code sequence.
[0029] With the first aspect of the present invention, an input signal is divided into multiple subbands, a low frequency subband signal consisting of multiple subbands on the low frequency side, and a subband signal. -high frequency band, consisting of multiple subbands on the high frequency side are generated, the resource amount representing resources of the input signal is calculated based on at least any one of the low frequency subband signal and of the input signal, the resource amount is subject to smoothing, the pseudohigh frequency subband energy, which is an estimated value of the high frequency subband signal energy, is calculated based on the smoothed resource amount and at a predetermined coefficient, the high frequency subband energy, which is the energy of the high frequency subband signal, is calculated from the high frequency subband signal, the subband energy high frequency and the Pseudo-high frequency subband energy are compared to select any one of multiple coefficients, the coefficient information to obtain the selected coefficient and the smoothing information with respect to the smoothing to generate high frequency encoded data are encoded, a low signal frequency, which is a low-frequency signal of the input signal, is encoded to generate low-frequency encoded data, and the low-frequency encoded data and high-frequency encoded data are multiplexed to obtain an output code sequence.
[0030] A decoding device according to a second aspect of the present invention includes: demultiplexing means configured to demultiplex input encoded data into low-frequency encoded data, coefficient information to obtain a coefficient, and smoothing information with respect to the smoothing ; low frequency decoding means configured to decode the low frequency encoded data to generate a low frequency signal; subband dividing means configured to divide the low frequency signal into multiple subbands to generate a low frequency subband signal for each of the subbands; resource amount calculating means configured to calculate resource amount based on the low frequency subband signals; standardization means configured to subject the resource amount to standardization based on standardization information; and generating means configured to generate a high frequency signal based on the coefficient obtained from the coefficient information, the amount of resource subjected to smoothing, and the low frequency subband signals.
[0031] The smoothing means may subject the resource amount to smoothing by performing a weighted average on the resource amount of a predetermined number of continuous frames of the low frequency signal.
[0032] The smoothing information can be information that indicates at least one of the numerous frames used for the weighted average or the weight used for the weighted average.
[0033] The generating means may include decoded high frequency subband energy calculating means configured to calculate the decoded high frequency subband energy, which is an estimated value of the subband energy that constitutes the high-frequency signal, based on the smoothed resource amount and coefficient, and high-frequency signal generation means configured to generate the high-frequency signal based on the decoded high-frequency subband energy and the sub signal -low frequency band.
[0034] The coefficient can be generated by learning with the amount of resource obtained from a wideband supervisory signal and the energy of the same subband, as a subband that constitutes the high frequency signal of the supervisory signal. broadband, as an explanatory variable and an explained variable.
[0035] The wideband supervisor signal may be a signal obtained by encoding a predetermined signal according to a predetermined encoding method and encoding algorithm and by decoding the predetermined encoded signal; with the coefficient being generated by learning using the broadband supervisor signal for each of multiple different encoding methods and encoding algorithms.
[0036] A decoding method or program according to the second aspect of the present invention includes the steps of: demultiplexing the input encoded data into low frequency encoded data, coefficient information to obtain a coefficient, and smoothing information with respect to standardization; decoding the low frequency encoded data to generate a low frequency signal; dividing the low frequency signal into multiple subbands to generate a low frequency subband signal for each of the subbands; calculate resource amount based on low frequency subband signals; subjecting the resource amount to standardization based on standardization information; and generating a high frequency signal based on the coefficient obtained from the coefficient information, the amount of resource subject to smoothing, and the low frequency subband signals.
[0037] With the second aspect of the present invention, input encoded data is demultiplexed into low-frequency encoded data, coefficient information to obtain a coefficient, and smoothing information with respect to smoothing, the low-frequency encoded data is decoded to generate a low frequency signal, the low frequency signal is divided into multiple subbands to generate a low frequency subband signal for each of the subbands, the resource amount is calculated based on the subband signals. low frequency band, the resource amount is subject to smoothing based on the smoothing information and a high frequency signal is generated based on the coefficient obtained from the coefficient information, the resource amount subject to smoothing, and the low frequency subband. ADVANTAGEOUS EFFECTS OF THE INVENTION
[0038] According to the first aspect and the second aspect of the present invention, music signals can be reproduced with higher sound quality by expanding the frequency band. BRIEF DESCRIPTION OF THE DRAWINGS
[0039] Figure 1 is a diagram illustrating an example of low-frequency energy spectrum after decoding that serves as an input signal and estimated high-frequency frequency enveloping.
[0040] Figure 2 is a diagram that illustrates an example of the original energy spectrum of a music signal with attack that accompanies rapid temporal change.
[0041] Figure 3 is a block diagram illustrating an example of a functional configuration of a frequency band expansion device according to a first embodiment of the present invention.
[0042] Figure 4 is a flowchart to describe the frequency band expansion processing by the frequency band expansion device of Figure 3.
[0043] Figure 5 is a diagram illustrating the energy spectrum of a signal to be inserted into the frequency band expansion device of Figure 3 and bandpass filter locations on the frequency axis.
[0044] Figure 6 is a diagram illustrating an example of a frequency characteristic in a vocal section and an estimated high frequency energy spectrum.
[0045] Figure 7 is a diagram illustrating an example of the energy spectrum of a signal to be inserted into the frequency band expansion device of Figure 3.
[0046] Figure 8 is a diagram that illustrates an example of the energy spectrum after the elevation of the input signal of Figure 7.
[0047] Figure 9 is a block diagram illustrating an example of a functional configuration of a coefficient learning device to perform learning of a coefficient to be used in a high frequency signal generation circuit of the band expansion device frequency of figure 3.
[0048] Figure 10 is a flowchart to describe an example of the coefficient learning processing by the coefficient learning device of figure 9.
[0049] Figure 11 is a block diagram illustrating an example of a functional configuration of an encoding device according to a second embodiment of the present invention.
[0050] Figure 12 is a flowchart to describe an example of encoding processing by the encoding device of Figure 11.
[0051] Figure 13 is a block diagram illustrating an example of a functional configuration of a decoding device according to the second embodiment of the present invention.
[0052] Figure 14 is a flowchart to describe an example of decoding processing by the decoding device of Figure 13.
[0053] Figure 15 is a block diagram illustrating an example of a functional configuration of a coefficient learning device to perform learning of a representative vector to be used in a high-frequency encoding circuit of the encoding device of figure 11 and a decoded high frequency subband energy estimation coefficient to be used in the high frequency decoding circuit of the decoding device of Fig. 13.
[0054] Figure 16 is a flowchart to describe an example of the coefficient learning processing by the coefficient learning device of Figure 15.
[0055] Figure 17 is a diagram illustrating an example of a code sequence that the encoding device of Figure 11 transmits.
[0056] Figure 18 is a block diagram illustrating an example of a functional configuration of an encoding device.
[0057] Figure 19 is a flowchart to describe encoding processing.
[0058] Figure 20 is a block diagram illustrating an example of a functional configuration of a decoding device.
[0059] Figure 21 is a flowchart to describe decoding processing.
[0060] Figure 22 is a flowchart to describe encoding processing.
[0061] Figure 23 is a flowchart to describe the decoding processing.
[0062] Figure 24 is a flowchart to describe encoding processing.
[0063] Figure 25 is a flowchart to describe encoding processing.
[0064] Figure 26 is a flowchart to describe encoding processing.
[0065] Figure 27 is a flowchart to describe encoding processing.
[0066] Figure 28 is a diagram that illustrates an example of configuration of a coefficient learning processing.
[0067] Figure 29 is a flowchart to describe the coefficient learning processing.
[0068] Figure 30 is a block diagram illustrating an example of a functional configuration of an encoding device.
[0069] Figure 31 is a flowchart to describe encoding processing.
[0070] Figure 32 is a block diagram illustrating an example of a functional configuration of a decoding device.
[0071] Figure 33 is a flowchart to describe the decoding processing.
[0072] Figure 34 is a block diagram illustrating an example of hardware configuration of a computer that performs processing in which the present invention is applied using a program. DESCRIPTION OF MODALITIES
[0073] In the following, embodiments of the present invention will be described in relation to the drawings. Note that description will be made in the following order. 1. First Mode (Case of Applying the Present Invention on the Frequency Band Expansion Device) 2. Second Mode (Case of Applying the Present Invention on the Coding Device and on the Decoding Device) 3. Third Mode (Case of Inclusion of the coefficient index in the High Frequency Coded Data) 4. Fourth Modality (Case of Inclusion of the coefficient index and the Difference of the Pseudohigh Frequency subband energy in the High Frequency Coded Data) 5. Fifth Modality (Case of Selection of the coefficient index Using Evaluated Value) 6. Sixth Modality (Case of Sharing Part of the Coefficients) 7. Seventh Modality (Case of Subjection of Resource Quantity to Uniformization) <1. First Mode>
[0074] With the first mode, components of the low-frequency signal after decoding to be obtained by decoding the encoded data using the high-frequency deletion encoding technique are subjected to processing to expand the frequency band (hereinafter referred to as as frequency band expansion processing). [Example of Functional Configuration of the Frequency Band Expansion Device]
[0075] Figure 3 illustrates an example of a functional configuration of a frequency band expansion device in which the present invention has been applied.
[0076] A frequency band expansion device 10 takes a component of the low frequency signal after decoding as an input signal, subjects the input signal thereof to frequency band expansion processing, and transmits a signal after processing frequency band expansion obtained as a result of this as an output signal.
[0077] The frequency band expansion device 10 is configured by a low pass filter 11, a delay circuit 12, band pass filters 13, a resource quantity calculation circuit 14, a sub energy estimation circuit. - high frequency band 15, a high frequency signal generating circuit 16, a high pass filter 17 and a signal adder 18.
[0078] The low pass filter 11 performs filtering an input signal with a predetermined cut-off frequency and supplies a low-frequency signal component, which is a low-frequency signal component, to the delay circuit 12 as a signal after of filtering.
[0079] In order to synchronize the addition time of a low-frequency signal component from the low-pass filter 11 and a high-frequency signal component described below, the delay circuit 12 delays the low-frequency signal component at fixed delay time to supply signal adder 18.
[0080] Bandpass filters 13 are configured by bandpass filters 13-1 to 13-N, each with a different pass band. The bandpass filter 13-i (1 <i<N) passes a predetermined bandpass signal from the input signals, and supplies it to the resource quantity calculation circuit 14 and the high frequency signal generation circuit 16 as one of multiple subband signals.
[0081] The resource amount calculation circuit 14 calculates a single or multiple resource amounts using at least any one of the multiple subband signals from the bandpass filters 13 or the input signal for supply to the resource estimation circuit. high frequency subband energy 15. Here, resource quantity is information that represents resources as a signal of the input signal.
[0082] The high frequency subband energy estimation circuit 15 calculates an estimated value of the high frequency subband energy, which is the energy of a high frequency subband signal for each subband of high frequency, based on a single or multiple resource quantities coming from the resource quantity calculation circuit 14, and supplies it to the high frequency signal generation circuit 16.
[0083] The high-frequency signal generating circuit 16 generates a high-frequency signal component, which is a component of the high-frequency signal, based on the multiple subband signals from the bandpass filters 13, and the multiple estimated values of the high frequency subband energy from the high frequency subband energy estimation circuit 15 to supply the high pass filter 17.
[0084] The high-pass filter 17 subjects the high-frequency signal component from the high-frequency signal generating circuit 16 to filtering with a cut-off frequency corresponding to a cut-off frequency in the low-pass filter 11 to supply the adder of sign 18.
[0085] The signal adder 18 adds the low frequency signal component from the delay circuit 12 and the high frequency signal component from the high pass filter 17, and transmits them as an output signal.
[0086] Note that, with the configuration of figure 3, in order to obtain a subband signal, bandpass filters 13 are applied, but are not restricted to these, and a band division filter described in PTL 1 can be applied, for example.
[0087] Also, similarly with the configuration of figure 3, in order to synthesize subband signals, signal adder 18 is applied, but not restricted to this, and a band synthesis filter described in PTL 1 can be applied. [Frequency Band Expansion Processing of the Frequency Band Expansion Device]
[0088] Next, the frequency band expansion processing by the frequency band expansion device of Figure 3 will be described in relation to the flowchart of Figure 4.
[0089] In step S1, the low pass filter 11 subjects the input signal to filtering with a predetermined cut-off frequency and supplies the component of the low-frequency signal that serves as a signal after filtering to the delay circuit 12.
[0090] The low pass filter 11 can define an optional frequency as a cut-off frequency, but, with the present modality, a predetermined band is taken as an expansion start band described below, and a corresponding cut-off frequency is defined. at the frequency of the lower end of its expansion start band. In this way, the low pass filter 11 supplies a low frequency signal component, which is a signal component of a lower frequency than the expansion start band, to the delay circuit 12 as a signal after filtering.
[0091] Also, the low pass filter 11 can also define the ideal frequency as a cut-off frequency according to the encoding technique with high frequency deletion of the input signal and encoding parameters such as bit rate and the like . As the encoding parameters, overhead information employed by the band expansion technique according to PTL 1 can be used, for example.
[0092] In step S2, the delay circuit 12 delays the component of the low frequency signal coming from the low pass filter 11 in the fixed delay time and supplies it to the signal adder 18.
[0093] In step S3, bandpass filters 13 (bandpass filters 13-1 through 13-N) divide the input signal into multiple subband signals, and supply each of the multiple subband signals after the division to the resource quantity calculation circuit 14 and to the high frequency signal generation circuit 16. Note that, regarding the processing of division of the input signal by bandpass filters 13, details of this will be described later.
[0094] In step S4, the resource amount calculation circuit 14 calculates a single or multiple resource amounts using at least one of the multiple subband signals from the bandpass filters 13 and the input signal for supplying the circuit of high frequency subband energy estimation 15. Note that, regarding the resource quantity calculation processing by the resource quantity calculation circuit 14, details of this will be described later.
[0095] In step S5, the high frequency subband energy estimation circuit 15 calculates multiple estimated values of the high frequency subband energy based on a single or multiple resource amounts from the amount calculation circuit. resource 14, and supplies them to the high frequency signal generating circuit 16. Note that, regarding the processing to calculate the estimated values of the high frequency subband energy by the high frequency subband energy estimating circuit 15 , details of this will be described later.
[0096] In step S6, the high-frequency signal generating circuit 16 generates a high-frequency signal component based on the multiple subband signals from the bandpass filters 13 and the multiple estimated values of the subband energy. high frequency band from the high frequency subband energy estimation circuit 15, and supplies it to the high pass filter 17. The high frequency signal component mentioned here is a signal component of higher frequency than the band. of expansion start. Note that, regarding the high frequency signal component generation processing by the high frequency signal generation circuit 16, details of this will be described later.
[0097] In step S7, the high-pass filter 17 subjects the high-frequency signal component from the high-frequency signal generation circuit 16 to filtering, thereby removing noise, such as distorted components, at an included low frequency into a high-frequency signal component and supplying the high-frequency signal component thereof to signal adder 18.
[0098] In step S8, the signal adder 18 adds the low frequency signal component from the delay circuit 12 and the high frequency signal component from the high pass filter 17 to supply them as an output signal.
[0099] According to the aforementioned processing, the frequency band can be expanded with respect to a component of the low frequency signal after decoding.
[00100] Below, details of each process from steps S3 to S6 of the flowchart of figure 4 will be described. [Band Pass Filter Processing Details]
[00101] First, details of the processing by bandpass filters 13 in step S3 of the flowchart of figure 4 will be described.
[00102] Note that, for convenience of description, below, the number N of bandpass filters 13 will be taken as N = 4.
[00103] For example, one of the 16 subbands obtained by the equal division of a Nyquist frequency of the input signal into 16 is taken as the expansion start band, and four subbands of the 16 subbands whose frequencies are more lower than the expansion start band are taken as the pass bands of bandpass filters 13-1 through 13-4, respectively.
[00104] Figure 5 illustrates locations on the geometric axis of the frequency of the pass bands of the pass band filters 13-1 to 13-4, respectively.
[00105] As illustrated in Figure 5, if it is said that, of the frequency bands (subbands) that are lower than the expansion start band, the index of the first subband of the high frequency is sb, the index of second subband is sb - 1 and the index of the first subband is sb - (I - 1), bandpass filters 13-1 to 13-4 assign, from the subbands with a frequency lower than the band of beginning of expansion, the sub-bands whose indices are sbat to sb - 3 as passing bands, respectively.
[00106] Note that, in the present modality, the pass bands of the passband filters 13-1 to 13-4 are four predetermined subbands of the 16 subbands obtained by the equal division of the Nyquist frequency of the input signal into 16, respectively, but without restrictions, and may be four predetermined subbands of the 256 subbands obtained by equal division of the Nyquist frequency of the input signal into 256, respectively. Also, the bandwidths of bandpass filters 13-1 through 13-4 may differ. [Details of Processing by the Resource Quantity Calculation Circuit]
[00107] Next, a description will be made in relation to the processing details by the resource quantity calculation circuit 14 in step S4 of the flowchart of figure 4.
[00108] The resource quantity calculation circuit 14 calculates a single or multiple resource quantities to be used for the high frequency subband energy estimation circuit 15 to calculate an estimated value of the subband energy of high frequency using at least any one of multiple subband signals from bandpass filters 13 and the input signal.
[00109] More specifically, the resource quantity calculation circuit 14 calculates, from the four subband signals from the bandpass filters 13, the subband energy signal (subband energy (below , also referred to as low frequency subband energy)) for each subband as a resource quantity for supplying the high frequency subband energy estimation circuit 15.
[00110] Specifically, the resource quantity calculation circuit 14 obtains low frequency subband energy power(ib, J) at a certain predetermined time interval J from four subband signals x(ib, n) supplied from bandpass filters 13 using the following Expression (1). Here, ib represents an index of a subband, and n represents a discrete-time index. Now, say the number of samples in a frame is FSIZE, and energy is represented by decibel. [Mathematical Expression 1]
[00111] In this way, the low frequency subband energy power(ib, J) obtained by the resource quantity calculation circuit 14 is supplied to the high frequency subband energy estimation circuit 15 as a quantity of resource. [Details of Processing by High Frequency Subband Power Estimation Circuit]
[00112] Next, a description will be made in relation to the processing details by the high frequency subband energy estimation circuit 15 in step S5 of the flowchart of figure 4.
[00113] The high frequency subband energy estimation circuit 15 calculates an estimated value of the subband energy (high frequency subband energy) of a band to be expanded (frequency expansion band) of a subband whose index is sb + 1 (expansion start band) and, subsequently, based on the four subband energies supplied from the resource quantity calculation circuit 14.
[00114] Specifically, if an index of the highest frequency subband of the frequency expansion band is said to be eb, the high frequency subband energy estimation circuit 15 estimates (eb - sb) energies of the subband in relation to the subbands whose indices are sb + 1up to eb.
[00115] An estimated value of the subband powerest(ib, J) whose index is ib in the frequency expansion band is represented, for example, by the following Expression (2) using the four energies of the subband power(ib, J) supplied from the resource quantity calculation circuit 14. [Mathematical Expression 2]

[00116] Here, in Expression (2), the coefficients Aib(kb) and Bib are coefficients with a different value for each subband ib. Say that the coefficients Aib(kb) and Bib are coefficients to be properly defined to obtain a suitable value for various input signals. Also, according to the change in the sb subband, the coefficients Aib(kb) and Bib also change to ideal values. Note that the derivation of the coefficients Aib(kb) and Bib will be described later.
[00117] In Expression (2), although an estimated value of a high-frequency subband energy is calculated by primary linear coupling using each energy of the multiple subband signals coming from bandpass filters 13, without restrictions, it it can be calculated using, for example, linear coupling of multiple low frequency subband energies of several frames before and after a time interval J or it can be calculated using a non-linear function.
[00118] In this way, the estimated value of the high frequency subband energy calculated by the high frequency subband energy estimation circuit 15 is supplied to the high frequency signal generating circuit 16. [Details of Processing by the High Frequency Signal Generation Circuit]
[00119] Next, a description will be made in relation to the processing details by the high frequency signal generation circuit 16 in step S6 of the flowchart of figure 4.
[00120] The high frequency signal generating circuit 16 calculates a low frequency subband energy power(ib, J) of each subband from the multiple subband signals supplied from the bandpass filters 13 based on the aforementioned Expression (1). The high frequency signal generating circuit 16 obtains a gain amount G(ib, J) by the following Expression (3) using the calculated multiple energies of the low frequency subband power(ib, J) and the estimated value of high frequency subband energy powerest(ib, J) calculated based on the aforementioned Expression (2) by the high frequency subband energy estimation circuit 15. [Mathematical Expression 3]

[00121] Here, in Expression (3), sbmap(ib) indicates a mapping source subband in the event that the ib subband is taken as a mapping destination subband, and is represented by the following Expression (4). [Mathematical Expression 4]

[00122] Note that, in Expression (4), INT(a) is a function to truncate below the decimal point of an a value.
[00123] Next, the high frequency signal generating circuit 16 calculates a subband signal x2(ib, n) after the gain adjustment by multiplying the output of bandpass filters 13 by the amount of gain G(ib , J) obtained by Expression (3) using the following Expression (5). [Mathematical Expression 5]

[00124] Additionally, the high frequency signal generating circuit 16 calculates a subband signal x3(ib, n) after adjusting the cosine transform gain from the subband signal x2(ib, n) ) after gain adjustment by performing cosine modulation from a frequency corresponding to the low end frequency of a subband whose index is sb -3 to a frequency corresponding to the high end frequency of a subband whose index is sb. [Mathematical Expression 6]

[00125] Note that, in Expression (6), π represents a circular constant. This Expression (6) means that each of the subband signals x2(ib, n) after the gain adjustment is shifted to a frequency on one side of the high frequency to the four-band value.
[00126] The high frequency signal generating circuit 16 calculates a component of the high frequency signal xhigh(n) of the subband signals x3(ib, n) after adjusting gain shifted to the high frequency side using the following Expression (7). [Mathematical Expression 7]

[00127] In this way, according to the high-frequency signal generating circuit 16, high-frequency signal components are generated based on the four low-frequency subband energies calculated based on the four subband signals coming from the bandpass filters 13 and at the estimated value of the high frequency subband energy from the high frequency subband energy estimating circuit 15, and are supplied to the high pass filter 17.
[00128] According to the aforementioned processing, with respect to the input signal obtained after decoding the data encoded by the high-frequency deletion encoding technique, low-frequency sub-band energies calculated from the multiple sub-signals band are taken as resource quantities and, based on these and appropriately defined coefficients, an estimated value of the high frequency subband energy is calculated and a component of the high frequency signal is generated in an adapted manner from the energies of the low frequency subband and the estimated value of the high frequency subband energy and, in this way, the subband energies in the frequency expansion band can be estimated with high precision and music signals can be reproduced with higher sound quality.
[00129] Although description has been made so far in relation to an example where the resource quantity calculation circuit 14 calculates only low frequency subband energies calculated from the multiple subband signals as resource quantities , in this case, a subband energy in the frequency expansion band may be able to be estimated with high precision, depending on the input signal types.
[00130] Therefore, the resource amount calculation circuit 14 also calculates a resource amount with a strong correlation to how to transmit a sound energy in the frequency expansion band, thereby enabling the estimation of a sub energy. -band in the frequency expansion band in the high frequency subband energy estimation circuit 15 is performed with higher accuracy. [Another Example of the Amount of Resource Calculated by the Resource Amount Calculation Loop]
[00131] Figure 6 illustrates an example of a frequency characteristic of a vocal section, which is a section in which vocal occupies most of a certain input signal, and a high-frequency energy spectrum obtained by calculating only the low frequency subband energies as resource quantities for estimating a high frequency subband energy.
[00132] As illustrated in Figure 6, with the frequency characteristic of a vocal section, the estimated high frequency energy spectrum is often located above the high frequency energy spectrum of the original signal. Unnatural sensations in relation to the human voice are readily perceived by the human ear and, in this way, estimation of a high frequency subband energy needs to be performed with high precision in a particular vocal section.
[00133] Also, as illustrated in figure 6, with the frequency characteristic of a vocal section, there is often a large recessed part from 4.9 kHz to 11.025 kHz.
[00134] Therefore, below, a description will be made in relation to an example in which a lowered degree of 4.9 kHz to 11.025 kHz in a frequency region is applied as an amount of resource to be used for estimating an energy of high frequency subband of a vocal section. Now, in the following, the resource amount that indicates this lowered degree will be referred to as immersion.
[00135] In the following, an example of the calculation of the dip(J) in the time interval J will be described.
[00136] First, from the input signal, signals in 2,048 sample sections included in several frames before and after including the time interval J are subjected to the 2,048-point FFT (Fast Fourier Transform) to calculate coefficients on the frequency axis . The absolute values of the calculated coefficients are subjected to db transform to obtain energy spectra.
[00137] Figure 7 illustrates an example of the energy spectra thus obtained. Here, in order to remove fine components from the energy spectra, elevation processing is performed to remove components of 1.3 kHz or less, for example. According to elevation processing, each dimension of the energy spectra is taken as a time series, and is subjected to a low-pass filter to perform filtering processing, whereby fine components of a spectrum peak can be smoothed. .
[00138] Figure 8 illustrates an example of the energy spectrum of an input signal after elevation. With the energy spectrum after the rise illustrated in figure 8, the difference between the minimum and maximum value of the energy spectrum included in a range equivalent to 4.9 kHz up to 11.025 kHz is taken as the immersion dip(J).
[00139] In this way, a resource amount with strong correlation with the subband energy in the frequency expansion band is calculated. Note that an example of immersion dip(J) calculation is not restricted to the aforementioned technique, and another technique can be employed.
[00140] Next, a description will be made in relation to another example of calculation of a resource quantity with strong correlation with the subband energy in the frequency expansion band. [Yet Another Example of Calculating the Amount of Resource Calculated by the Resource Amount Calculation Loop]
[00141] Of a certain input signal, with the frequency characteristic of an attack section, which is a section that includes a music signal with attack, as described in relation to Figure 2, often the energy spectrum in the high frequency side is generally flat. In the technique to calculate only low frequency subband energies as resource amounts, the frequency expansion band subband energy is estimated without using a resource amount that represents temporal fluctuation peculiar to the input signal that includes a attack section and thus it is difficult to estimate the subband energy of the overall flat frequency expansion band viewed in an attack section with high accuracy.
[00142] Therefore, in the following, a description will be made in relation to an example in which temporal fluctuation of a low frequency subband energy is applied as a resource quantity to be used for estimating a subband energy of high frequency of an attack section.
[00143] Temporal fluctuation powerd(J) of a low frequency subband energy in a certain time interval J is obtained by the following Expression (8), for example. [Mathematical Expression 8]

[00144] According to Expression (8), the temporal fluctuation powerd(J) of a low frequency subband energy represents a ratio between the sum of four low frequency subband energies in the time interval J and the sum of four energies of the low frequency subband in the time interval (J - 1), which is one frame before time interval J, and the larger this value, the greater the temporal fluctuation of the energy between the frames, that is, it is conceivable that the signal included in time interval J has a strong attack nature.
[00145] Also, when comparing the statistically average energy spectrum illustrated in figure 1 and the attack section energy spectrum (music signal with attack) illustrated in figure 2, the attack section energy spectrum increases in the direction from the right at the middle frequency. With attack sections, such a frequency characteristic is often displayed.
[00146] Therefore, in the following, description will be made in relation to an example in which, as a resource quantity to be used for estimating a high frequency subband energy of a strike section, slope in the middle frequency of this is employed.
[00147] Slope slope(J) of the middle frequency in a certain time interval J is obtained by the following Expression (9), for example. [Mathematical Expression 9]

[00148] In Expression (9), a coefficient w(ib) is a weighting coefficient adjusted to weight by the high frequency subband energy. According to Expression (9), slope(J) represents a ratio between the sum of the four energies of the low frequency subband weighted by the high frequency and the sum of the four energies of the low frequency subband. For example, in the event that the four energies of the low frequency subband become the energy for the middle frequency subband, when the energy spectrum of the middle frequency increases in the upper right direction, slope(J) has a large value, and when the energy spectrum of the middle frequency falls in the lower right direction, it has a small value.
[00149] Also, often the slope of the frequency of the middle fluctuates enormously before and after a lead section, and thus the temporal fluctuation sloped(J) of the slope represented by the following Expression (10) can be taken as a quantity of feature to be used for estimating a substituted high frequency energy of an attack section. [Mathematical Expression 10]

[00150] Also, similarly, the temporal fluctuation dipd(J) of the aforementioned dip(J) represented by the following Expression (11) can be taken as a resource quantity to be used for estimating a high frequency subband energy of an attack section. [Mathematical Expression 11]

[00151] According to the above-mentioned technique, a resource amount with a strong correlation with the subband energy of the frequency expansion band is calculated and, in this way, estimation of the subband energy of the frequency expansion band in the high frequency subband power estimation circuit 15 can be performed with higher accuracy.
[00152] Although description has been made so far in relation to an example in which a resource quantity with a strong correlation with the subband energy of the frequency expansion band is calculated, in the following, description will be made in relation to an example where a high frequency subband energy is estimated using the resource amount thus calculated. [Details of Processing by High Frequency Subband Power Estimation Circuit]
[00153] Now, description will be made in relation to an example where a high frequency subband energy is estimated using the immersion and low frequency subband energies described with respect to figure 8 as resource quantities.
[00154] Specifically, in step S4 of the flowchart of Figure 4, the resource quantity calculation circuit 14 calculates a low frequency subband energy and the dip of four subband signals for each subband of the bandpass filters 13 as resource quantities for supplying the high frequency subband power estimation circuit 15.
[00155] In step S5, the high-frequency subband energy estimation circuit 15 calculates an estimated value for a high-frequency subband energy based on the four low-frequency subband energies and immersion from the resource quantity calculation circuit 14.
[00156] Here, between the subband energies and the immersion, a range (scale) of a value to be obtained differs and, in this way, the high frequency subband energy estimation circuit 15 performs the following conversion into the immersion value, for example.
[00157] The high frequency subband energy estimation circuit 15 calculates the subband energy with the highest frequency of the four energies of the low frequency subband and the immersion value in relation to a large number of input signals and obtains a mean value and standard deviation with respect to each of these in advance. Now, say that an average value of the subband energies is powerave, the standard deviation of the subband energies is powerstd, an average value of the dip is dipave, and the standard deviation of the dip is dipstd.
[00158] The high frequency subband energy estimation circuit 15 converts the dip(J) value of the dip using these values, such as the following Expression (12), to obtain a dips(J) dip after the conversation . [Mathematical Expression 12]

[00159] According to the conversion indicated in Expression (12) being performed, the high frequency subband energy estimation circuit 15 can convert the dip(J) value into a variable (immersion) dips(J ) statistically equal to the average and dispersion of the energies of the low frequency subband and, in this way, an average of a value that the immersion has can be defined, in general, equal to a range of a value that the energies of the sub -band have.
[00160] With the frequency expansion band, an estimated powerest(ib, J) value of a subband energy whose index is ib represented by the following Expression (13) using linear coupling between the four subband energies of low frequency power(id, J) from the resource quantity calculation circuit 14 and the immersion dips(J) indicated in Expression (12), for example. [Mathematical Expression 13]

[00161] Here, in Expression (13), the coefficients Cib(kb), Dib and Eib are coefficients with a different value for each subband id. Say that the Cib(kb), Dib and Eib coefficients are coefficients to be properly defined to obtain a suitable value for various input signals. Also, according to the change in the sb subband, the Cib(kb), Dib and Eib coefficients also change to ideal values. Note that the derivation of the Cib(kb), Dib and Eib coefficients will be described later.
[00162] In Expression (13), although an estimated value of a high frequency subband energy is calculated by the primary linear coupling, without restrictions, for example, it can be calculated using linear couplings of multiple resource quantities of several frames before and after time interval J or can be calculated using a non-linear function.
[00163] According to the aforementioned processing, the immersion value peculiar to a vocal section is used to estimate a high frequency subband energy, thus compared to a case where only the subband energies Low frequency are taken as resource quantities, increasing the accuracy of estimating a high frequency subband energy in a vocal section and reducing unnatural sensations that are readily perceived by the human ear, brought about by a spectrum of sub energy. -high frequency band which is overestimated, so the high frequency energy spectrum of the original signal using the technique where only energies from the low frequency subband are taken as resource amounts and thus music signals can be reproduced with higher sound quality.
[00164] Incidentally, in relation to immersion (decreased degree in the frequency characteristic in a vocal section) calculated as a resource amount by the aforementioned technique, in the event that the number of subband divisions is 16, the frequency resolution is low, and thus this lowered degree cannot be expressed with the energies of the low-frequency subband alone.
[00165] Therefore, the number of subband divisions increases (for example, 256 divisions, equivalent to 16 beats), the number of band divisions by bandpass filters 13 increases (for example, 64, equivalent to 16 beats) and the number of low frequency subband energies to be calculated by the resource quantity calculation circuit 14 increases (eg 64, equivalent to 16 times), thereby increasing the frequency resolution and enabling one degree lowered is expressed with low frequency subband energies only.
[00166] Thus, it is believed that a high frequency subband energy can be estimated, in general, with the same precision as estimating a high frequency subband energy using the aforementioned immersion as a resource quantity. , using low frequency subband energies only.
[00167] However, the amount of calculation increases by increasing the number of subband divisions, the number of band divisions and the number of energies of the low frequency subband. If we consider that any technique can estimate a high frequency subband energy with similar accuracy, it is believed that a technique to estimate a high frequency subband energy without increasing the number of subband divisions using dip as a resource quantity is effective in one aspect of calculation quantity.
[00168] Although description has been made so far in relation to techniques for estimating a high frequency subband energy using the immersion and low frequency subband energies, a resource amount to be used for estimating a high frequency subband energy is not restricted to this combination, and one or multiple above-described resource quantities (low frequency subband energies, dip, temporal fluctuation of the low frequency subband energies, slope, fluctuation temporal slope and temporal fluctuation of immersion) can be employed. Thus, accuracy can be further increased by estimating a high frequency subband energy.
[00169] Also, as stated, with an input signal, a parameter peculiar to a section in which estimating a high frequency subband energy is difficult is employed as a resource quantity to be used for estimating a high frequency subband energy. high frequency subband, thereby enabling the estimation accuracy of its section to increase. For example, temporal fluctuation of low frequency subband energies, slope, temporal fluctuation of the slope and temporal fluctuation of immersion are parameters peculiar to attack sections, and these parameters are employed as resource quantities, thereby enabling the accuracy of estimating a high frequency subband energy in an attack section increases.
[00170] Note that, in the event that resource quantities different from the energies of the low frequency subband and the immersion, i.e., temporal fluctuation of the low frequency subband energies, slope, temporal fluctuation of the slope and fluctuation immersion time, are also employed to perform estimation of a high frequency subband energy, a high frequency subband energy can be estimated by the same technique as the aforementioned technique.
[00171] Note that the techniques for calculating resource quantities mentioned here are not restricted to the aforementioned techniques, and another technique may be employed. [How to Obtain Cib(kb), Dib and Eib Coefficients]
[00172] Next, a description will be made in relation to how to obtain the coefficients Cib(kb), Dib and Eib in the aforementioned Expression (13).
[00173] As a method to obtain the Cib(kb), Dib and Eib coefficients, in order to obtain suitable coefficients, the Cib(kb), Dib and Eib coefficients for various input signals at the time of estimating the energy of sub- band of frequency expansion band, a technique will be employed in which learning is performed using a wideband supervisory signal (hereafter referred to as wideband supervisory signal) in advance, and the Cib(kb), Dib, and Eib coefficients are determined based on their learning outcomes.
[00174] At the time of learning the Cib(kb), Dib and Eib coefficients, a coefficient learning device will be applied, in which bandpass filters with the same pass bandwidths as the passband filters 13-1 up to 13-14 described in relation to Figure 5 are arranged at a frequency higher than the expansion start band. The coefficient learning device performs learning when a wideband supervisory signal is input. [Example of Functional Configuration of Coefficient Learning Device]
[00175] Figure 9 illustrates an example of a functional configuration of a coefficient learning device to perform learning of the Cib(kb), Dib and Eib coefficients.
[00176] Regarding signal components with frequency lower than the start band of expansion of the wideband supervisor signal to be inserted into a coefficient 20 learning device of Figure 9, it is desirable that an input signal with band restricted to be inserted in the frequency band expansion device 10 of Fig. 3 is a signal encoded by the method equal to the encoding method subject at the time of encoding.
[00177] The coefficient learning device 20 is configured by bandpass filters 21, a high frequency subband energy calculation circuit 22, a resource quantity calculation circuit 23 and a coefficient estimation circuit 24 .
[00178] Bandpass filters 21 are configured by bandpass filters 21-1 to 21-(K+N), each with a different pass band. The bandpass filter 21-i(1<i<K+N) passes a predetermined bandpass signal of an input signal and supplies it to the high frequency subband energy calculation circuit 22 or the high frequency subband energy calculation circuit. calculating resource amount 23 as one of multiple subband signals. Note that from bandpass filters 21-1 through 21-(K+N), bandpass filters 21-1 through 21-K pass a higher frequency signal than the expansion start band.
[00179] The high frequency subband energy calculation circuit 22 calculates a high frequency subband energy for each subband for each fixed time interval for multiple high frequency subband signals from the pass filters band 21 for supplying the coefficient estimation circuit 24.
[00180] The resource amount calculation circuit 23 calculates the resource amount equal to a resource amount calculated by the resource amount calculation circuit 14 of the frequency band expansion device 10 of Figure 3 for each frame equal to a fixed time interval, in which a high frequency subband energy is calculated by the high frequency subband energy calculation circuit 22. That is, the resource quantity calculation circuit 23 calculates one or multiple resource quantities using at least one of the multiple subband signals from the bandpass filters 21 and the wideband supervisor signal for supplying the coefficient estimation circuit 24.
[00181] The coefficient estimating circuit 24 estimates coefficients (coefficient data) to be used in the high frequency subband energy estimating circuit 15 of the frequency band expansion device 10 of figure 3 based on the energy of the high frequency subband from the high frequency subband energy calculation circuit 22 and the resource quantities from the resource quantity calculation circuit 23 for each fixed time interval. [Coefficient Learning Device Coefficient Learning Processing]
[00182] Next, coefficient learning processing by the coefficient learning device of figure 9 will be described in relation to the flowchart of figure 10.
[00183] In step S11, bandpass filters 21 divide an input signal (wideband supervisor signal) into (K+N) subband signals. The bandpass filters 21-1 through 21-K supply multiple subband signals of higher frequency than the expansion start band to the high frequency subband energy calculation circuit 22. Also, the bandpass filters 21-(K+1) to 21-(K+N) supply multiple subband signals of frequency lower than the expansion start band to the resource quantity calculation circuit 23.
[00184] In step S12, the high frequency subband power circuit 22 calculates a high frequency subband energy power(ib, J) for each subband for each fixed time interval for multiple signals. high frequency subband from bandpass filters 21 (bandpass filters 21-1 to 21-K). The high frequency subband energy power(ib, J) is obtained by the aforementioned Expression (1). The high frequency subband energy calculation circuit 22 supplies the calculated high frequency subband energy to the coefficient estimation circuit 24.
[00185] In step S13, the resource amount calculation circuit 23 calculates a resource amount for each time interval equal to a fixed time interval in which a high frequency subband energy is calculated by the calculation circuit of the high frequency subband energy 22.
[00186] With the resource quantity calculation circuit 14 of the frequency band expansion device 10 of Figure 3, it was considered that four low frequency subband energies and one dip are calculated as resource quantities and similarly , also with the resource quantity calculation circuit 23 of the coefficient learning device 20, description will be made considering that the four subband energies of low frequency and immersion are calculated.
[00187] Specifically, the resource quantity calculation circuit 23 calculates four low-frequency subband energies using four subband signals with the same bands as the four subband signals to be input into the calculation circuit. resource amount 14 of frequency band expansion device 10, from bandpass filters 21 (bandpass filters 21-(K+1) to 21-(K+4)). Also, the resource quantity calculation circuit 23 calculates a dip from the broadband supervisor signal and calculates a dips(J) dip based on the aforementioned Expression (12). The resource quantity calculation circuit 23 supplies the four energies of the low frequency subband and the immersion dips(J) calculated to the coefficient estimation circuit 24 as resource quantities.
[00188] In step S14, the coefficient estimation circuit 24 performs estimation of the coefficients Cib(kb), Dib and Eib based on a large number of combinations between (eb - sb) energies of the high frequency subband and on the resource quantities (four low frequency subband energies and dips(J)) supplied from the high frequency subband energy calculation circuit 22 and the resource quantity calculation circuit 23 in the time interval. For example, the coefficient estimation circuit 24 takes, in relation to a certain high frequency subband, five resource quantities (four low frequency subband energies and dips(J) immersion) as explanatory variables and takes the high frequency subband energy power(ib, J) as an explained variable to perform regression analysis using the least squares method, thereby shifting the Cib(kb), Dib and Eib coefficients of Expression (13) .
[00189] Note that it goes without saying that the estimation technique for the Cib(kb), Dib and Eib coefficients is not restricted to the aforementioned technique and several common parameter identification methods can be employed.
[00190] According to the aforementioned processing, learning the coefficients to be used for estimating a high frequency subband energy is performed using the broadband supervisor signal in advance and, in this way, adequate output results can be obtained for various input signals to be inputted into the frequency band expansion device 10 and, consequently, music signals can be reproduced with higher sound quality.
[00191] Note that the coefficients Aib(kb) and Bib in the aforementioned Expression (2) can also be obtained by the aforementioned coefficient learning method.
[00192] Description has been made so far in relation to coefficient learning processing considering that, with the high frequency subband energy estimation circuit 15 of the frequency band expansion device 10, a promise of an estimated value of each high frequency subband energy is calculated by the linear coupling between the four low frequency subband energies and immersion. However, the technique for estimating a high frequency subband energy in the high frequency subband energy estimation circuit 15 is not restricted to the above-mentioned example, and a high frequency subband energy can be calculated by resource quantity calculation circuit 14 which calculates one or multiple resource quantities (temporal fluctuation of low frequency subband energy, slope, temporal fluctuation of the slope and temporal fluctuation of an embedding) other than an embedding, or linear coupling between multiple resource amounts of multiple frames before and after time interval J can be employed or a non-linear function can be employed. That is, with coefficient learning processing, it is sufficient for the coefficient estimation circuit 24 to calculate (learn) the coefficients with conditions equal to the conditions in relation to resource quantities, time interval and a function to be used at the time a high frequency subband energy is calculated by the high frequency subband energy estimating circuit 15 of the frequency band expansion device 10. <2. Second Mode>
[00193] In the second embodiment, the input signal is subjected to encoding processing and decoding processing in the encoding technique with high frequency characteristic by an encoding device and a decoding device. [Example of Functional Configuration of Encoding Device]
[00194] Figure 11 illustrates an example of a functional configuration of an encoding device in which the present invention has been applied.
[00195] An encoding device 30 is configured by a low pass filter 31, a low frequency encoding circuit 32, a subband division circuit 33, a resource quantity calculation circuit 34, a calculation circuit of the pseudo-high-frequency subband power 35, a pseudo-high-frequency sub-band power difference calculating circuit 36, a high-frequency encoding circuit 37, a multiplexing circuit 38, and a low-frequency decoding circuit 39.
[00196] The low pass filter 31 subjects an input signal to filtering at a predetermined cut-off frequency and supplies a lower frequency signal (hereinafter referred to as the low frequency signal) than the cut-off frequency to the coding circuit in low frequency 32, to the subband division circuit 33 and to the resource amount calculation circuit 34 as a signal after filtering.
[00197] The low-frequency encoding circuit 32 encodes the low-frequency signal from the low-pass filter 31 and supplies low-frequency encoded data obtained as a result of this to the multiplexing circuit 38 and the low-frequency decoding circuit 39.
[00198] The subband division circuit 33 equally divides the input signal and the low frequency signal from the low pass filter 31 into multiple subband signals with predetermined bandwidth for supply to the quantity calculation circuit of resource 34 or to the pseudo-high frequency subband energy difference calculation circuit 36. More specifically, the subband division circuit 33 supplies multiple subband signals (hereinafter, referred to as subband signals of low frequency) obtained with the low frequency signals as input to the resource quantity calculation circuit 34. Also, the subband division circuit 33 supplies, of the multiple subband signals obtained with the input signal as input , higher frequency subband signals (hereinafter, referred to as high frequency subband signals) than a cut-off frequency defined in the low-pass filter 31 to the energy difference calculation circuit d and pseudohigh frequency subband 36.
[00199] Resource amount calculation circuit 34 calculates one or multiple resource amounts using at least any one of multiple subband signals of low frequency subband signals from subband division circuit 33 and the low-frequency signal from the low-pass filter 31 for supplying the pseudo-high-frequency subband energy calculation circuit 35.
[00200] The pseudo-high-frequency sub-band energy calculation circuit 35 generates a pseudo-high-frequency sub-band energy based on the one or multiple resource quantities from the resource quantity calculation circuit 34 to supply the circuit calculation of the pseudo-high frequency subband energy difference 36.
[00201] The pseudo-high-frequency sub-band energy difference calculation circuit 36 calculates the pseudo-high-frequency sub-band energy difference described below based on the high-frequency sub-band signal from the division circuit of sub-band 33 and in the pseudo-high-frequency sub-band energy from the pseudo-high-frequency sub-band energy calculation circuit 35 for supply to the high-frequency encoding circuit 37.
[00202] The high-frequency encoding circuit 37 encodes the pseudo-high-frequency sub-band energy difference from the pseudo-high-frequency sub-band energy difference calculation circuit 36 to supply high-frequency encoded data obtained therefrom from this to the multiplexing circuit 38.
[00203] The multiplexing circuit 38 multiplexes the low-frequency encoded data from the low-frequency encoding circuit 32 and the high-frequency encoded data from the high-frequency encoding circuit 37 for transmission as an output code sequence.
[00204] The low-frequency decoding circuit 39 decodes the low-frequency encoded data from the low-frequency encoding circuit 32, as appropriate, to supply decoded data obtained therefrom to the subband division circuit 33 and the subband circuitry. calculation of the amount of resource 34. [Encoding Device Encoding Processing]
[00205] In the following, encoding processing by the encoding device 30 of Figure 11 will be described in relation to the flowchart of Figure 12.
[00206] In step S111, the low-pass filter 31 subjects an input signal to filtering at a predetermined cut-off frequency to supply a low-frequency signal that serves as a signal after filtering to the low-frequency encoding circuit 32, by subband division circuit 33 and resource quantity calculation circuit 34.
[00207] In step S112, the low-frequency encoding circuit 32 encodes the low-frequency signal from the low-pass filter 31 to supply low-frequency encoded data obtained as a result to the multiplexing circuit 38.
[00208] Note that, regarding the coding of the low frequency signal of step S112, it is sufficient that a suitable coding system is selected according to the coding efficiency or a circuit scale is requested, and the present invention does not depend of this coding system.
[00209] In step S113, the subband division circuit 33 equally divides the input signal and the low frequency signal into multiple subband signals with a predetermined bandwidth. Subband division circuit 33 supplies low frequency subband signals obtained with the low frequency signal as input to resource quantity calculation circuit 34. Also, subband division circuit 33 supplies, of the multiple subband signals with the input signals as input, high frequency subband signals with a band higher than the band frequency limit defined in the low pass filter 31 to the power difference calculation circuit. pseudohigh frequency subband 36.
[00210] In step S114, the resource amount calculation circuit 34 calculates one or multiple resource amounts using at least any one of the multiple subband signals of the low frequency subband signals coming from the frequency division circuit. subband 33 and the low-frequency signal from the low-pass filter 31 for supplying the pseudo-high-frequency subband energy calculation circuit 35. Note that the resource quantity calculation circuit 34 of Fig. 11 basically has , the same configuration and function of the resource quantity calculation circuit 14 of figure 3, and the processing of step S114 is basically the same processing of step S4 of the flowchart of figure 4 and, in this way, detailed description of this will be omitted .
[00211] In step S115, the pseudo-high-frequency sub-band energy calculation circuit 35 generates a pseudo-high-frequency sub-band energy based on one or multiple resource quantities from the resource quantity calculation circuit 34 for supplying the pseudo-high-frequency sub-band power difference calculation circuit 36. Note that the pseudo-high-frequency sub-band power calculation circuit 35 of Fig. 11 basically has the same configuration and function as the power estimation circuit. high frequency subband power 15 of Fig. 3 and the processing of step S115 is basically the same processing as step S5 of the flowchart of Fig. 4 and thus detailed description thereof will be omitted.
[00212] In step S116, the pseudo-high-frequency sub-band energy difference calculation circuit 36 calculates the pseudo-high-frequency sub-band energy difference based on the high-frequency sub-band signal from the high frequency subband signal. sub-band 33 and in the pseudo-high-frequency sub-band energy from the pseudo-high-frequency sub-band energy calculation circuit 35 for supply to the high-frequency encoding circuit 37.
[00213] More specifically, the pseudo-high-frequency sub-band energy difference calculation circuit 36 calculates a high-frequency sub-band energy power(ib, J) at a certain fixed time interval J with respect to the sub signal. -high frequency band coming from the subband division circuit 33. Now, in the present embodiment, say that all of the subband of the low frequency subband signal and the subband of the subband signal high frequency bands are identified using the ib index. The subband energy calculation technique is the same as the first modality technique, that is, the technique using Expression (1) can be applied.
[00214] Next, the pseudo-high-frequency sub-band energy difference calculation circuit 36 obtains the difference (pseudo-high-frequency sub-band energy difference) powerdiff(ib, J) between the high-frequency sub-band energy power (ib, J) and the pseudo-high-frequency subband energy powerlh(ib, J) from the pseudo-high-frequency subband energy calculation circuit 35 at time interval J. The difference of the pseudo-high-frequency subband energy frequency powerdiff(ib, J) is obtained by the following Expression (14). [Mathematical Expression 14]

[00215] In Expression (14), the index sb + 1 represents the index of the lowest frequency subband of the high frequency subband signals. Also, index eb represents the index of the highest frequency subband to be encoded from the high frequency subband signals.
In this way, the pseudo-high-frequency sub-band energy difference calculated by the pseudo-high-frequency sub-band energy difference calculation circuit 36 is supplied to the high-frequency coding circuit 37.
[00217] In step S117, the high-frequency encoding circuit 37 encodes the pseudo-high-frequency sub-band energy difference from the pseudo-high-frequency sub-band energy difference calculation circuit 36 to supply high-encoded data frequency obtained as a result of this to the multiplexing circuit 38.
[00218] More specifically, the high-frequency coding circuit 37 determines which cluster of multiple clusters in the pseudohigh-frequency subband energy difference characteristic space defined in advance a vector converted from the subband energy difference of pseudo-high frequency from the pseudo-high-frequency sub-band energy difference calculation circuit 36 (hereinafter, referred to as pseudo-high-frequency sub-band difference vector) belongs. Here, the pseudohigh frequency subband energy difference vector over a certain time interval J indicates an (eb - sb)-dimensional vector with the pseudohigh frequency subband energy difference value powerdiff(ib, J) for each index ib as each element. Also, the characteristic space of the pseudohigh frequency subband energy difference is also the (eb - sb)-dimensional space.
[00219] The high-frequency coding circuit 37 measures, with the pseudo-high-frequency sub-band energy difference characteristic space, the distance between each representative vector of multiple clusters defined in advance and the sub-band energy difference vector. pseudo-high-frequency band, obtains an index of a cluster with the shortest distance (hereinafter, referred to as the pseudo-high-frequency sub-band energy difference ID), and supplies it to the multiplexing circuit 38 as high-frequency encoded data.
[00220] In step S118, the multiplexing circuit 38 multiplexes the low-frequency encoded data transmitted from the low-frequency encoding circuit 32 and the high-frequency encoded data transmitted from the high-frequency encoding circuit 37, and transmits a string of exit code.
[00221] Incidentally, as an encoding device according to the high-frequency characteristic encoding technique, it was disclosed in Japanese Unexamined Patent Application Publication 200717908 a technique wherein a pseudo-high-frequency subband signal is generated from a low frequency subband signal, the pseudohigh frequency subband signal and the energy of a high frequency subband signal are compared for each subband, the energy gain for each subband is calculated to match the pseudohigh frequency subband energy and the energy of the high frequency subband signal and this is included in a code sequence as high frequency characteristic information.
[00222] On the other hand, according to the above processing, as information to estimate a high frequency subband energy at the time of decoding, it is sufficient that only the pseudohigh frequency subband energy difference ID is included in the exit code string. Specifically, for example, in the event that the number of clusters defined in advance is 64, as information to restore a high-frequency signal in the decoding device, it is sufficient that only 6-bit information for a time interval is added in the sequence of code and, if compared to a technique disclosed in Japanese Unexamined Patent Application Publication 2007-17908, volume of information to be included in the code sequence can be reduced and, in this way, coding efficiency can increase and, consequently, signals of music can be played with higher sound quality.
[00223] Also, with the aforementioned processing, if there is room for bulk computation, a low-frequency signal obtained by the low-frequency decoding circuit 39 which decodes the low-frequency encoded data from the low-frequency encoding circuit 32 can be inputted into the subband division circuit 33 and into the resource amount calculation circuit 34. With the decoding processing by the decoding device, a resource amount is calculated from the low frequency signal decoded from the data encoded at low frequency and the energy of a high frequency subband is estimated based on the amount of its resource. Therefore, with encoding processing, too, in the event that the pseudo-high-frequency subband energy difference ID to be calculated based on the resource amount calculated from the decoded low-frequency signal is included in the sequence of code, with the decoding processing by the decoding device, a high frequency subband energy can be estimated with higher accuracy. In this way, music signals can be played back with better sound quality. [Example of Functional Configuration of the Decoding Device]
[00224] In the following, an example of functional configuration of a decoding device corresponding to the encoding device 30 of figure 11 will be described in relation to figure 13.
[00225] A decoding device 40 is configured by a demultiplexing circuit 41, a low frequency decoding circuit 42, a subband division circuit 43, a resource quantity calculation circuit 44, a decoding circuit at high frequency 45, a decoded high frequency subband energy calculation circuit 46, a decoded high frequency signal generating circuit 47 and a synthesizing circuit 48.
[00226] The demultiplexing circuit 41 demultiplexes an input code sequence into high-frequency encoded data and low-frequency encoded data, supplies the low-frequency encoded data to the low-frequency decoding circuit 42, and supplies the high-encoded data frequency to the high frequency decoding circuit 45.
[00227] The low-frequency decoding circuit 42 performs decoding of the low-frequency encoded data from the demultiplexing circuit 41. The low-frequency decoding circuit 42 supplies a low-frequency signal obtained as a result of the decoding (hereinafter referred to as as decoded low frequency signal) to the subband division circuit 43, the resource quantity calculation circuit 44 and the synthesizing circuit 48.
The subband division circuit 43 equally divides the decoded low frequency signal from the low frequency decoding circuit 42 into multiple subband signals with a predetermined bandwidth and supplies the subband signals obtained (decoded low frequency subband signals) to the resource quantity calculation circuit 44 and to the decoded high frequency signal generation circuit 47.
[00229] The resource amount calculation circuit 44 calculates one or multiple resource amounts using at least any one of multiple subband signals from the decoded low frequency subband signals coming from the subband division circuit 43 and the decoded low-frequency signal for supplying the decoded high-frequency subband energy calculation circuit 46.
[00230] The high-frequency decoding circuit 45 performs decoding of the high-frequency encoded data from the demultiplexing circuit 41 and uses a pseudo-high-frequency subband energy difference ID obtained therefrom to supply a coefficient for estimate the energy of a high frequency subband (hereinafter, referred to as decoded high frequency subband energy estimation coefficient) prepared in advance for each ID (index) to the subband energy calculation circuit of high frequency decoded 46.
[00231] The high frequency subband energy calculation decoding circuit 46 calculates a decoded high frequency subband energy based on the one or multiple resource quantities and the subband energy estimation coefficient decoded high-frequency signal coming from the high-frequency decoding circuit 45 for supplying the decoded high-frequency signal generating circuit 47.
[00232] The decoded high frequency signal generating circuit 47 generates a decoded high frequency signal based on the decoded low frequency subband signals from the subband division circuit 43 and the high subband energy decoded frequency from the decoded high frequency subband energy calculation circuit 46 for supply to the synthesizing circuit 48.
[00233] The synthesizing circuit 48 synthesizes the decoded low-frequency signal from the low-frequency decoding circuit 42 and the decoded high-frequency signal from the decoded high-frequency signal generating circuit 47, and transmits this as a signal about to leave. [Decoding Device Decoding Process]
[00234] Next, the decoding processing by the decoding device of Figure 13 will be described in relation to the flowchart of Figure 14.
[00235] In step S131, the demultiplexing circuit 41 demultiplexes an input code sequence into high-frequency encoded data and low-frequency encoded data, supplies the low-frequency encoded data to the low-frequency decoding circuit 42 and supplies the high frequency encoded data to the high frequency decoding circuit 45.
[00236] In step S132, the low-frequency decoding circuit 42 performs decoding of the low-frequency encoded data from the demultiplexing circuit 41 and supplies a decoded low-frequency signal obtained therefrom to the sub-band division circuit 43 , resource quantity calculation circuit 44 and synthesizing circuit 48.
[00237] In step S133, the subband division circuit 43 equally divides the decoded low frequency signal from the low frequency decoding circuit 42 into multiple subband signals with a predetermined bandwidth and supplies the signals of decoded low frequency subband obtained to the resource quantity calculation circuit 44 and to the decoded high frequency signal generation circuit 47.
[00238] In step S134, the resource amount calculation circuit 44 calculates one or multiple resource amounts from at least any one of multiple subband signals from the decoded low frequency subband signals coming from the subband division circuit 43 and the decoded low frequency signal from the low frequency decoding circuit 42 for supplying the decoded high frequency subband energy calculation circuit 46. Note that the resource quantity calculation circuit 44 of figure 13 has basically the same configuration and function as the resource quantity calculation circuit 14 of figure 3, and the processing of step S134 is basically the same as the processing of step S4 of the flowchart of figure 4 and of this way, detailed description of this will be omitted.
[00239] In step S135, the high-frequency decoding circuit 45 performs decoding of the high-frequency encoded data from the demultiplexing circuit 41, uses a pseudo-high-frequency subband energy difference ID obtained therefrom to supply a decoded high frequency subband energy estimation coefficient prepared in advance for each ID (index) to the decoded high frequency subband energy calculation circuit 46.
[00240] In step S136, the decoded high-frequency subband energy calculation circuit 46 calculates a decoded high-frequency subband energy based on the one or multiple resource quantities from the decoded high frequency subband energy calculation circuit. resource 44 and in the decoded high frequency subband energy estimation coefficient from the high frequency decoding circuit 45 for supplying the decoded high frequency signal generating circuit 47. Note that the sub energy calculation circuit -decoded high frequency band 46 of Fig. 13 basically has the same configuration and function as the high frequency subband energy estimation circuit 15 of Fig. 3, and the processing of step S136 is basically the same as processing of step S5 of the flowchart of figure 4 and, in this way, detailed description of this will be omitted.
[00241] In step S137, the decoded high-frequency signal generating circuit 47 transmits a decoded high-frequency signal based on the decoded low-frequency subband signal from the subband division circuit 43 and the power decoded high frequency subband energy calculation circuit 46 from decoded high frequency subband energy calculation circuit. Note that the decoded high frequency signal generating circuit 47 of Fig. 13 basically has the same configuration and function as the circuit of generating the high frequency signal 16 of Fig. 3, and the processing of step S137 is basically the same as the processing of step S6 of the flowchart of Fig. 4, and thus detailed description thereof will be omitted.
[00242] In step S138, the synthesizing circuit 48 synthesizes the decoded low-frequency signal from the low-frequency decoding circuit 42 and the decoded high-frequency signal from the decoded high-frequency signal generating circuit 47 to transmit them as an output signal.
[00243] According to the above processing, the estimation coefficient of the high frequency subband energy at the time of decoding is employed according to features of the difference between the pseudohigh frequency subband energy calculated in advance at the time of encoding and the real high frequency subband energy and thus estimation accuracy of a high frequency subband energy at the time of decoding can increase and, consequently, music signals can be reproduced with higher quality of sound.
[00244] Also, according to the above processing, information to generate a high-frequency signal included in the code sequence is only the pseudo-high-frequency subband energy difference ID, and thus the decoding processing can be effectively carried out.
[00245] Although a description has been made in relation to the encoding processing and decoding processing in which the present invention has been applied, below, a description will be made in relation to a technique to calculate the representative vector of each of the multiple groupings in the characteristic space of the pseudo-high-frequency sub-band energy difference defined in advance in the high-frequency encoding circuit 37 of the encoding device 30 of Fig. 11 and a decoded high-frequency sub-band energy estimation coefficient to be transmitted by the high-frequency decoding circuit 45 of the decoding device 40 of Fig. 13. [Technique of Calculation of Vector Representatives of Multiple Clusters in Characteristic Space of Pseudohigh-Frequency Subband Energy Difference and Sub-Band Energy Estimation Coefficient decoded high frequency band corresponding to Each Cluster]
[00246] As a method to obtain vectors representative of the multiple clusters and a decoded high frequency subband energy estimation coefficient of each cluster, a coefficient needs to be prepared to estimate a high frequency subband energy at the time of decoding with high precision according to a pseudohigh frequency subband energy difference vector to be calculated at the time of encoding. Therefore, a technique will be applied to perform learning using a broadband supervisory signal in advance, and to determine these based on their learning outcomes. [Example of Functional Configuration of Coefficient Learning Device]
[00247] Figure 15 illustrates an example of a functional configuration of a coefficient learning device to perform learning of vectors representing the multiple clusters and a decoded high frequency subband energy estimation coefficient of each cluster.
[00248] It is desirable that, of a wideband supervisor signal to be input into the coefficient learning device 50 of Fig. 15, a signal component equal to or less than a cutoff frequency to be defined in the low pass filter of the device encoding 30 is a decoded low-frequency signal obtained by an input signal in the encoding device 30 which passes through the low-pass filter 31, encoded by the low-frequency encoding circuit 32 and further decoded by the low-frequency decoding circuit 42 of the decoding device 40.
[00249] The coefficient learning device 50 is configured by a low pass filter 51, a subband division circuit 52, a resource quantity calculation circuit 53, a subband energy calculation circuit of pseudo high frequency 54, a pseudo high frequency subband energy difference calculating circuit 55, a pseudo high frequency subband energy difference grouping circuit 56, and a coefficient estimating circuit 57.
[00250] Note that the low pass filter 51, the subband division circuit 52, the resource quantity calculation circuit 53 and the pseudo high frequency subband energy calculation circuit 54 of the coefficient learning device 50 of Fig. 15 have basically the same configuration and function as the low pass filter 31, the subband division circuit 33, the resource quantity calculation circuit 34, and the pseudo high frequency subband energy calculation circuit 35 of figure 11, respectively, and thus description of these will be omitted.
[00251] Specifically, the pseudo-high-frequency sub-band power difference calculation circuit 55 has the same configuration and function as the pseudo-high-frequency sub-band power difference calculation circuit 36 of Fig. 11 and not only supplies the pseudo-high-frequency sub-band energy difference calculated to the pseudo-high-frequency sub-band energy difference grouping circuit 56, but also supplies a high-frequency sub-band energy to be calculated at the time of calculation of the pseudohigh frequency subband energy difference to the coefficient estimation circuit 57.
The pseudo-high-frequency sub-band energy difference grouping circuit 56 subjects a pseudo-high-frequency sub-band energy difference vector obtained from the pseudo-high-frequency sub-band energy difference from the calculation circuit from the difference of the pseudohigh frequency subband energy 55 to the cluster to calculate a representative vector in each cluster.
[00253] The coefficient estimating circuit 57 calculates a high frequency subband energy estimating coefficient for each grouping, subject to grouping by the pseudohigh frequency subband energy difference 56 grouping circuit, on the basis of in the high frequency subband energy from the pseudo high frequency subband energy difference calculation circuit 55 and in one or multiple resource quantities from the resource quantity calculation circuit 53. [Coefficient Learning Device Coefficient Learning Processing]
[00254] Next, coefficient learning processing by the coefficient learning device 50 of Figure 15 will be described in relation to the flowchart of Figure 16.
[00255] Note that the processing of steps S151 to S155 of the flowchart of figure 16 is the same as the processing of steps S111 and S113 to S116 of the flowchart of figure 12, except that a signal to be input into the coefficient 50 learning device is a broadband supervisor signal and thus its description will be omitted.
[00256] Specifically, in step S156, the pseudo-high-frequency sub-band energy difference clustering circuit 56 calculates the vector representative of each cluster by a large number of pseudo-high-frequency sub-band energy difference vectors ( many time intervals) obtained from the pseudo-high-frequency sub-band energy difference from the pseudo-high-frequency sub-band energy difference calculation circuit 55 which is grouped into 64 clusters, for example. As an example of a grouping technique, grouping according to the k-means method can be applied, for example. The pseudohigh frequency subband energy difference clustering circuit 56 takes the center of gravity vector of each cluster obtained as a result of performing clustering according to the k-means method as the representative vector of each cluster. Note that one technique for grouping and the number of groupings is not restricted to those mentioned above, and another technique can be employed.
[00257] Also, the pseudo-high-frequency sub-band energy difference grouping circuit 56 measures the distance with the 64 representative vectors using a pseudo-high-frequency sub-band energy difference vector obtained from the energy difference. from the pseudo-high-frequency sub-band energy difference calculation circuit 55 in time interval J to determine a CID(J) index of a cluster to which a representative vector to provide the shortest distance belongs. Now, say the CID(J) index takes an integer from 1 to the number of clusters (64 in this example). The pseudo-high frequency subband energy difference grouping circuit 56 transmits a representative vector in this way and also supplies the CID(J) index to the coefficient estimation circuit 57.
[00258] In step S157, the coefficient estimation circuit 57 performs, of a large number of combinations between (eb - sb) energies of the high frequency subband and resource quantities supplied from the calculation circuit of the energy difference of pseudohigh frequency subband 55 and resource quantity calculation circuit 53 in the same time interval, calculating a decoded high frequency subband energy estimation coefficient in each cluster for each cluster (which belongs to the same grouping) with the same CID(J) index. Now, say that the technique for calculating a coefficient by the coefficient estimating circuit 57 is the same as the technique used by the coefficient estimating circuit 24 in the coefficient learning device 20 of Figure 9, but needless to say, another technique can be employed.
[00259] According to the aforementioned processing, learning the representative vector of each of the multiple clusters in the characteristic space of the pseudo-high-frequency subband energy difference defined in advance in the high-frequency encoding circuit 37 of the encoding device 30 of Fig. 11 and a decoded high frequency subband energy estimation coefficient to be transmitted by the high frequency decoding circuit 45 of the decoding device 40 of Fig. 13 and, in this way, suitable output results can be obtained for several input signals to be inputted to the encoding device 30 and for various input code sequences to be inputted to the decoding device 40, and consequently music signals can be reproduced with higher sound quality.
[00260] Additionally, in relation to encoding and decoding for signals, coefficient data for calculating a high frequency subband energy in the pseudohigh frequency subband energy calculation circuit 35 of the encoding device 30 or in the decoded high frequency subband energy calculation circuit 46 of decoding device 40 can be handled as follows. Specifically, it is considered that different coefficient data are employed according to the type of an input signal, and the coefficient of these can also be recorded in the header of a code sequence.
[00261] For example, improvement in coding efficiency can be accomplished by exchanging the coefficient data using a signal such as speech, jazz or the like.
[00262] Figure 17 illustrates a code sequence thus obtained.
[00263] A code sequence A of Fig. 17 is coded speech, in which α coefficient data ideal for speech is recorded in a header.
[00264] On the other hand, the code sequence B of figure 17 is jazz encoded, and ideal β coefficient data for jazz is recorded in the header.
[00265] An arrangement can be made in which such multiple coefficient data are prepared by learning with the same type of music signals, with the encoding device 30, the coefficient data of these are selected with genre information recorded in the header of an input signal. Alternatively, a gender can be determined by performing signal waveform analysis to select coefficient data. That is, the signal gender analysis technique is not restricted to a particular technique.
[00266] Also, if computing time permits, an arrangement can be made in which the aforementioned learning device is housed in the encoding device 30, processing is performed using a coefficient dedicated to the signals and, as illustrated in a sequence of code C in figure 17, the coefficient of these is finally recorded in the header.
[00267] Advantages for using this technique will be described below.
[00268] Regarding the form of a high frequency subband energy, there are many similar parts in an input signal. Learning of a coefficient for estimating a high frequency subband energy is individually performed for each input signal using this characteristic that many input signals have and thus redundancy due to the existence of similar parts of a subband energy. high frequency band can be reduced and coding efficiency can be increased. Also, estimating a high frequency subband energy can be performed with higher accuracy compared to statistical learning of a coefficient to estimate a high frequency subband energy using multiple signals.
[00269] Also in this way an arrangement can be made in which coefficient data to be learned from an input signal at the time of encoding is input once for several frames. <3. Third Mode> [Example of Functional Configuration of Encoding Device]
[00270] Note that, although description has been made in which the pseudo-high-frequency subband energy difference ID is transmitted from the encoding device 30 to the decoding device 40 as high-frequency encoded data, a coefficient index for obtaining a decoded high frequency subband energy estimation coefficient can be taken as high frequency encoded data.
[00271] In a case like this, the encoding device 30 is configured as illustrated in Figure 18, for example. Note that, in figure 18, a part corresponding to the case of figure 11 is denoted with the same reference number, and description of this will be omitted as appropriate.
[00272] The encoding device 30 of Fig. 18 differs from the encoding device 30 of Fig. 11 in that a low frequency decoding circuit 39 is not provided, and other points are the same.
[00273] With the encoding device 30 of Fig. 18, the resource amount calculating circuit 34 calculates a low frequency subband energy as a resource amount using the low frequency subband signal supplied from the circuit. subband division 33 for supplying the pseudo-high frequency subband energy calculation circuit 35.
[00274] Also, with the pseudohigh frequency subband energy calculation circuit 55, multiple decoded high frequency subband energy estimation coefficients obtained in advance by regression analysis and coefficient indices to identify these coefficients of Estimates of the decoded high frequency subband energy are recorded in a correlated manner.
[00275] Specifically, multiple sets of a coefficient Aib(kb) and a coefficient Bib of each subband used to calculate the aforementioned Expression (2) are prepared in advance as multiple estimation coefficients of the decoded high frequency subband energy . For example, these coefficients Aib(kb) and Bibha were obtained by regression analysis using the least squares method with a low frequency subband energy as an explained variable and with a high frequency subband energy as a non-explanatory variable. With regression analysis, an input signal consisting of a low frequency subband signal and a high frequency subband signal is employed as a wideband supervisory signal.
[00276] The pseudo high frequency subband energy calculation circuit 35 calculates the pseudo high frequency subband energy of each subband on the high frequency side using the high subband energy estimation coefficient decoded frequency and the resource amount from the resource amount calculation circuit 34 for supplying the pseudo-high frequency subband energy difference calculation circuit 36.
[00277] The pseudo-high-frequency sub-band energy difference calculation circuit 36 compares a high-frequency sub-band energy obtained from the high-frequency sub-band signal supplied from the sub-division circuit -band 33 and the pseudo-high-frequency sub-band energy from the pseudo-high-frequency sub-band energy calculation circuit 35.
[00278] As a result of the comparison, the pseudo-high-frequency sub-band energy difference calculation circuit 36 supplies, of the multiple decoded high-frequency sub-band energy estimation coefficients, a coefficient index of a coefficient of estimation of the decoded high frequency subband energy according to which a pseudo high frequency subband energy which approximates the subband energy with the highest frequency was obtained, to the high frequency coding circuit 37 In other words, a coefficient index of a decoded high-frequency subband energy estimation coefficient is selected according to which a decoded high-frequency signal most closely approximates a high-frequency signal of a high-frequency signal. input to be played at the time of decoding, ie a true value is obtained. [Encoding Device Encoding Processing]
[00279] Next, encoding processing to be performed by the encoding device 30 of figure 18 will be described in relation to the flowchart of figure 19. Note that the processing of steps S181 to S183 is the same as the processing of steps S111 to S113 of figure 12 and thus description of this will be omitted.
[00280] In step S184, the resource amount calculation circuit 34 calculates a resource amount using the low frequency subband signal from the subband division circuit 33 to supply the power calculation circuit. pseudohigh frequency subband 35.
[00281] Specifically, the resource quantity calculation circuit 34 performs the calculation of the aforementioned Expression (1) to calculate, with respect to each subband ib (however, sb - 3 < ib < sb), an energy of sub -low frequency band power(ib, J) of frame J (however, 0 < J) as a resource quantity. That is, the low frequency subband energy power(ib, J) is calculated by converting a root mean square of the sample value of each sample of a low frequency subband signal that makes up the J frame into a logarithm.
[00282] In step S185, the pseudo-high-frequency sub-band energy calculation circuit 35 calculates a pseudo-high-frequency sub-band energy based on the resource amount supplied from the resource amount calculation circuit 34 to supply to the pseudo-high frequency subband energy difference calculation circuit 36.
[00283] For example, the pseudo-high-frequency subband energy calculation circuit 35 performs calculation of the aforementioned Expression (2) using the coefficient Aib(kb) and the coefficient Bib recorded in advance as sub-energy estimation coefficients. decoded high frequency band and the power(kb, J) low frequency subband energy (however, sb - 3<kb<sb) to calculate a powerest(ib, J) pseudohigh frequency subband energy.
[00284] Specifically, the low frequency subband energy power(kb, J) of each subband on the low frequency side supplied as a resource amount is multiplied by the coefficient Aib(kb) for each subband, the Bib coefficient is further added to the sum of the low frequency subband energies multiplied by the coefficient and is taken as a powerest(ib, J) pseudohigh frequency subband energy. This pseudo-high-frequency subband energy is calculated with respect to each subband on the high-frequency side whose index is sb + 1through eb.
[00285] Also, the pseudo-high-frequency sub-band energy calculation circuit 35 performs calculation of a pseudo-high-frequency sub-band energy for each pre-recorded decoded high-frequency sub-band energy estimation coefficient. For example, let's say that K decoded high-frequency subband energy estimation coefficients whose indices are 1 through K (however, 2 <K) were prepared in advance. In this case, the pseudohigh frequency subband energy of each subband is calculated for each K decoded high frequency subband energy estimation coefficients.
[00286] In step S186, the pseudo-high-frequency sub-band energy difference calculation circuit 36 calculates the pseudo-high-frequency sub-band energy difference based on the high-frequency sub-band signal from the high frequency subband signal. sub-band 33 and in the pseudo-high-frequency sub-band energy from the pseudo-high-frequency sub-band energy calculation circuit 35.
[00287] Specifically, the pseudo-high-frequency sub-band energy difference calculation circuit 36 performs the same calculation as the above-mentioned Expression (1) with respect to the high-frequency sub-band signal coming from the sub-band division circuit. band 33 to calculate a high frequency subband energy power(ib, J) in frame J. Note that in the present embodiment, say every subband of a low frequency subband signal and every subband band of a high frequency subband signal are identified with an index ib.
[00288] Next, the pseudo-high-frequency subband energy difference calculation circuit 36 performs the same calculation as above-mentioned Expression (14) to obtain the difference between the high-frequency subband energy power(ib, J) and the powerest(ib,J) pseudohigh frequency subband energy in frame J. Thus, the powersest(ib,J) pseudohigh frequency subband energy is obtained with respect to each subband on the high side. frequency whose index is sb + 1through eb for each decoded high frequency subband energy estimation coefficient.
[00289] In step S187, the pseudo-high-frequency sub-band energy difference calculation circuit 36 calculates the following Expression (15) for each decoded high-frequency sub-band energy estimate coefficient to calculate the sum of the squares of the pseudohigh frequency subband energy difference. [Mathematical Expression 15]

[00290] Note that, in Expression (15), the sum of squared difference E(J, id) indicates the sum of squares of the difference of the pseudohigh frequency subband energy of frame J obtained with respect to a coefficient of decoded high frequency subband energy estimate whose coefficient index is id. Also, in Expression (15), powerdiff(ib, J, id) indicates the difference of the pseudohigh-frequency subband energy powerdiff(ib, J) of frame J from a subband whose index is ib obtained with respect to a decoded high frequency subband energy estimation coefficient whose coefficient index is id. The sum of squared difference E(J, id) is calculated against the K decoded high frequency subband energy estimation coefficients.
[00291] The sum of squared difference E(J, id) thus obtained indicates a degree of similarity between the high frequency subband energy calculated from the real high frequency signal and the pseudo high subband energy frequency calculated using a decoded high frequency subband energy estimation coefficient whose coefficient index is id.
[00292] Specifically, the sum of squares difference E(J, id) indicates error of an estimated value in relation to a true value of a pseudohigh frequency subband energy. In this way, the smaller the sum of squares difference E(J, id), a decoded high-frequency signal closer and closer to the actual high-frequency signal is obtained by the calculation using a sub-energy estimation coefficient. high frequency band decoded. In other words, it can be said that a decoded high frequency subband energy estimation coefficient according to which the sum of squares difference E(J, id) becomes minimal is an estimation coefficient plus suitable for frequency band expansion processing to be performed at the time of decoding the output code sequence.
[00293] Therefore, the pseudo-high frequency subband energy difference calculation circuit 36 selects, from the K sums of squares of difference E(J, id), the sum of squares difference according to which the value becomes. if minimal and supplies a coefficient index that indicates a decoded high frequency subband energy estimate coefficient corresponding to the sum of the squared difference thereof to the high frequency encoding circuit 37.
[00294] In step S188, the high-frequency encoding circuit 37 encodes the coefficient index supplied from the pseudo-high-frequency subband energy difference calculation circuit 36 and supplies the high-frequency encoded data obtained therefrom from this to the multiplexing circuit 38.
[00295] For example, in step S188, entropy encoding is performed on the coefficient index. Thus, the volume of information of the high frequency encoded data transmitted to the decoding device 40 can be compressed. Note that the high-frequency encoded data can be any information, as long as the optimal decoded high-frequency subband energy estimate coefficient is obtained from the information, for example, the coefficient index can become encoded data in high frequency without change.
[00296] In step S189, the multiplexing circuit 38 multiplexes the high-frequency encoded data obtained from the low-frequency encoding circuit 32 and the high-frequency encoded data supplied from the high-frequency encoding circuit 37, transmits an exit code sequence obtained as a result of this and the encoding processing ends.
[00297] In this way, the high-frequency encoded data obtained by coefficient index encoding is transmitted as an output code sequence along with the low-frequency encoded data and, in this way, a subband energy estimation coefficient of decoded high frequency more suitable for frequency band expansion processing can be obtained in the decoding device 40 receiving the input of this output code sequence. Thus, signals with higher sound quality can be obtained. [Example of Functional Configuration of the Decoding Device]
[00298] Also, the decoding device 40 which inputs the output code sequence transmitted from the encoding device 30 of Fig. 18 as an input code sequence and decoding it is configured as shown in Fig. 20, for example . Note that, in figure 20, a part corresponding to the case of figure 20 is denoted with the same reference number, and description of this will be omitted.
[00299] The decoding device 40 of Fig. 20 is the same as the decoding device 40 of Fig. 13, wherein the decoding device 40 is configured by the demultiplexing circuit 41 to the downmix circuit 48, but differs from the decoding device 40 of Fig. 13 , in which the decoded low frequency signal from the low frequency decoding circuit 42 is not supplied to the resource quantity calculation circuit 44.
[00300] In the decoding device 40 of Fig. 20, the high-frequency decoding circuit 45 pre-recorded the estimate coefficient of the decoded high-frequency subband equal to the estimate coefficient of the decoded high-frequency subband that the circuit Pseudo-high frequency subband energy calculation method 35 of Fig. 18 records. Specifically, the set of coefficient Aib(kb) and coefficient Bib that serve as decoded high frequency subband energy estimation coefficients obtained in advance by regression analysis was recorded in a manner with a coefficient index.
[00301] The high-frequency decoding circuit 45 decodes the high-frequency encoded data supplied from the demultiplexing circuit 41 and supplies a decoded high-frequency subband energy estimate coefficient indicated by the coefficient index obtained as a result of this to decoded high frequency subband energy calculation circuit 46. [Decoding Device Decoding Process]
[00302] Next, the decoding processing to be performed by the decoding device 40 of figure 20 will be described in relation to the flowchart of figure 21.
[00303] This decoding processing is started when the output code sequence transmitted from the encoding device 30 is supplied to the decoding device 40 as an input code sequence. Note that the processing of steps S211 to S213 is the same as the processing of steps S131 to S133 of Fig. 14, and thus description of this will be omitted.
[00304] In step S214, the resource amount calculation circuit 44 calculates a resource amount using the decoded low frequency subband signal from the subband division circuit 43 and supplies it to the calculation circuit of the decoded high frequency subband energy 46. Specifically, the resource quantity calculation circuit 44 performs the calculation of the aforementioned Expression (1) to calculate the low frequency subband energy power(ib, J) in the frame J (however, 0 < J) with respect to each ib subband on the low frequency side as a resource quantity.
[00305] In step S215, the high-frequency decoding circuit 45 performs decoding of the high-frequency encoded data supplied from the demultiplexing circuit 41 and supplies a decoded high-frequency subband energy estimation coefficient indicated by a coefficient index obtained as a result of this to the decoded high frequency subband energy calculation circuit 46. That is, from the multiple decoded high frequency subband energy estimation coefficients recorded in advance in the high frequency decoding circuit 45, a decoded high frequency subband energy estimate coefficient indicated by the coefficient index obtained by decoding is transmitted.
[00306] In step S216, the decoded high frequency subband energy calculation circuit 46 calculates a decoded high frequency subband energy based on the resource amount supplied from the resource amount calculation circuit 44 and in the estimation coefficient of the decoded high-frequency subband energy supplied from the high-frequency decoding circuit 45, and supplies it to the decoded high-frequency signal generating circuit 47.
[00307] Specifically, the decoded high-frequency subband energy calculation circuit 46 performs the calculation of the above-mentioned Expression (2) using the coefficient Aib(kb) and the coefficient Bib which serve as sub energy estimation coefficients. -decoded high frequency band and the low frequency subband energy power(kb, J) (however, sb - 3 < kb < sb) which serves as a resource quantity to calculate a high subband energy decoded frequency. Thus, a decoded high frequency subband energy is obtained with respect to each subband on the high frequency side whose index is sb + 1 to eb.
[00308] In step S217, the decoded high frequency signal generating circuit 47 generates a decoded high frequency signal based on the decoded low frequency subband signal supplied from the subband division circuit 43 and in the decoded high frequency subband energy supplied from the decoded high frequency subband energy calculation circuit 46.
[00309] Specifically, the decoded high frequency signal generating circuit 47 performs the calculation of the aforementioned Expression (1) using the decoded low frequency subband signal to calculate a low frequency subband energy with respect to each subband on the low frequency side. The decoded high-frequency signal generating circuit 47 performs the calculation of the aforementioned Expression (3) using the obtained low-frequency subband energy and the decoded high-frequency subband energy obtained to calculate the amount of gain G(ib, J) for each subband on the high frequency side.
[00310] Additionally, the decoded high frequency signal generating circuit 47 performs the calculations of the aforementioned Expression (5) and Expression (6) using the amount of gain G(ib, J) and the low subband signal frequency decoded to generate a high frequency subband signal x3(ib,n) with respect to each subband on the high frequency side.
[00311] Specifically, the decoded high frequency signal generating circuit 47 subjects a decoded low frequency subband signal x(ib, n) to amplitude modulation according to a ratio of a subband energy of low frequency and a decoded high frequency subband energy, and further subject a decoded low frequency subband signal x2(ib,n) obtained therefrom to frequency modulation. Thus, a frequency component signal in a subband on the low frequency side is converted to a frequency component signal on a subband on the high frequency side to obtain a high frequency subband signal x3( ib,n).
[00312] In this way, the processing to obtain a high frequency subband signal in each subband is, in more detail, the following processing.
[00313] Say that four consecutively arranged sub-bands in a frequency region will be referred to as a band block and the frequency band has been divided such that a band block (hereinafter, particularly referred to as low block frequency) is configured by four subbands whose indices are sbaté sb - 3 on the low frequency side. At this time, for example, a band consisting of subbands whose indices on the high frequency side are sb + 1 to sb + 4 is taken as a band block. Now, in the following, the high frequency side, that is, a band block consisting of a subband whose index is equal to or greater than sb + 1, will be particularly referred to as a high frequency block.
[00314] Now, let it be said that attention is paid to a subband that constitutes a block of high frequency to generate a high frequency subband signal of the subband thereof (hereinafter, referred to as subband of interest). First, the decoded high frequency signal generating circuit 47 identifies a subband of a low frequency block with the same position relationship as a position of the subband of interest in the high frequency block.
[00315] For example, in the event that the index of the sub-band of interest is sb + 1, the sub-band of interest is a band with the lowest frequency of the high-frequency block and, thus, the sub- band of a low frequency block with the same position ratio as the subband of interest is a subband whose index is sb - 3.
[00316] In this way, in the event that the subband of a low frequency block with the same position ratio of the subband of interest has been identified, a high frequency subband signal of the subband of interest is generated using the low frequency subband energy of the subband thereof, the decoded low frequency subband signal and the decoded high frequency subband energy of the subband of interest.
[00317] Specifically, the decoded high frequency subband energy and the low frequency subband energy are replaced by Expression (3), and a gain amount according to a ratio of these energies is calculated. The decoded low frequency subband signal is multiplied by the calculated gain amount and additionally the decoded low frequency subband signal multiplied by the gain amount is subjected to frequency modulation by calculating Expression (6) and is taken as a high frequency subband signal of the subband of interest.
[00318] According to the aforementioned processing, the high frequency subband signal of each subband on the high frequency side is obtained. In response to this, the decoded high-frequency signal generating circuit 47 further performs the calculation of the aforementioned Expression (7) to obtain the sum of the obtained high-frequency subband signals and to generate a decoded high-frequency signal. The decoded high frequency signal generating circuit 47 supplies the obtained decoded high frequency signal to the synthesizing circuit 48 and processing proceeds from step S217 to step S218.
[00319] In step S218, the synthesizing circuit 48 synthesizes the decoded low-frequency signal from the low-frequency decoding circuit 42 and the decoded high-frequency signal from the decoded high-frequency signal generating circuit 47 to transmit this as an output signal. Afterwards, decoding processing ends.
[00320] As described above, according to the decoding device 40, a coefficient index is obtained from high-frequency encoded data obtained by demultiplexing the input code sequence and a decoded high-frequency subband energy is calculated using a decoded high frequency subband energy estimation coefficient indicated by the coefficient index thereof, and in this way the estimation accuracy of a high frequency subband energy can be increased. Thus, music signals can be reproduced with better sound quality. <4. Fourth Mode> [Encoding Device Encoding Processing]
[00321] Also, although description has been made so far in relation to a case where only one coefficient index is included in the high frequency encoded data as an example, other information may be included in the high frequency encoded data.
[00322] For example, if an arrangement is made in which a coefficient index is included in the high-frequency encoded data, a decoded high-frequency subband energy estimation coefficient may be known at the side of the decoding device 40 whereby a decoded high frequency subband energy closer to a high frequency subband energy of the actual high frequency signal is obtained.
[00323] However, a difference is caused between the real high frequency subband energy (true value) and the decoded high frequency subband energy (estimated value) obtained on the side of the decoding device 40 in general, by the same value of the powerdiff(ib,J) pseudohigh frequency subband energy difference calculated by the pseudohigh frequency subband energy difference calculation circuit 36.
[00324] Therefore, if an arrangement is made in which not only a coefficient index, but also the pseudohigh frequency subband energy difference between the subbands are included in the high frequency encoded data, gross error from a decoded high frequency subband energy to the real high frequency subband energy can be known at the side of the decoding device 40. Thus, the estimation accuracy for a high frequency subband energy can be known increase using this error.
[00325] The following will be described in relation to encoding processing and decoding processing in the event that pseudo-high-frequency subband energy difference is included in the high-frequency encoded data in relation to the flowcharts of Figure 22 and of figure 23.
[00326] First, the encoding processing to be performed by the encoding device 30 of figure 18 will be described in relation to the flowchart of figure 22. Note that the processing from step S241 to step S246 is the same as the processing from step S181 to step S186 of figure 19 and thus description of this will be omitted.
[00327] In step S247, the pseudo-high frequency subband energy difference calculation circuit 36 performs the calculation of Expression (15) to calculate the sum of squares difference E(J, id) for each estimation coefficient of the decoded high frequency subband energy.
[00328] The pseudo-high frequency subband energy difference calculation circuit 36 selects, from the sum of squared difference E(J, id), the sum of squared difference according to which the value becomes minimum and supplies a coefficient index which indicates a decoded high frequency subband energy estimate coefficient corresponding to the sum of the squared difference from this to the high frequency encoding circuit 37.
[00329] Additionally, the pseudo-high-frequency sub-band energy difference calculation circuit 36 supplies the pseudo-high-frequency sub-band energy difference powerdiff(ib, J) of the sub-bands, obtained with respect to an estimation coefficient of the decoded high frequency subband energy corresponding to the selected sum of squares difference, to the high frequency encoding circuit 37.
[00330] In step S248, the high-frequency encoding circuit 37 encodes the coefficient index and the pseudo-high-frequency sub-band energy difference supplied from the pseudo-high-frequency sub-band energy difference calculation circuit 36 and supplies the high frequency encoded data obtained therefrom to the multiplexing circuit 38.
[00331] Thus, the difference of the pseudo-high-frequency sub-band energy of the sub-bands on the high-frequency side whose indices are sb + 1 to eb, that is, estimation error of a high-frequency sub-band energy is supplied to the decoding device 40 as high frequency encoded data.
[00332] In the event that the high-frequency encoded data was obtained, thereafter, the processing of step S249 is performed and the encoding processing ends, but the processing of step S249 is the same as the processing of step S189 of Figure 19 and, in this way, its description will be omitted.
[00333] As described above, if an arrangement is made in which the pseudo-high-frequency sub-band energy difference is included in the high-frequency encoded data, with the decoding device 40, the estimation accuracy of a sub-band energy -high frequency band can increase further and music signals with higher sound quality can be obtained. [Decoding Device Decoding Process]
[00334] Next, the decoding processing to be performed by the decoding device 40 of figure 20 will be described in relation to the flowchart of figure 23. Note that the processing from step S271 to step S274 is the same as the processing from step S211 to step S214 and thus description of it will be omitted.
[00335] In step S275, the high-frequency decoding circuit 45 performs the decoding of the high-frequency encoded data supplied from the demultiplexing circuit 41. Then, the high-frequency decoding circuit 45 supplies an energy estimation coefficient decoded high-frequency subband indicated by a coefficient index obtained by decoding and the difference of the pseudo-high-frequency subband energy of the subbands obtained by decoding to the decoded high-frequency subband energy calculation circuit 46.
[00336] In step S276, the decoded high frequency subband energy calculation circuit 46 calculates a decoded high frequency subband energy based on the resource amount supplied from the resource amount calculation circuit 44 and in the estimation coefficient of the decoded high-frequency subband energy supplied from the high-frequency decoding circuit 45. Note that, in step S276, the same processing as in step S216 of Fig. 21 is performed.
[00337] In step S277, the decoded high-frequency subband energy calculation circuit 46 adds the difference of the pseudo-high-frequency subband energy supplied from the high-frequency decoding circuit 45 into the subband energy of decoded high frequency and supplies it to the decoded high frequency signal generating circuit 47 as the final decoded high frequency subband energy. That is, the difference of the pseudohigh frequency subband energy of the same subband is added to the calculated decoded high frequency subband energy of each subband.
[00338] Subsequently, the processing of step S278 to step S279 is performed and the decoding processing ends, but these processes are the same as those of steps S217 and S218 of figure 21 and, in this way, their description will be omitted.
[00339] In this way, the decoding device 40 obtains a coefficient index and the pseudo-high frequency subband energy difference from the high-frequency encoded data obtained by demultiplexing the input code sequence. Then, the decoding device 40 calculates a decoded high-frequency subband energy using the decoded high-frequency subband energy estimation coefficient indicated by the coefficient index and the pseudo-high-frequency subband energy difference. . Thus, the estimation accuracy for a high frequency subband energy can be increased and music signals can be reproduced with higher sound quality.
[00340] Note that the difference between the estimated values of the high frequency subband energy generated between the encoding device 30 and the decoding device 40, i.e. the difference between the pseudo-high frequency subband energy and the energy decoded high-frequency subband (hereafter referred to as estimated difference between devices) can be taken into account.
[00341] In a case like this, for example, the pseudo-high-frequency sub-band energy difference that serves as high-frequency encoded data is corrected with the estimated difference between the devices, or the sub-band energy difference High-frequency pseudo-frequency is included in the high-frequency encoded data and, on the side of the decoding device 40, the pseudo-high-frequency sub-band energy difference is corrected with the estimated difference between the devices. Additionally, an arrangement can be made in which, on the side of the decoding device 40, the estimated difference between the devices is recorded and the decoding device 40 adds the estimated difference between the devices in the pseudo-high frequency subband energy difference. to perform correction. Thus, a decoded high frequency signal closer to the real high frequency signal can be obtained. <5. Fifth Mode>
[00342] Note that a description has been made in which, in the coding device 30 of Figure 18, the pseudo-high frequency subband energy difference calculation circuit 36 selects the ideal coefficient index among multiple coefficient indices with the sum of the difference of squares E(J, id) as an index, but a coefficient index can be selected using an index other than the sum of the difference of squares.
[00343] For example, an evaluated value can be used in which the residual mean square value, the maximum value, the mean value and the like between a high frequency subband energy and a pseudohigh frequency subband energy are taken in consideration. In such a case, the encoding device 30 of Fig. 18 performs the encoding processing illustrated in the flowchart of Fig. 24.
[00344] Next, the encoding processing by the encoding device 30 will be described in relation to the flowchart of figure 24. Note that the processing from step S301 to step S305 is the same as the processing from step S181 to step S185 of figure 19 , and description of this will be omitted. In the event that processing from step S301 to step S305 was performed, the pseudohigh frequency subband energy of each subband was calculated for each K decoded high frequency subband energy estimation coefficients.
[00345] In step S306, the pseudo-high frequency subband energy difference calculation circuit 36 calculates the evaluated value Res(id, J) with the current frame J which serves as an object to be processed which is employed for each K decoded high frequency subband energy estimation coefficients.
[00346] Specifically, the pseudo-high-frequency sub-band energy difference calculation circuit 36 performs the same calculation as the aforementioned Expression (1) using the high-frequency sub-band signal of each sub-band supplied from the subband division circuit 33 for calculating the high frequency subband energy power(ib, J) in frame J. Note that, in the present embodiment, every subband of a low frequency subband signal and every subband of a high frequency subband signal can be identified using the ib index.
[00347] In the event that the high frequency subband energy power(ib, J) is obtained, the pseudohigh frequency subband energy difference calculation circuit 36 calculates the following Expression (16) to calculate a root mean square value. residual Resstd(id, J). [Mathematical Expression 16]

[00348] Specifically, the difference between the high frequency subband energy power(ib, J) and the pseudohigh frequency subband energy powerest(ib, id, J) in frame J is obtained with respect to each sub -band on the high frequency side whose index is sb + 1up to eb and the sum of the squares of this difference is taken as the residual root mean square Resstd(id, J). Note that the powerest(ib, id, J) pseudohigh frequency subband energy indicates a pseudohigh frequency subband energy in frame J of a subband whose index is ib, obtained with respect to the coefficient of estimation of decoded high frequency subband energy whose coefficient index is id.
[00349] Next, the pseudo-high frequency subband energy difference calculation circuit 36 calculates the following Expression (17) to calculate the maximum residual value Resmax(id, J). [Mathematical Expression 17]

[00350] Note that, in Expression (17), maxib{|power(ib, J) - powersest(ib, id, J)|} indicates the maximum of the absolute values of the difference between the high frequency subband energy power(ib, J) of each subband whose index is sb + 1up to eb is the powerest(ib, id, J) pseudohigh frequency subband energy. In this way, the maximum value of the absolute values of the difference between the high frequency subband energy power(ib, J) and the pseudohigh frequency subband energy powersest(ib, id, J) in frame J is taken as a maximum residual value Resmax(id, J).
[00351] Also, the pseudo-high frequency subband energy difference calculation circuit 36 calculates the following Expression (18) to calculate the residual average value Resave(id, J). [Mathematical Expression 18]

[00352] Specifically, the difference between the high frequency subband energy power(ib, J) and the pseudo high frequency subband energy powerest(ib, id, J) in frame J is obtained with respect to each sub -band on the high frequency side whose index is sb + 1 to eb, and the sum of their difference is obtained. The absolute value of a value obtained by dividing the sum of the difference obtained by the number of subbands (eb - sb) on the high frequency side is taken as a residual average Resave(id, J) value. This Resave(id, J) residual mean value indicates the magnitude of a mean value of the estimated error of the subbands with the signal being taken into account.
[00353] Additionally, in the event that the residual root mean value Resstd(id, J), the maximum residual value Resmax(id, J) and the residual mean value Resave(id, J) are obtained, the calculation circuit of the pseudohigh frequency subband energy difference 36 computes the following Expression (19) to calculate the final evaluated value Res(id, J). [Mathematical Expression 19]

[00354] Specifically, the residual root mean value Resstd(id, J), the residual maximum value Resmax(id, J) and the residual mean value Resave(id, J) are added with weight to obtain the final evaluated value Res( id, J). Note that in Expression (19), Wmax and Wave are weights determined in advance and examples of these are Wmax = 0.5 and Wave = 0.5.
[00355] The pseudo-high-frequency sub-band energy difference calculation circuit 36 performs the aforementioned processing to calculate the evaluated value Res(id, J) for each K decoded high-frequency sub-band energy estimation coefficients , that is, for every K id coefficient indices.
[00356] In step S307, the pseudo-high frequency subband energy difference calculation circuit 36 selects the coefficient index id based on the evaluated value Res(id, J) for each coefficient index id obtained.
[00357] The evaluated value Res(id, J) obtained in the aforementioned processing indicates a degree of similarity between the high frequency subband energy calculated from the real high frequency signal and the pseudo high frequency subband energy calculated using a decoded high frequency subband energy estimation coefficient whose coefficient index is id, that is, it indicates the magnitude of the estimated error of a high frequency component.
[00358] In this way, the smaller the evaluated value Res(id, J), the closer the real high-frequency signal is a decoded high-frequency signal obtained by calculating with a subband energy estimation coefficient of high frequency decoded. Therefore, the pseudo-high frequency subband energy difference calculation circuit 36 selects, from the K evaluated values Res(id, J), an evaluated value according to which the value becomes minimum and supplies an index of coefficient indicating a decoded high frequency subband energy estimation coefficient corresponding to the evaluated value thereof to the high frequency coding circuit 37.
[00359] In the event that the coefficient index has been transmitted to the high-frequency encoding circuit 37, thereafter, the processes of step S308 and step S309 are performed, and the encoding processing ends, but these processes are the same of step S188 and step S189 of Fig. 19 and thus description of these will be omitted.
[00360] As described above, in the coding device 30, the evaluated value Res(id, J) calculated from the root mean square Resstd(id, J), the maximum residual value Resmax(id, J) and the value Resave(id, J) residual averages are employed, and a coefficient index of the ideal decoded high frequency subband energy estimation coefficient is selected.
[00361] In the event that the evaluated value Res(id, J) is employed, if compared to the case of employing the sum of squares difference, the estimation accuracy of a high frequency subband energy can be evaluated using a lot more rating scales and in this way a more suitable decoded high frequency subband energy estimation coefficient can be selected. Thus, in the decoding device 40 that receives input from an output code sequence, a decoded high frequency subband energy estimation coefficient better suited to frequency band expansion processing can be obtained and signals with higher quality of sound can be obtained. <Modification 1>
[00362] Also, in the event that the coding processing described above is performed for each frame of an input signal, with a constant region where there is little temporal fluctuation in relation to the high frequency subband energies of the subbands in the High frequency side of the input signal, a different coefficient index can be selected for each of the continuous frames.
[00363] Specifically, with consecutive frames constituting a constant region of the input signal, the high frequency subband energies of the frames are almost the same, and thus the same coefficient index needs to be continuously selected with these frames . However, with a section of these continuous frames, the coefficient index to be selected changes for each frame, and as a result of this, high frequency audio components to be reproduced on the side of decoding device 40 may not be stationary. Consequently, with audio being played, unnatural sensations are perceptibly brought about.
[00364] Therefore, in the event of selecting a coefficient index in the coding device 30, results of estimating high frequency components in the temporally previous frame can be taken into account. In such a case, the encoding device 30 of Fig. 18 performs the encoding processing illustrated in the flowchart of Fig. 25.
[00365] Next, the encoding processing by the encoding device 30 will be described in relation to the flowchart of figure 25. Note that the processing from step S331 to step S336 is the same as the processing from step S301 to step S306 of figure 24 and, in this way, description of this will be omitted.
[00366] In step S337, the pseudo-high frequency subband energy difference calculation circuit 36 calculates an evaluated value ResP(id, J) using the last frame and the current frame.
[00367] Specifically, the pseudo-high-frequency sub-band energy difference calculation circuit 36 records, with respect to the temporally previous frame (J - 1) after the J-frame to be processed, a pseudo-high-frequency sub-band energy frequency of each subband obtained by using a high frequency subband energy estimation coefficient decoded with the finally selected coefficient index. The finally selected coefficient index mentioned here is a coefficient index encoded by the high frequency encoding circuit 37 and transmitted to the decoding device 40.
[00368] Next, let us say that the coefficient index id selected in the table (J - 1) is particularly idselected(J - 1). Also, considering that a pseudo-high frequency subband energy of a subband whose index is ib (however, sb + 1 < ib < eb), by using a high frequency subband energy estimation coefficient decoded coefficient index idselected(J - 1), powerest(ib, idselected(J - 1), J - 1) is obtained, and the description continues.
[00369] The pseudo-high frequency subband energy difference calculation circuit 36 first calculates the following Expression (20) to calculate an estimated residual root mean square ResPstd(id, J). [Mathematical Expression 20]

[00370] Specifically, with respect to each subband on the high frequency side whose index is sb + 1through eb, the difference between the pseudohigh frequency subband energy powersest(ib, idselected(J - 1), J - 1) of frame (J - 1) and the powerest(ib, id, J) pseudohigh frequency subband energy of frame J is obtained. The sum of the squares of their difference is taken as the estimated residual root mean square ResPstd(id, J). Note that the powerest(ib, id, J) pseudohigh frequency subband energy indicates a pseudohigh frequency subband energy of frame J of a subband whose index is ib, obtained with respect to an estimation coefficient of the decoded high frequency subband energy whose coefficient index is id.
[00371] This estimated residual root mean square ResPstd(id, J) is the sum of the squared difference of the pseudohigh frequency subband energies between temporally consecutive frames and, in this way, the smaller the estimated residual mean square ResPstd value (id, J), smaller is the temporal change of an estimated value of a high frequency component.
[00372] Next, the pseudo-high frequency subband energy difference calculation circuit 36 calculates the following Expression (21) to calculate the estimated maximum residual value ResPmax(id, J). [Mathematical Expression 21]

[00373] Note that, in Expression (21), maxib{|powerest(ib, idselected(J - 1), J - 1) - powersest(ib, id, J)|} indicates the maximum of the absolute values of the difference between the pseudohigh frequency subband energy powersest(ib, idselected(J - 1), J - 1) of each subband whose index is sb + 1up to eb and the pseudohigh frequency subband energy powersest(ib, id, J). In this way, the maximum value of the absolute values of the difference of the pseudo-high frequency subband energies between the temporally consecutive frames is taken as the estimated maximum residual value ResPmax(id, J).
[00374] The estimated maximum residual value ResPmax(id, J) indicates that the smaller the value, the closer the estimated results of the high frequency components between consecutive frames.
[00375] In the event that the estimated maximum residual value ResPmax(id, J) is obtained, then the pseudo-high frequency subband energy difference calculation circuit 36 calculates the following Expression (22) to calculate the average value estimated residual ResPave(id, J). [Mathematical Expression 22]

[00376] Specifically, with respect to each subband on the high frequency side whose index is sb + 1to eb, the difference between the pseudohigh frequency subband energy powersest(ib, idselected(J - 1), J - 1) of frame (J - 1) and the powerest(ib, id, J) pseudohigh frequency subband energy of frame J is obtained. The absolute value of a value obtained by dividing the sum of the subband difference by the number of subbands (eb - sb) on the high frequency side is taken as the estimated residual average value ResPave(id, J). This estimated residual mean value ResPave(id, J) indicates the magnitude of a mean value of the estimated difference of the subbands between frames, taking the signal into account.
[00377] Additionally, in the event that the estimated residual root mean value ResPstd(id, J), the estimated maximum residual value ResPmax(id, J) and the estimated residual mean value ResPave(id, J) are obtained, the circuit Pseudo-high frequency subband energy difference calculation method 36 calculates the following Expression (23) to calculate an evaluated value ResP(id, J). [Mathematical Expression 23]

[00378] Specifically, the estimated residual root mean value ResPstd(id, J), the estimated maximum residual value ResPmax(id, J) and the estimated residual mean value ResPave(id, J) are added with weight to obtain an evaluated value ResP(id, J). Note that in Expression (23), Wmax and Wave are weights determined in advance and examples of these are Wmax = 0.5 and Wave = 0.5.
[00379] In this way, after the evaluated value ResP(id, J) is calculated using the last frame and the current frame, processing proceeds from step S337 to step S338.
[00380] In step S338, the pseudo-high frequency subband energy difference calculation circuit 36 calculates the following Expression (24) to calculate the final evaluated value Resall(id, J). [Mathematical Expression 24]

[00381] Specifically, the obtained evaluated value Res(id, J) and the evaluated value ResP(id, J) are added with weight. Note that, in Expression (24), Wp(J) is weight to be defined by the following Expression (25), for example. [Mathematical Expression 25]

[00382] Also, powerr(J) of Expression (25) is a value to be determined by the following Expression (26). [Mathematical Expression 26]

[00383] This powerr(J) indicates the average of the difference of the energies of the high frequency subband of the frame (J - 1) and the frame J. Also, according to Expression (25), when the powerr(J) ) is a value in a predetermined range close to 0, the smaller the powerr(J) is, Wp(J) becomes a value closer to 1, and when the powerr(J) is greater than a value in a default range becomes 0.
[00384] Here, in the event that the powerr(J) is a value in a predetermined range close to 0, an average of the difference of the high frequency subband energies between consecutive frames is small to some degree. In other words, temporal fluctuation of a high frequency component of the input signal is small and, consequently, the current frame of the input signal is a constant region.
[00385] The more constant the high frequency component of the input signal, the weight Wp(J) becomes a value closer to 1 and, conversely, the more non-constant the high frequency component of the input signal , the weight Wp(J) becomes a value closer to 0. In this way, with the evaluated value Resall(id, J) indicated in Expression (24), the smaller the temporal fluctuation of a high frequency component of the input signal, the greater a contribution ratio of the rated value ResP(id, J) with a result of the comparison to a result of the estimation of a high frequency component in a last frame as a rating scale.
[00386] As a result of this, with a constant region of the input signal, a decoded high frequency subband energy estimation coefficient according to which an approximate high frequency component of an estimation result of a component of high frequency in the last frame is obtained is selected and even on the decoding device side 40, audio with high quality sound more natural can be reproduced. Conversely, with a non-constant region of the input signal, the term of the evaluated value ResP(id, J) in the evaluated value Resall(id, J) becomes 0 and a decoded high frequency signal closer to the high frequency signal real is obtained.
[00387] The pseudo-high-frequency sub-band energy difference calculation circuit 36 performs the aforementioned processing to calculate the evaluated value Resall(id, J) for each K decoded high-frequency sub-band energy estimation coefficients .
[00388] In step S339, the pseudo-high frequency subband energy difference calculation circuit 36 selects the coefficient index id based on the evaluated value Resall(id, J) for each subband energy estimate coefficient. decoded high frequency band obtained.
[00389] The evaluated value Resall(id, J) obtained in the above processing is a value evaluated by performing linear coupling on the evaluated value Res(id, J) and on the evaluated value ResP(id, J) using weight. As stated, the smaller the value of the evaluated value Res(id, J), a decoded high frequency signal closer to the real high frequency signal is obtained. Also, the smaller the value of the evaluated value ResP(id, J), a decoded high-frequency signal closer to the decoded high-frequency signal of the last frame is obtained.
[00390] In this way, the smaller the evaluated value Resall(id, J), the more suitable decoded high frequency signal is obtained. Therefore, the pseudo-high frequency subband energy difference calculation circuit 36 selects, from the K evaluated values Resall(id, J), an evaluated value according to which the value becomes minimum and supplies an index of coefficient indicating a decoded high frequency subband energy estimation coefficient corresponding to the evaluated value thereof to the high frequency coding circuit 37.
[00391] After the coefficient index is selected, the processes of step S340 and step S341 are performed and the encoding processing ends, but these processes are the same as those of step S308 and step S309 of figure 24 and in this way , description of these will be omitted.
[00392] As described above, in the coding device 30, the evaluated value Resall(id, J) obtained by performing linear coupling on the evaluated value Res(id, J) and on the evaluated value ResP(id, J) is used, and the coefficient index of the ideal decoded high frequency subband energy estimation coefficient is selected.
[00393] In the event of employment of the assessed value Resall(id, J), in the same way as in the case of employment of the assessed value Res(id, J), a high frequency subband energy estimation coefficient most suitable decoded can be selected by many more rating scales. Furthermore, if the evaluated value Resall(id, J) is employed, on the side of the decoding device 40, temporal fluctuation in a constant region of a high frequency component of a signal to be reproduced can be suppressed, and signals with higher sound quality can be obtained. <Modification 2>
[00394] Incidentally, with frequency band expansion processing, while trying to obtain audio with higher sound quality, subbands on the lower frequency side become important in relation to audibility. Specifically, of the subbands on the high frequency side, the higher the estimation accuracy of a subband the closer to the lower frequency side, the higher the sound quality the audio can be reproduced.
[00395] Therefore, in the event that a value evaluated in relation to each of the decoded high frequency subband energy estimation coefficients is calculated, weight can be placed on a subband on one side of the lower frequency. In such a case, the encoding device 30 of Fig. 18 performs the encoding processing illustrated in the flowchart of Fig. 26.
[00396] Next, the encoding processing by the encoding device 30 will be described in relation to the flowchart of figure 26. Note that the processing from step S371 to step S375 is the same as the processing from step S331 to step S335 of figure 25 and, in this way, description of this will be omitted.
[00397] In step S376, the pseudo-high frequency subband energy difference calculation circuit 36 calculates the evaluated value ResWband(id, J), with the current frame J serving as an object to be processed being employed, for each K decoded high frequency subband energy estimation coefficients.
[00398] Specifically, the pseudo-high-frequency sub-band energy difference calculation circuit 36 performs the same calculation as the aforementioned Expression (1) using the high-frequency sub-band signal of each sub-band supplied from the subband division circuit 33 to calculate the high frequency subband energy power(ib, J) in frame J.
[00399] In the event that the high frequency subband energy power(ib, J) is obtained, the pseudo high frequency subband energy difference calculation circuit 36 calculates the following Expression (27) to calculate an average value residual quadratic ResstdWband(id, J). [Mathematical Expression 27]

[00400] Specifically, with respect to each subband on the high frequency side whose index is sb + 1 to eb, the difference between the energy of the high frequency subband power(ib, J) and the subband energy of powerest(ib, id, J) pseudohigh frequency in frame J is obtained, and this difference is multiplied by the weight Wband(ib) for each subband. The sum of squares of the difference multiplied by the weight Wband(ib) is taken as the residual root mean square ResstdWband(id, J).
[00401] Here, the weight Wband(ib) (however, sb + 1 < ib < and b) is defined by the following Expression (28), for example. The value of this Wband(ib) weight increases in the event that a subband of it is on one side of the lower frequency. [Mathematical Expression 28]

[00402] Next, the pseudo-high frequency subband energy difference calculation circuit 36 calculates the maximum residual value ResmaxWband(id, J). Specifically, the maximum value of the absolute value of the values obtained by multiplying the difference between the high frequency subband energy power(ib, J) whose index is sb + 1 to eb and the pseudohigh frequency subband energy powerest(ib , id, J) of each subband by the weight Wband(ib) is taken as the maximum residual value ResmaxWband(id, J).
[00403] Also, the pseudo-high frequency subband energy difference calculation circuit 36 calculates the residual average value ResaveWband(id, J).
[00404] Specifically, for each subband whose index is sb + 1 to eb, the difference between the power(ib,J) high frequency subband energy and the powerest pseudohigh frequency subband energy (ib, id, J) is obtained and is multiplied by the weight Wband(ib), and the sum of the difference multiplied by the weight Wband(ib) is obtained. Then, the absolute value of a value obtained by dividing the sum of the difference obtained by the number of subbands (eb - sb) on the high frequency side is taken as the residual average value ResaveWband(id, J).
[00405] Additionally, the pseudo-high frequency subband energy difference calculation circuit 36 calculates the evaluated value ResWband(id, J). Specifically, the sum of the residual root mean square ResstdWband(id, J), the maximum residual value ResmaxWband(id, J) multiplied by the weight Wmax and the residual mean value ResaveWband(id, J) multiplied by the weight Wave is taken as the evaluated value ResWband(id, J).
[00406] In step S377, the pseudo-high frequency subband energy difference calculation circuit 36 calculates the evaluated value ResPWband(id, J) with the last frame and the current frame being employed.
[00407] Specifically, the pseudo-high-frequency sub-band energy difference calculation circuit 36 records, with respect to the temporally previous frame (J - 1) after the J-frame to be processed, a pseudo-high-frequency sub-band energy frequency of each subband obtained by using a high frequency subband energy estimation coefficient decoded with the finally selected coefficient index.
[00408] The pseudo-high frequency subband energy difference calculation circuit 36 first calculates an estimated residual root mean square ResPstdWband(id, J). Specifically, with respect to each subband on the high frequency side whose index is sb + 1through eb, the difference between the powerest(ib, idselected(J - 1), J - 1) pseudohigh frequency subband energy and the energy of pseudohigh frequency subband powerest(ib, id, J) is obtained and is multiplied by the weight Wband(ib). Then, the sum of squares of the difference multiplied by the weight Wband(ib) is taken as the estimated residual root mean square ResPstdWband(id, J).
[00409] Next, the pseudo-high frequency subband energy difference calculation circuit 36 calculates an estimated maximum residual value ResPmaxWband(id, J). Specifically, the maximum value of the absolute value of the values obtained by multiplying the difference between the powerest(ib, idselected(J - 1), J - 1) pseudohigh frequency subband energy and the pseudohigh frequency subband energy powersest(ib, id, J) of each subband whose index is sb + 1up to and b by the weight Wband(ib) is taken as the estimated maximum residual value ResPmaxWband(id, J).
[00410] Next, the pseudo-high frequency subband energy difference calculation circuit 36 calculates an estimated residual average value ResPaveWband(id, J). Specifically, for each subband whose index is sb + 1through eb, the difference between the powerest(ib, idselected(J - 1), J - 1) pseudohigh frequency subband energy and the sub-band energy powerest(ib, id, J) pseudohigh frequency band is obtained and is multiplied by the weight Wband(ib). The absolute value of a value obtained by dividing the sum of the difference multiplied by the weight Wband(ib) by the number of subbands on the high frequency side is then taken as the estimated residual average value ResPaveWband(id, J).
[00411] Additionally, the pseudo-high frequency subband energy difference calculation circuit 36 obtains the sum of the estimated residual root mean square ResPstdWband(id, J), the estimated maximum residual value ResPmaxWband(id, J) multiplied by Wmax weight and the estimated residual mean value ResPaveWband(id, J) multiplied by the Wave weight, and takes this as an evaluated ResPWband(id, J) value.
[00412] In step S378, the pseudo-high frequency subband energy difference calculation circuit 36 adds the evaluated value ResWband(id, J) and the evaluated value ResPWband(id, J) multiplied by the weight Wp(J) in Expression (25) to calculate the final evaluated value ResallWband(id, J). This rated value ResallWband(id, J) is calculated for each K decoded high frequency subband energy estimation coefficients.
[00413] Subsequently, the processes from step S379 to step S381 are performed, and the encoding processing ends, but these processes are the same as the processes from step S339 to step S341 of figure 25 and, in this way, their description will be omitted . Note that, in step S379, of the K coefficient indices, a coefficient index according to which the evaluated value ResallWband(id, J) becomes the minimum is selected.
[00414] In this way, weighting is performed for each subband to put weight on a subband on one side of the lower frequency, thereby enabling audio with higher sound quality to be obtained on the side of the decoding device 40 .
[00415] Note that, although a description was previously made that decoded high frequency subband energy estimation coefficients are selected based on the evaluated value ResallWband(id, J), subband energy estimation coefficients Decoded high frequency can be selected based on the rated ResWband(id, J) value. <Modification 3>
[00416] Additionally, the human auditory perception has a characteristic to the effect that the more a frequency band has amplitude (energy), the more the human auditory perception perceives this and, in this way, an evaluated value in relation to each coefficient of decoded high frequency subband energy estimate can be calculated to put weight on a subband with more energy.
[00417] In such a case, the decoding device 30 of Figure 18 performs the encoding processing illustrated in the flowchart of Figure 27. Next, the encoding processing by the encoding device 30 will be described in relation to the flowchart of Figure 27 Note that the processes from step S401 to step S405 are the same as the processes from step S331 to step S335 of figure 25 and, thus, their description will be omitted.
[00418] In step S406, the pseudo-high frequency subband energy difference calculation circuit 36 calculates an evaluated value ResWpower(id, J), with the current frame J serving as an object to be processed being employed, for each K decoded high frequency subband energy estimation coefficients.
[00419] Specifically, the pseudo-high-frequency sub-band energy difference calculation circuit 36 performs the same calculation as the above-mentioned Expression (1) to calculate a high-frequency sub-band energy power(ib, J) in the frame J using the high frequency subband signal of each subband supplied from the subband division circuit 33.
[00420] In the event that the high frequency subband energy power(ib, J) is obtained, the pseudo high frequency subband energy difference calculation circuit 36 calculates the following Expression (29) to calculate an average value residual quadratic ResstdWpower(id, J). [Mathematical Expression 29]

[00421] Specifically, with respect to each subband on the high frequency side whose index is sb + 1up to eb, the difference between the high frequency subband energy power(ib, J) and the subband energy of pseudohigh frequency powerest(ib, id, J) is obtained, and this difference is multiplied by the weight Wpower(power(ib, J)) for each subband. The sum of squares of the difference multiplied by the weight Wpower(power(ib, J)) is then taken as a residual root mean square ResstdWpower(id, J).
[00422] Here, the weight Wpower(power(ib, J))(however, sb + 1 < ib < eb) is defined by the following Expression (30), for example. The value of this weight Wpower(power(ib, J)) increases in the event that the high frequency subband energy power(ib, J) of a subband is greater. [Mathematical Expression 30]

[00423] Next, the pseudo-high frequency subband power difference calculation circuit 36 calculates a maximum residual value ResmaxWpower(id, J). Specifically, the maximum value of the absolute value of the values obtained by multiplying the difference between the power(ib, J) high frequency subband energy and the powerest(ib, id, J) pseudohigh frequency subband energy of each subband whose index is sb + 1up to and b by the weight Wpower(power(ib, J)) is taken as the maximum residual value
[00424] Also, the pseudo-high frequency subband energy difference calculation circuit 36 calculates a residual average value.
[00425] Specifically, for each subband whose index is sb + 1up to eb, the difference between the high frequency subband energy power(ib, J) and the pseudohigh frequency subband energy power( ib, id, J) is obtained and is multiplied by the weight Wpower(power(ib, J)), and the sum of the difference multiplied by the weight Wpower(power(ib, J)) is obtained. The absolute value of a value obtained by dividing the sum of the difference obtained by the number of subbands (eb - sb) on the high frequency side is then taken as the residual average ResaveWpower(id, J) value.
[00426] Additionally, the pseudo-high frequency subband power difference calculation circuit 36 calculates an evaluated value ResWpower(id, J). Specifically, the sum of the residual root mean square ResstdWpower(id, J), the maximum residual value ResmaxWpower(id, J) multiplied by the weight Wmax and the residual mean value ResaveWpower(id, J) multiplied by the weight Wave is taken as the evaluated value ResWpower(id, J).
[00427] In step S407, the pseudo-high frequency subband power difference calculation circuit 36 calculates an evaluated value ResPWpower(id, J) with the last frame and the current frame being employed.
[00428] Specifically, the pseudo-high-frequency sub-band energy difference calculation circuit 36 records, with respect to the temporally previous frame (J - 1) after the J-frame to be processed, a pseudo-high-frequency sub-band energy frequency of each subband obtained by using a high frequency subband energy estimation coefficient decoded with the finally selected coefficient index.
[00429] The pseudo-high frequency subband power difference calculation circuit 36 first calculates an estimated residual root mean square ResPstdWpower(id, J). Specifically, for each subband on the high frequency side whose index is sb + 1through eb, the difference between the pseudohigh frequency subband energy powersest(ib, idselected(J - 1), J - 1) and the pseudohigh frequency subband energy powersest(ib, id, J) is obtained and is multiplied by the weight Wpower(power(ib, J)). Then, the sum of squares of the difference multiplied by the weight Wpower(power(ib, J)) is taken as the estimated residual root mean square ResPstdWpower(id, J).
[00430] Next, the pseudo-high frequency subband power difference calculation circuit 36 calculates an estimated maximum residual value ResPmaxWpower(id, J). Specifically, the maximum value of the absolute value of the values obtained by multiplying the difference between the powerest(ib, idselected(J - 1), J - 1) pseudohigh frequency subband energy and the pseudohigh frequency subband energy powerest(ib, id, J) of each subband whose index is sb + 1up to and b by the weight Wpower(power(ib, J)) is taken as the estimated maximum residual value ResPmaxWpower(id, J).
[00431] Next, the pseudo-high frequency subband power difference calculation circuit 36 calculates an estimated residual average value ResPaveWpower(id, J). Specifically, for each subband whose index is sb + 1through eb, the difference between the powerest(ib, idselected(J - 1), J - 1) pseudohigh frequency subband energy and the sub-band energy powerest(ib, id, J) pseudohigh frequency band is obtained and is multiplied by the weight Wpower(power(ib, J)). The absolute value of a value obtained by dividing the sum of the difference multiplied by the weight Wpower(power(ib, J)) by the number of subbands (eb - sb) on the high frequency side is then taken as the average value. estimated residual ResPaveWpower(id, J).
[00432] Additionally, the pseudo-high frequency subband power difference calculation circuit 36 obtains the sum of the estimated residual root mean square ResPstdWpower(id, J), the estimated maximum residual value ResPmaxWpower(id, J) multiplied by Wmax weight and the estimated residual average value ResPaveWpower(id, J) multiplied by the Wave weight, and takes this as an evaluated ResPWpower(id, J) value.
[00433] In step S408, the pseudo-high frequency subband energy difference calculation circuit 36 adds the rated value ResWpower(id, J) and the rated value ResPWpower(id, J) multiplied by the weight Wp(J) in Expression (25) to calculate the final evaluated value ResallWpower(id, J). This rated value ResallWpower(id, J) is calculated for each of the K decoded high frequency subband energy estimation coefficients.
[00434] Subsequently, the processes from step S409 to step S411 are performed, and the encoding processing ends, but these processes are the same as the processes from step S339 to step S341 of figure 25 and, in this way, their description will be omitted . Note that, in step S409, of the K coefficient indices, a coefficient index according to which the evaluated value ResallWpower(id, J) becomes the minimum is selected.
[00435] In this way, weighting is performed for each subband to put weight on a subband with higher power, thereby enabling audio with higher sound quality to be obtained on the side of the decoding device 40.
[00436] Note that description has been made so far in which the selection of a decoded high frequency subband energy estimation coefficient is performed based on the evaluated value ResallWpower(id, J), but an energy estimation coefficient Decoded high frequency subband can be selected based on the rated value ResWpower(id, J). <6. Sixth Modality> [Coefficient Learning Device Configuration]
[00437] Incidentally, the set of coefficient Aib(kb) and coefficient Bib that serve as decoded high-frequency subband energy estimation coefficients was recorded in the decoding device 40 of Figure 20 in a manner correlated with a coefficient index . For example, in the event that the decoded high frequency subband energy estimate coefficients of the 128 coefficient indices are recorded in the decoding device 40, a larger region needs to be prepared as a recording region, such as memory. , to record these decoded high frequency subband energy estimation coefficients or the like.
[00438] Therefore, an arrangement can be made in which a part of the various decoded high frequency subband energy estimation coefficients is taken as common coefficients and, in this way, the recording region used to record the estimation coefficients of the decoded high-frequency subband power is reduced. In a case like this, a coefficient learning device that obtains estimation coefficients from the high frequency subband energy decoded by the learning is configured as illustrated in Figure 28, for example.
[00439] A coefficient learning device 81 is configured by a subband division circuit 91, a high frequency subband energy calculation circuit 92, a resource quantity calculation circuit 93 and a circuit of coefficient estimation 94.
[00440] Multiple music data to be used for learning and the like are supplied to this 81 coefficient learning device as wideband supervisory signals. Wideband supervisory signals are signals in which multiple high frequency subband components and multiple low frequency subband components are included.
[00441] The subband division circuit 91 is configured by a bandpass filter and the like, divides a supplied wideband supervisor signal into multiple subband signals and supplies them to the subband energy calculation circuit of high frequency 92 and to the resource quantity calculation circuit 93. Specifically, the high frequency subband signal of each subband on the high frequency side whose index is sb + 1 until and b is supplied to the calculation circuit of the high frequency subband energy 92 and the low frequency subband signal of each subband on the low frequency side whose index is sb - 3 to sbe supplied to the resource quantity calculation circuit 93.
[00442] The high frequency subband energy calculation circuit 92 calculates the high frequency subband energy of each high frequency subband signal supplied from the subband division circuit 91 to supply to coefficient estimating circuit 94. Resource amount calculation circuit 93 calculates a low frequency subband energy as a resource amount based on each low frequency subband signal supplied from the circuit. of subband division 91 to supply the coefficient estimation circuit 94.
[00443] The coefficient estimation circuit 94 generates a high frequency subband energy estimation coefficient decoded by performing regression analysis using the high frequency subband energy from the subband energy calculation circuit. - high frequency band 92 and the resource amount from the resource amount calculation circuit 93 for transmission to the decoding device 40. [Description of Coefficient Learning Device]
[00444] Next, the coefficient learning processing to be performed by the coefficient learning device 81 will be described in relation to the flowchart of figure 29.
[00445] In step S431, the subband division circuit 91 divides each of the multiple supplied wideband supervisor signals into multiple subband signals. Then, the subband division circuit 91 supplies the high frequency subband signal of a subband whose index is sb + 1up to eb to the high frequency subband energy calculation circuit 92 and supplies the low frequency subband signal of a subband whose index is sb - 3to sb to resource quantity calculation circuit 93.
[00446] In step S432, the high frequency subband energy calculation circuit 92 performs the same calculation as the aforementioned Expression (1) on each high frequency subband signal supplied from the sub division circuit -band 91 to calculate a high frequency subband energy for supplying the coefficient estimating circuit 94.
[00447] In step S433, the resource quantity calculation circuit 93 performs the calculation of the aforementioned Expression (1) on each low frequency subband signal supplied from the subband division circuit 91 to calculate a low frequency subband power as a resource amount to supply the coefficient estimation circuit 94.
Thus, the high frequency subband energy and the low frequency subband energy with respect to each frame of the multiple wideband supervisor signals are supplied to the coefficient estimation circuit 94.
[00449] In step S434, the coefficient estimation circuit 94 performs regression analysis using the least squares method to calculate a coefficient Aib(kb) and a coefficient Bib for each subband ib (however, sb + 1 < ib <eb) whose index is sb + 1 to eb.
[00450] Note that with the regression analysis, the low frequency subband energy supplied from the resource quantity calculation circuit 93 is taken as an explanatory variable and the high frequency subband energy supplied from the high frequency subband energy calculation circuit 92 is taken as an explained variable. Also, regression analysis is performed by the low frequency subband energies and the high frequency subband energies of all frames that constitute all wideband supervisory signals supplied to the coefficient 81 learning device being used. .
[00451] In step S435, the coefficient estimating circuit 94 obtains the residual vector of each frame of the wideband supervisory signals using the coefficient Aib(kb) and the coefficient Bib obtained for each subband ib.
[00452] For example, the coefficient estimation circuit 94 subtracts the sum of the sum of the total sum of the low frequency subband energy power(kb, J) (however, sb - 3 < kb < sb) multiplied by the coefficient Aib( kb) and by the coefficient Bib of the high frequency subband energy power(ib, J) for each subband ib (however, sb + 1 < ib < eb) of frame J to obtain the residual. A vector consisting of the residue of each subband ib of frame J is taken as a residual vector.
[00453] Note that the residual vector is calculated with respect to all frames that constitute all wideband supervisory signals supplied to the coefficient learning device 81.
[00454] In step S436, the coefficient estimation circuit 94 normalizes the residual vector obtained with respect to each of the frames. For example, the coefficient estimation circuit 94 obtains, for each subband ib, residual scatter values of the subbands ib of the residual vectors of all frames and divides the residual of the subband ib into each residual vector. by the square root of the dispersion values of this, thus normalizing the residual vectors.
[00455] In step S437, the coefficient estimation circuit 94 performs grouping on the normalized residual vectors of all frames by the k-means or the like method.
[00456] For example, say that an average frequency enveloping of all frames obtained at the time of performing the estimate of a high frequency subband energy using the coefficient Aib(kb) and the coefficient Bib will be referred to as a SA mid frequency enveloping. Also, say that predetermined frequency envelope whose energy is greater than that of the SA mid frequency envelope will be referred to as an SH frequency envelope and predetermined frequency envelope whose energy is less than that of the SA mid frequency envelope will be referred to as an SL frequency envelope.
[00457] At this time, the grouping of the residual vectors is performed so that the residual vectors of the coefficients according to which frequency envelopes that approach the SA medium frequency envelope, the SH frequency envelope and the SL frequency envelope were obtained belonging to a CA grouping, a CH grouping and a CL grouping, respectively. In other words, clustering is performed in such a way that the residual vector of each frame belongs to any one of the CA cluster, the CH cluster, or the CL cluster.
[00458] With frequency band expansion processing to estimate a high-frequency component based on a correlation between a low-frequency component and a high-frequency component, when calculating a residual vector using the coefficient Aib(kb ) and the Bib coefficient obtained by the regression analysis, residual error increases since a subband belongs to one side of the higher frequency in its characteristics. Therefore, when performing grouping on an unchanged residual vector, processing is performed such that weight is placed on a subband on one side of the higher frequency.
[00459] On the other hand, in the coefficient learning device 81, residual vectors are normalized with the residual dispersion value of each sub-band, according to which, grouping can be performed with uniform weight being placed in each sub-band. band, whereas the residual dispersion of each subband is equal in appearance.
[00460] In step S438, the coefficient estimation circuit 94 selects either cluster CA, cluster CH, or cluster CL as a cluster to be processed.
[00461] In step S439, the coefficient estimation circuit 94 calculates the coefficient Aib(kb) and the coefficient Bib of each subband ib (however, sb + 1 < ib < eb) by regression analysis using the tables of residual vectors that belong to the cluster selected as the cluster to be processed.
[00462] Specifically, if it is said that the frame of a residual vector that belongs to the cluster to be processed will be referred to as a frame to be processed, the energies of the low frequency subband and the energies of the high frequency subband of all frames to be processed are taken as explanatory variables and explained variables, and the regression analysis employing the method of least squares is performed. Thus, the coefficient Aib(kb) and the coefficient Bib are obtained for each subband ib.
[00463] In step S440, the coefficient estimation circuit 94 obtains, in relation to all frames to be processed, residual vectors using the coefficient Aib(kb) and the coefficient Bib obtained by the processing of step S439. Note that, in step S440, the same processing as in step S435 is performed and the residual vector of each frame to be processed is obtained.
[00464] In step S441, the coefficient estimation circuit 94 normalizes the residual vector of each frame to be processed obtained in the processing of step S440 by performing the same processing of step S436. That is, normalization of a residual vector is performed by the residual error being divided by the square root of a dispersion value for each subband.
[00465] In step S442, the coefficient estimation circuit 94 performs grouping in the normalized residual vectors of all frames to be processed by the k-means or the like method. The number of clusters mentioned here is determined as follows. For example, in the event of trying to generate decoded high frequency subband energy estimation coefficients from 128 coefficient indices in the coefficient learning device 81, a number obtained by multiplying the number of frames to be processed by 128 and additionally , by dividing this by the number of all frames is taken as the number of clusters. Here, the number of all frames is a total number of all frames of all wideband supervisory signals supplied to coefficient learning device 81.
[00466] In step S443, the coefficient estimation circuit 94 obtains the vector of the center of gravity of each cluster obtained by processing step S442.
[00467] For example, the cluster obtained by clustering step S442 corresponds to a coefficient index, a coefficient index is assigned to each cluster in the learning device of coefficient 81, and the coefficient of estimating the high subband energy decoded frequency of each coefficient index is obtained.
[00468] Specifically, it is said that, in step S438, the grouping CA was selected as the grouping to be processed and F groups were obtained by the grouping of step S442. Now, if you pay attention to a CF cluster, which is one of the F clusters, the high frequency subband energy estimate coefficient decoded from the CF cluster coefficient index is taken as the coefficient Aib(kb) obtained with respect to the CA cluster from step S439, which is a linear correlation term. Also, the sum of a vector obtained by subjecting the center of gravity vector of the CF cluster obtained in step S443 to the inverse normalization processing performed in step S441 (reverse normalization) and the Bib coefficient obtained in step S439 is taken as the Bib coefficient , which is a constant term of the decoded high frequency subband energy estimate coefficient. The reverse normalization mentioned here is processing to multiply each factor of the CF cluster center of gravity vector by the same normalization value (square root of the dispersion values for each subband) in the event that the normalization performed in step S441 is for divide residual error by the square root of the dispersion values for each subband, for example.
[00469] Specifically, the set of the coefficient Aib(kb) obtained in step S439 and the coefficient Bib obtained as described above becomes the high frequency subband energy estimation coefficient decoded from the coefficient index of the cluster CF. In this way, each of the F clusters obtained by the cluster commonly has the coefficient Aib(kb) obtained in relation to the cluster CA as a linear correlation term of the decoded high frequency subband energy estimate coefficient.
[00470] In step S444, the coefficient learning device 81 determines whether all the clusters of the CA cluster, the CH cluster and the CL cluster have been processed or not as the cluster to be processed. In the event that determination is made, in step S444, that not all groupings have been processed, processing returns to step S438, and the aforementioned processing is repeated. That is, the next cluster is selected as an object to be processed, and a decoded high frequency subband energy estimate coefficient is calculated.
[00471] On the other hand, in the event that determination is made, in step S444, that all clusters were processed, a desired predetermined number of decoded high frequency subband energy estimation coefficients was obtained and in this way , processing proceeds to step S445.
[00472] In step S445, the coefficient estimating circuit 94 transmits the obtained coefficient index and the decoded high frequency subband energy estimating coefficient to the decoding device 40 for recording them therein, and the learning processing of the coefficient ends.
[00473] For example, the decoded high-frequency subband energy estimate coefficients to be transmitted to the decoding device 40 include several decoded high-frequency subband energy estimate coefficients with the same coefficient Aib(kb) as a linear correlation term. Therefore, the coefficient learning device 81 correlates these common coefficients Aib(kb) with an index of the linear correlation term (pointer), which is information to identify the coefficients Aib(kb) and also correlates the coefficient indices with the index of the linear correlation term and the Bib coefficient, which is a constant term.
[00474] Then, the coefficient learning device 81 supplies the index of the linear correlation term (pointer) and the correlated Aib(kb) coefficient, and the coefficient index and index of the correlated linear correlation term (pointer) and the coefficient Bib to the decoding device 40 to store them in the memory of the high frequency decoding circuit 45 of the decoding device 40. In this way, at the time of recording the multiple decoded high frequency subband energy estimation coefficients, in With respect to common linear correlation terms, if linear correlation term indices (pointers) are stored in a recording region for the decoded high frequency subband energy estimation coefficients, the recording region can be significantly reduced.
[00475] In this case, the indices of the linear correlation term and the coefficients Aib(kb) are recorded in the memory of the high-frequency decoding circuit 45 in a correlated manner and, in this way, an index of the linear correlation term and the Bib coefficient can be obtained from an index of coefficient and additionally the coefficient Aib(kb) can be obtained from the index of the linear correlation term.
[00476] Note that, as a result of the analysis by the present applicant, even if the linear correlation terms of the multiple decoded high frequency subband energy estimation coefficients were normalized around three patterns, it was known that there are almost no sound quality related deterioration in the audibility of audio subjected to frequency band expansion processing. In this way, according to the coefficient learning device 81, the recording region used for recording the decoded high frequency subband energy estimation coefficients can be further reduced without deteriorating the audio sound quality after processing. frequency band expansion.
[00477] As described above, the coefficient learning device 81 generates and transmits the decoded high frequency subband energy estimate coefficient of each coefficient index from the supplied wideband supervisor signal.
[00478] Note that, with the coefficient learning processing of figure 29, a description was made that residual vectors are normalized, but in either step S436 or step S441, or in both, normalization of the residual vectors may not be fulfilled.
[00479] Alternatively, although normalization of the residual vectors can be performed, the sharing of the linear correlation terms of the decoded high frequency subband energy estimation coefficients may not be performed. In such a case, after the normalization processing of step S436, the normalized residual vectors are subjected to grouping in the same number of groups as the number of decoded high frequency subband energy estimation coefficients to be obtained. Regression analysis is performed for each cluster using the frame of a residual vector that belongs to each cluster, and the decoded high frequency subband energy estimation coefficient of each cluster is generated. <7. Seventh Mode> [Example of Functional Configuration of Encoding Device]
[00480] Incidentally, description has been made so far that, at the time of encoding an input signal, the coefficient Aib(kb) and the coefficient Bib according to which a high-frequency envelope can be estimated with the best precision are selected from a low frequency envelope of the input signal. In this case, coefficient index information indicating the coefficient Aib(kb) and the coefficient Bibé is included in the output code sequence and is transmitted alongside the decoding and, at the time of decoding the output code sequence, an envelope of high frequency is generated by using the coefficient Aib(kb) and the coefficient Bib corresponding to the coefficient index.
[00481] However, in the event that temporal fluctuation of a low-frequency envelope is large, even if estimation of a high-frequency envelope has been performed using the same coefficient Aib(kb) and coefficient Bib for consecutive frames of the input signal , the temporal fluctuation of the high-frequency envelope increases.
[00482] In other words, in the event that the temporal fluctuation of a low frequency subband energy is large, even if a decoded high frequency subband energy has been calculated using the same coefficients Aib(kb) and Bib coefficient, the temporal fluctuation of the decoded high frequency subband energy increases. This is because a low frequency subband energy is employed for calculating a decoded high frequency subband energy and, in this way, when the temporal fluctuation of this low frequency subband energy is large, a decoded high frequency subband energy to be obtained also fluctuates enormously temporarily.
[00483] Also, although description has been made so far in which the multiple sets of coefficient Aib(kb) and coefficient Bib are prepared in advance by learning with a wideband supervisory signal, this wideband supervisory signal is a signal obtained by encoding the input signal and additionally decoding the input signal after encoding.
[00484] The sets of coefficient Aib(kb) and coefficient Bib obtained by such learning are coefficient sets suitable for a case to encode the actual input signal using the coding system and the coding algorithm during the coding of the signal. entry at the time of learning.
[00485] At the time of generating a broadband supervisor signal, a different broadband supervisor is obtained depending on which type of encoding system is employed for encoding/decoding the input signal. Also, if the encoders (coding algorithms) differ, although the same encoding system is employed, a different wideband supervisory signal is obtained.
[00486] In this way, in the event that only a signal obtained by encoding/decoding the input signal using an encoding system and a particular encoding algorithm was employed as a broadband supervisor signal, it may be difficult to estimate an envelope frequency with high precision from the coefficient Aib(kb) and the coefficient Bib obtained. That is, there may not be enough capacity to handle the difference between encoding systems or encoding algorithms.
[00487] Therefore, an arrangement can be made in which smoothing a low-frequency envelope and generating suitable coefficients are performed, thereby enabling a high-frequency envelope to be estimated with high accuracy, regardless of the temporal fluctuation of an envelope frequency, coding system and the like.
[00488] In a case like this, an encoding device that encodes the input signal is configured as shown in figure 30. Note that, in figure 30, a part corresponding to the case of figure 18 is denoted with the same number of reference, and description thereof will be omitted as appropriate. The encoding device 30 of Fig. 30 differs from the encoding device 30 of Fig. 18 in that a parameter determining unit 121 and a smoothing unit 122 are innovatively provided, and other points are the same.
[00489] The parameter determination unit 121 generates a parameter regarding the smoothing of a low frequency subband energy to be calculated as a resource amount (hereinafter referred to as smoothing parameter) based on the signal of high frequency subband supplied from the subband division circuit 33. The parameter determining unit 121 supplies the generated smoothing parameter to the pseudohigh frequency subband energy difference calculation circuit 36 and to the unit of standardization 122.
[00490] Here, the smoothing parameter is information or the like that indicates the value of how many temporally consecutive low frequency subband energy frames is used to smooth the low frequency subband energy of the current frame that serves as an object to be processed, for example. That is, a parameter to be used for smoothing processing of a low frequency subband energy is determined by the parameter determination unit 121.
[00491] The smoothing unit 122 smoothes the low frequency subband power that serves as a resource quantity supplied from the resource quantity calculation circuit 34 using the smoothing parameter supplied from the parameter determination unit 121 for supplying the pseudo-high frequency subband energy calculation circuit 35.
[00492] In the pseudo-high frequency subband energy calculation circuit 35, the multiple decoded high frequency subband energy estimation coefficients obtained by regression analysis, a coefficient group index and a coefficient index to identify these decoded high frequency subband energy estimation coefficients are recorded in a correlated manner.
[00493] Specifically, encoding is performed on an input signal according to each of the multiple different encoding systems and encoding algorithms, and a signal obtained by further decoding a signal obtained by encoding is prepared as a band supervisor signal wide.
[00494] For each of these multiple wideband supervisory signals, a low frequency subband energy is taken as an explanatory variable and a high frequency subband energy is taken as an explained variable. According to the regression analysis (learning) using the least squares method, the multiple sets of the coefficient Aib(kb) and the coefficient Bib of each subband are obtained and recorded in the subband energy calculation circuit of pseudohigh frequency 35.
[00495] Here, with learning using a wideband supervisory signal, multiple sets of the coefficient Aib(kb) and the coefficient Bib of each subband (hereinafter referred to as coefficient sets) are obtained. Say that a group of multiple coefficient sets obtained from a broadband supervisor signal in this way will be referred to as a coefficient group, information to identify a coefficient group will be referred to as an index of the coefficient group and information to identify a coefficient set that belongs to a coefficient group will be referred to as a coefficient index.
[00496] With the pseudo-high frequency subband energy calculation circuit 35, a coefficient set of multiple coefficient groups is recorded in a manner correlated with a coefficient group index and a coefficient index to identify the set. of coefficient of this. That is, a coefficient set (coefficient Aib(kb) and coefficient Bib) that serves as a decoded high frequency subband energy estimation coefficient, recorded in the pseudohigh frequency subband energy calculation circuit 35 is identified by a coefficient group index and a coefficient index.
[00497] Note that, at the time of learning a coefficient set, a low-frequency subband energy that serves as an explanatory variable can be smoothed by the same smoothing processing of a low-frequency subband energy that serves as a resource amount in the smoothing unit 122.
[00498] The pseudo-high-frequency sub-band energy calculation circuit 35 calculates the pseudo-high-frequency sub-band energy of each sub-band on the high-frequency side using, for each sub-band energy estimation coefficient. The decoded high frequency subband energy estimation coefficient is recorded, the decoded high frequency subband energy estimation coefficient and the amount of resource after smoothing supplied from the smoothing unit 122 for supplying the subband energy difference calculation circuit pseudo-high frequency 36.
[00499] The pseudo-high-frequency sub-band energy difference calculation circuit 36 compares a high-frequency sub-band energy obtained from the high-frequency sub-band signal supplied from the sub-division circuit. -band 33 and the pseudo-high-frequency sub-band energy from the pseudo-high-frequency sub-band energy calculation circuit 35.
[00500] Then, the pseudo-high-frequency subband energy difference calculation circuit 36 supplies, as a result of the comparison, the multiple decoded high-frequency subband energy estimation coefficients, the coefficient group index and the coefficient index of the decoded high frequency subband energy estimation coefficient according to which a pseudohigh frequency subband energy closest to a high frequency subband energy was obtained, to the encoding circuit at high frequency 37. Also, the pseudo-high frequency subband energy difference calculation circuit 36 also supplies smoothing information indicating the smoothing parameter supplied from the parameter determining unit 121 to the high-coding circuit frequency 37.
[00501] In this way, multiple coefficient groups are prepared in advance by the learning to handle the difference of coding systems or coding algorithms and are recorded in the pseudo-high frequency subband energy calculation circuit 35, according to that, a more suitable decoded high frequency subband energy estimation coefficient can be employed. Thus, on the decoding side of the output code sequence, estimation of a high-frequency envelope can be performed with higher precision, independent of encoding systems or encoding algorithms. [Encoding Device Encoding Processing]
[00502] Next, the encoding processing to be performed by the encoding device 30 of figure 30 will be described in relation to the flowchart of figure 31. Note that the processes from step S471 to step S474 are the same as the processes from step S181 to step S184 of Fig. 19 and thus description of these will be omitted.
[00503] Meanwhile, the high frequency subband signal obtained in step S473 is supplied from the subband division circuit 33 to the pseudo high frequency subband energy difference calculation circuit 36 and to the determination unit of parameter 121. Also, in step S474, as a resource quantity, the low frequency subband energy power(ib, J) of each ib subband (sb - 3 < ib < sb) on the low side frequency of frame J that serves as an object to be processed is calculated and supplied to smoothing unit 122.
[00504] In step S475, the parameter determination unit 121 determines the number of frames to be used for smoothing a resource amount, based on the high frequency subband signal of each subband on the high side frequency supplied from the subband division circuit 33.
[00505] For example, the parameter determination unit 121 performs the calculation of the aforementioned Expression (1) with respect to each sub-band (however, sb + 1 < ib < and b) on the high frequency side of the J frame it serves as an object to be processed to obtain a subband energy and additionally obtain the sum of these subband energies.
[00506] Similarly, the parameter determination unit 121 obtains, in relation to a temporally previous frame (J - 1) before the J frame, the subband energy of each subband ib on the high frequency side and obtains additionally the sum of these subband energies. The parameter determination unit 121 compares a value obtained by subtracting the sum of subband energies obtained with respect to frame (J - 1) from the sum of subband energies obtained with respect to frame J (referred to below as difference of the sum of the subband energy) and a predetermined threshold.
[00507] For example, the parameter determination unit 121 determines, in the event that the difference of the sum of the subband energy is equal to or greater than the threshold, the number of frames to be used for smoothing a quantity of resource (hereafter referred to as the number of frames ns) as ns = 4 and, in the event that the subband energy sum difference is less than the threshold, determines the number of frames ns as ns = 16. The parameter determination unit 121 supplies the number of frames determined ns to the pseudo-high frequency subband energy difference calculation circuit 36 and to the smoothing unit 122 as the smoothing parameter.
[00508] Now, an arrangement can be made in which the difference of the sum of subband energy and multiple thresholds are compared, and the number of frames n is determined to be any one of three or more values.
[00509] In step S476, the smoothing unit 122 calculates the following Expression (31) using the smoothing parameter supplied from the parameter determination unit 121 to smooth the resource quantity supplied from the quantity calculation circuit. resource 34, and supplies it to the pseudohigh frequency subband energy calculation circuit 35. That is, the low frequency subband energy power(ib, J) of each subband on the low frequency side of the frame J to be processed supplied as the resource quantity is smoothed. [Mathematical Expression 31]

[00510] Note that, in Expression (31), ns is the number of frames ns that serves as a smoothing parameter, and the greater this number of frames ns, the more frames are used for smoothing the low subband energy frequency that serves as a resource amount. Also, say that the low frequency subband energies of the subbands of the value of several frames before frame J are held in smoothing unit 122.
[00511] Also, the weight SC(l) by which the low frequency subband energy power(ib, J) is multiplied is the weight to be determined by the following Expression (32), for example. The SC(l) weight for each frame has a large value, as does the SC(l) weight by which a temporally approximated frame of the J frame to be processed is multiplied. [Mathematical Expression 32]

[00512] In this way, in the smoothing unit 122, the quantity is smoothed by performing addition weighted by the weighting of SC(1) in the value of the last n frames of the low frequency subband energies to be determined by the number of frames ns that include the current J frame. Specifically, a weighted average of the low-frequency subband energies of the same subbands as the J frame. (ib, J) after standardization.
[00513] Here, the greater the number of frames ns to be used for smoothing, the smaller is the temporal fluctuation of the low frequency subband energy powersmooth(ib, J). Thus, in the event of estimating a subband energy on the high frequency side using the low frequency subband energy powersmooth(ib, J), temporal fluctuation of an estimated value of a subband energy on the high frequency side can be reduced.
[00514] However, unless the number of frames ns is set to a smaller value, as much as possible for a transient input signal, such as attack or the like, that is, an input signal where the temporal fluctuation of the high frequency component is large, tracking the temporal change of the input signal is delayed. Consequently, on the decoding side, during the reproduction of an output signal obtained by the decoding, it is likely that unnatural sensations in audibility are caused.
[00515] Therefore, in the parameter determination unit 121, in the event that the aforementioned difference of the sum of the subband energy is equal to or greater than the threshold, the input signal is considered as a transient signal in which the energy subband bandwidth on the high frequency side fluctuates enormously temporarily and the number of frames n is determined to be a smaller value (eg ns = 4). Thus, even when the input signal is a transient signal (signal with attack), the low frequency subband energy is properly smoothed, the temporal fluctuation of the estimated value of the subband energy on the high frequency side is reduced and also, tracking delay of change in high frequency components can be suppressed.
[00516] On the other hand, in the event that the difference of the sum of the subband energy is less than the threshold, in the parameter determination unit 121, the input signal is considered as a constant signal with less temporal fluctuation of the subband energy on the high frequency side and the number of frames n is determined as a larger value (eg ns = 16). Thus, the low frequency subband energy is properly smoothed and the temporal fluctuation of the estimated value of the subband energy on the high frequency side can be reduced.
[00517] In step S477, the pseudo-high-frequency sub-band energy calculation circuit 35 calculates a pseudo-high-frequency sub-band energy based on the low-frequency sub-band energy powersmooth(ib, J) of each subband on the low frequency side supplied from the smoothing unit 122, and supplies it to the pseudohigh frequency subband energy difference calculation circuit 36.
[00518] For example, the pseudo-high frequency subband energy calculation circuit 35 performs the calculation of the aforementioned Expression (2) using the coefficient Aib(kb) and the coefficient Bib recorded in advance as sub energy estimation coefficients. -decoded high frequency band and the low frequency subband energy powersmooth(ib, J) (however, sb - 3 < ib <sb) to calculate the powerest(ib, J) pseudohigh frequency subband energy .
[00519] Note that here the low frequency subband energy power(kb, J) of Expression (2) is replaced by the smoothed low frequency subband energy powersmooth(kb, J) (however, sb - 3 < kb < sb).
[00520] Specifically, the low frequency subband energy powersmooth(kb, J) of each subband on the low frequency side is multiplied by the coefficient Aib(kb) for each subband and additionally the coefficient Bib is added to the sum of the energies of the low frequency subband multiplied by the coefficient and is taken as the powerest(ib, J) pseudohigh frequency subband energy. This pseudo-high-frequency subband energy is calculated with respect to each subband on the high-frequency side whose index is sb + 1 through eb.
[00521] Also, the pseudo-high-frequency sub-band energy calculation circuit 35 performs calculation of a pseudo-high-frequency sub-band energy for each pre-recorded decoded high-frequency sub-band energy estimation coefficient. Specifically, with respect to all recorded coefficient groups, the calculation of a pseudo-high frequency subband energy is performed for each coefficient set (coefficient Aib(kb) and coefficient Bib) of the coefficient groups.
[00522] In step S478, the pseudo-high-frequency sub-band energy difference calculation circuit 36 calculates the pseudo-high-frequency sub-band energy difference based on the high-frequency sub-band signal from the high frequency subband signal. sub-band 33 and in the pseudo-high-frequency sub-band energy from the pseudo-high-frequency sub-band energy calculation circuit 35.
[00523] In step S479, the pseudo-high-frequency sub-band energy difference calculation circuit 36 calculates the aforementioned Expression (15) for each decoded high-frequency sub-band energy estimate coefficient to calculate the sum of the squares of the pseudohigh frequency subband energy difference (sum of the difference of squares E(J, id)).
[00524] Note that the processes of step S478 and step S479 are the same as the processes of step S186 and step S187 of figure 19 and, in this way, detailed description of these will be omitted.
[00525] When calculating the sum of squares difference E(J, id) for each pre-recorded high frequency subband energy estimation coefficient, the pseudohigh subband energy difference calculation circuit frequency 36 selects, from the sum of the difference of squares of this, the sum of the difference of squares according to which the value becomes the minimum.
[00526] Then, the pseudohigh frequency subband energy difference calculation circuit 36 supplies a coefficient group index and a coefficient index to identify a decoded high frequency subband energy estimate coefficient corresponding to the sum of the selected difference of squares and the smoothing information that indicates the smoothing parameter to the high frequency coding circuit 37.
[00527] Here, the smoothing information may itself be the value of the number of frames ns which serves as the smoothing parameter determined by the parameter determination unit 121 or it may be an indicator or the like that indicate the number of frames ns. For example, in the event that the smoothing information is taken as a 2-bit indicator that indicates the number of frames ns, the indicator value is set to 0 when the number of frames ns = 1, the indicator value is set to 1 when the number of frames ns = 4, the indicator value is set to 2 when the number of frames ns = 8 and the indicator value is set to 3 when the number of frames ns = 16.
[00528] In step S480, the high-frequency coding circuit 37 encodes the coefficient group index, the coefficient index and the smoothing information supplied from the pseudo-high frequency subband energy difference calculation circuit 36 and supplies the high frequency encoded data obtained therefrom to the multiplexing circuit 38.
[00529] For example, in step S480, the encoding of entropy or the like are performed on the coefficient group index, the coefficient index and the smoothing information. Note that the high-frequency encoded data can be any type of information, as long as the data comprises information from which the ideal decoded high-frequency subband energy estimation coefficient or the ideal smoothing parameter is obtained, by example, an index of the coefficient group or the like can be taken as high-frequency encoded data without change.
[00530] In step S481, the multiplexing circuit 38 multiplexes the low-frequency encoded data supplied from the low-frequency encoding circuit 32 and the high-frequency encoded data supplied from the high-frequency encoding circuit 37, transmits an exit code sequence obtained as a result of this, and the encoding processing ends.
[00531] In this way, the high-frequency encoded data obtained by encoding the coefficient group index, the coefficient index and the smoothing information are transmitted as an output code sequence, according to which, the decoding 40 that receives input from this output code sequence can estimate a high frequency component with higher accuracy.
[00532] Specifically, based on a coefficient group index and a coefficient index, of the multiple decoded high frequency subband energy estimation coefficients, the most appropriate coefficient for frequency band expansion processing can be obtained and a high frequency component can be estimated with high precision, independent of coding systems or coding algorithms. Furthermore, if a low-frequency subband energy that serves as a resource amount is smoothed according to the smoothing information, the temporal fluctuation of a high-frequency component obtained by estimation can be reduced and senseless audio unnatural audibility can be obtained regardless of whether the input signal is constant or transient. [Example of Functional Configuration of the Decoding Device]
[00533] Also, the decoding device 40 which inputs the output code sequence transmitted from the encoding device 30 of Fig. 30 as an input code sequence is configured as illustrated in Fig. 32, for example. Note that, in figure 32, a part corresponding to the case of figure 20 is denoted with the same reference number, and description of this will be omitted.
[00534] The decoding device 40 of Fig. 32 differs from the decoding device 40 of Fig. 20 in that a smoothing unit 151 is innovatively provided, and other points are the same.
[00535] In the decoding device 40 of Fig. 32, the high frequency decoding circuit 45 pre-records the same decoded high frequency subband energy estimate coefficient as a high subband energy estimate coefficient. decoded frequency that the pseudo-high frequency subband energy calculation circuit 35 of Fig. 30 records. Specifically, a set of coefficient Aib(kb) and coefficient Bib that serve as decoded high frequency subband energy estimation coefficients, obtained in advance by regression analysis, is recorded in a manner correlated with an index of the group of coefficient and a coefficient index.
[00536] The high-frequency decoding circuit 45 decodes the high-frequency encoded data supplied from the demultiplexing circuit 41 and, as a result, obtains a coefficient group index, a coefficient index and smoothing information. The high frequency decoding circuit 45 supplies a decoded high frequency subband energy estimation coefficient identified from the coefficient group index and coefficient index obtained to the high subband energy calculation circuit decoded frequency 46 and also supplies the smoothing information to the smoothing unit 151.
[00537] Also, the resource amount calculation circuit 44 supplies the low frequency subband energy calculated as a resource amount to the smoothing unit 151. The smoothing unit 151 smoothes the low subband energy frequency supplied from the resource quantity calculation circuit 44 according to the smoothing information from the high frequency decoding circuit 45 and supplies it to the decoded high frequency subband energy calculation circuit 46. [Decoding Device Decoding Process]
[00538] Next, the decoding processing to be performed by the decoding device 40 of Fig. 32 will be described in relation to the flowchart of Fig. 33.
[00539] This decoding processing is started when the output code sequence transmitted from the encoding device 30 is supplied to the decoding device 40 as an input code sequence. Note that the processes from step S511 to step S513 are the same as the processes from step S211 to step S213 of figure 21 and, thus, their description will be omitted.
[00540] In step S514, the high-frequency decoding circuit 45 performs decoding of the high-frequency encoded data supplied from the demultiplexing circuit 41.
[00541] The high-frequency decoding circuit 45 supplies, from the already recorded multiple decoded high-frequency subband energy estimation coefficients, a decoded high-frequency subband energy estimation coefficient indicated by the group index of coefficient and by the coefficient index obtained by decoding the high-frequency encoded data to the decoded high-frequency subband energy calculation circuit 46. Also, the high-frequency decoding circuit 45 supplies the smoothing information obtained by decoding the data encoded at high frequency to the smoothing unit 151.
[00542] In step S515, the resource amount calculation circuit 44 calculates a resource amount using the decoded low frequency subband signal from the subband division circuit 43 and supplies it to the smoothing unit 151 Specifically, according to the calculation of the aforementioned Expression (1), the low frequency subband energy power(ib, J) is calculated as an amount of resource with respect to each subband ib on the low frequency side .
[00543] In step S516, the smoothing unit 151 smoothes the low frequency subband energy power(ib, J) supplied from the resource amount calculation circuit 44 as a resource amount based on the information of smoothing supplied from the high frequency decoding circuit 45.
[00544] Specifically, the smoothing unit 151 performs the calculation of the aforementioned Expression (31) based on the number of frames ns indicated by the smoothing information to calculate a low-frequency subband energy powersmooth(ib, J) against to each ib subband on the low frequency side and supplies it to the decoded high frequency subband energy calculation circuit 46. Now, say the energies of the low frequency subband of the subbands of the value of several frames before frame J are kept in smoothing unit 151.
[00545] In step S517, the decoded high frequency subband energy calculation circuit 46 calculates a decoded high frequency subband energy based on the low frequency subband energy from the smoothing unit 151 and in the decoded high frequency subband energy estimation coefficient from the high frequency decoding circuit 45 and supplies it to the decoded high frequency signal generating circuit 47.
[00546] Specifically, the decoded high frequency subband energy calculation circuit 46 performs the calculation of the aforementioned Expression (2) using the coefficient Aib(kb) and the coefficient Bib which serve as the estimation coefficients of the sub energy. -decoded high frequency band and the powersmooth(ib, J) low frequency subband energy to calculate a decoded high frequency subband energy.
[00547] Note that here the low frequency subband energy power(kb, J) of Expression (2) is replaced by the smoothed low frequency subband energy powersmooth(kb, J) (however, sb - 3 < kb < sb). According to this calculation, the decoded high frequency subband energy powerest(ib, J) is obtained with respect to each subband on the high frequency side whose index is sb + 1 to eb.
[00548] In step S518, the decoded high-frequency signal generating circuit 47 generates a decoded high-frequency signal based on the decoded low-frequency subband signal supplied from the subband division circuit 43 and in the decoded high frequency subband energy supplied from the decoded high frequency subband energy calculation circuit 46.
[00549] Specifically, the decoded high frequency signal generating circuit 47 performs the calculation of the aforementioned Expression (1) using the decoded low frequency subband signal to calculate a low frequency subband energy with respect to each subband on the low frequency side. Then, the decoded high frequency signal generating circuit 47 performs the calculation of the aforementioned Expression (3) using the obtained low frequency subband energy and the decoded high frequency subband energy to calculate the gain amount. G(ib, J) for each subband on the high frequency side.
[00550] Also, the decoded high frequency signal generating circuit 47 performs the calculations of the aforementioned Expression (5) and Expression (6) using the gain amount G(ib, J) and the low subband signal frequency decoded to generate a high frequency subband signal x3(ib,n) with respect to each subband on the high frequency side.
[00551] Additionally, the decoded high-frequency signal generating circuit 47 performs the calculation of the aforementioned Expression (7) to obtain the sum of the obtained high-frequency subband signals and to generate a decoded high-frequency signal. The decoded high-frequency signal generating circuit 47 supplies the obtained decoded high-frequency signal to the synthesizing circuit 48, and processing proceeds from step S518 to step S519.
[00552] In step S519, the synthesizing circuit 48 synthesizes the decoded low-frequency signal from the low-frequency decoding circuit 42 and the decoded high-frequency signal from the decoded high-frequency signal generating circuit 47, and the transmits as an output signal. Afterwards, decoding processing ends.
[00553] As described above, according to the decoding device 40, a decoded high frequency subband energy is calculated using a decoded high frequency subband energy estimation coefficient identified by the coefficient group index and by the coefficient index obtained from the high-frequency encoded data, whereby the estimation accuracy of a high-frequency subband energy can increase. Specifically, multiple decoded high frequency subband energy estimation coefficients according to which the difference of encoding systems or encoding algorithms can be handled are recorded in advance in the decoding device 40. optimal decoded high frequency subband energy estimation coefficient identified by a coefficient group index and a coefficient index is selected and employed, according to which, the high frequency components can be estimated with high precision.
[00554] Also, in the decoding device 40, a low frequency subband energy is smoothed according to the smoothing information to calculate a decoded high frequency subband energy. In this way, the temporal fluctuation of a high-frequency envelope can be suppressed and audio without unnatural feeling in audibility can be obtained, regardless of whether the input signal is constant or transient.
[00555] Although description has been made so far in which the number of frames ns changes as a smoothing parameter, the weight SC(l) by which the energies of the low frequency subband power(ib, J) are multiplied in the smoothing time, with the number of frames ns as a fixed value, can be taken as a smoothing parameter. In such a case, the parameter determination unit 121 changes the weight SC(1) as a smoothing parameter, thereby changing the smoothing characteristics.
[00556] In this way, the SC(l) weight is also taken as a smoothing parameter, according to which the temporal fluctuation of a high frequency envelope can be adequately suppressed for a constant input signal and an input signal transient on the decoding side.
[00557] For example, in the event that the weight SC(l) of the aforementioned Expression (31) is taken as the weight to be determined by a function indicated in the following Expression (33), a degree of tracking for a more transient signal that the case of using the weight indicated in Expression (32) can be improved. [Mathematical Expression 33]

[00558] Note that, in Expression (33), ns indicates the number of frames ns of an input signal to be used for smoothing.
[00559] In the event that the SC(1) weight is taken as a smoothing parameter, the parameter determination unit 121 determines the SC(1) weight that serves as a smoothing parameter based on the subband signal high frequency. Smoothing information indicating the SC(1) weight which serves as a smoothing parameter is taken as high frequency encoded data and is transmitted to the decoding device 40.
[00560] In this case, also, for example, the value of the SC(l) weight itself, that is, the SC(0) weight up to the SC(ns - 1) weight, can be taken as uniformity information, or multiples SC(l) weights are prepared in advance, and from these, an index indicating the selected SC(l) weight can be taken as commonality information.
[00561] In the decoding device 40, the weight SC(1) obtained by decoding the encoded data at high frequency and identified by the smoothing information is employed to perform smoothing of a low frequency subband energy. Additionally, both the SC(l) weight and the number of frames ns are taken as smoothing parameters, and an index indicating the SC(l) weight, an indicator indicating the number of frames ns and the like can be taken as information of standardization.
[00562] Additionally, although description has been made in relation to a case where the third mode is applied as an example where multiple coefficient groups are prepared in advance and a low frequency subband energy that serves as an amount of feature is standardized, this example can be applied to any of the above-mentioned first modality up to fifth modality. That is, in a case where this example is also applied to any of the modalities, an amount of resource is smoothed according to a smoothing parameter and the amount of resource after smoothing is used to calculate the estimated energy value of subband of each subband on the high frequency side.
[00563] The above-described series of processing can be performed not only by hardware, but also by software. In the event of execution of the processing series using software, a program constituting the software thereof is installed from program recording media on a computer embedded in dedicated hardware or, for example, on a general purpose personal computer or congeners, according to which, various functions can be performed by installing various programs.
[00564] Fig. 34 is a block diagram illustrating an example of hardware configuration of a computer that performs the aforementioned series of processing using a program.
[00565] In the computer, a CPU 501, a ROM (Read Only Memory) 502 and a RAM (Random Access Memory) 503 are mutually connected by a bus 504.
[00566] Additionally, an input/output interface 505 is connected to the bus 504. An input unit 506 is connected to the input/output interface 506 consisting of a keyboard, a mouse, a microphone and the like, an output unit 507 consisting of a display, a speaker and the like, a storage unit 508 consisting of a hard disk, a non-volatile memory and the like, a communication unit 509 consisting of a network interface and the like, and a disk drive 510 that drives a removable 511 media such as a magnetic disk, optical disk, magneto-optical disk, semiconductor memory, or the like.
[00567] In the computer thus configured, the aforementioned processing series is performed by the CPU 501 which loads a program stored in the storage unit 508 into the RAM 503 via the input/output interface 505 and the bus 504 and executes it, for example .
[00568] The program that the computer (CPU 501) runs is provided by recording on removable media 511, which is a packaged media consisting, for example, of a magnetic disk (including a floppy disk), an optical disk (CD- ROM (Read Exclusive Compact Disc), DVD (Digital Versatile Disc), etc.), a magneto-optical disc, semiconductor memory, or the like, or is provided via cable or wireless transmission media such as a local area network, the Internet, a digital satellite broadcast or the like.
[00569] The program can be installed on the storage unit 508 via the input/output interface 505 by mounting the removable media 511 on the disk drive 510. Also, the program can be installed on the storage unit 508 by receiving on the unit 509 communication via cable media or wireless transmission. Additionally, the program can be installed in ROM 502 or storage unit 508 in advance.
[00570] Note that the program that the computer runs can be a program whose processing is performed in a time series manner along the sequence described in this Specification, or a program whose processing is performed in parallel or in the required timing, such as called being performed, or the like.
[00571] Note that the embodiments of the present invention are not restricted to the aforementioned embodiments, and various modifications can be made without departing from the essence of the present invention. LIST OF REFERENCE SIGNALS 10 frequency band expansion device 11 low pass filter 12 delay circuit 13 , 13-1 to 13-N bandpass filter 14 resource quantity calculation circuit 15 sub power estimation circuit high frequency band 16 high frequency signal generation circuit 17 high pass filter 18 signal adder 20 coefficient learning device 21, 21-1 to 21-(K+N) band pass filter 22 power calculation circuit high frequency subband 23 resource quantity calculation circuit 24 coefficient estimation circuit 30 encoding device 31 low pass filter 32 low frequency encoding circuit 33 subband division circuit 34 quantity calculation circuit Feature 35 pseudo-high-frequency sub-band power calculation circuit 36 pseudo-high-frequency sub-band power difference calculation circuit 37 high-frequency coding circuit ia 38 multiplexing circuit 40 decoding device 41 demultiplexing circuit 42 low frequency decoding circuit 43 subband division circuit 44 resource quantity calculation circuit 45 high frequency decoding circuit 46 power calculation circuit decoded high frequency subband 47 decoded high frequency signal generating circuit 48 synthesizing circuit 50 coefficient learning device 51 low pass filter 52 subband division circuit 53 resource quantity calculation circuit 54 pseudohigh frequency subband energy calculation 55 pseudohigh frequency subband energy difference calculation circuit 56 pseudohigh frequency subband energy difference grouping circuit 57 coefficient estimating circuit 121 determination unit of parameter 122 uniformity unit 151 uniformity unit

权利要求:
Claims (6)
[0001]
1. Decoding device (40), characterized in that it comprises: a demultiplexing circuit (41) configured to demultiplex input coded data into low frequency coded data, coefficient information to obtain a coefficient, and smoothing information with respect to the standardization; a low frequency decoding circuit (42) configured to decode the low frequency encoded data to generate a low frequency signal; a subband division circuit (43) configured to divide the low frequency signal into a plurality of subbands to generate a low frequency subband signal for each of the subbands; a resource amount calculating circuit (44) configured to calculate a resource amount based on the low frequency subband signals, wherein the resource amount includes low frequency subband energy of the subband signals. low frequency band; a smoothing circuit configured to subject the resource amount to smoothing by performing a weighted average on the resource amount of a predetermined number of continuous frames of the low frequency signal based on the smoothing information; and a generating circuit (47) configured to generate a high frequency signal based on the coefficient obtained from the coefficient information, the amount of resource subjected to smoothing, and the low frequency subband signals.
[0002]
2. Decoding device according to claim 1, characterized by the fact that the smoothing information is information that indicates at least one of the number of frames used for the weighted average or the weight used for the weighted average.
[0003]
3. Decoding device according to claim 1, characterized in that the generating circuit includes a decoded high frequency subband energy calculation circuit configured to calculate the decoded high frequency subband energy, which is an estimated value of the subband energy that constitutes the high frequency signal, based on the smoothed resource amount and coefficient, and a high frequency signal generating circuit configured to generate the high frequency signal based on in the decoded high frequency subband energy and in the low frequency subband signal.
[0004]
4. Decoding device according to claim 1, characterized in that the coefficient is generated by learning with the amount of resource obtained from a wideband supervisory signal and the energy of the same subband, such as a subband. -band which constitutes the high frequency signal of the broadband supervisor signal, as an explanatory variable and an explained variable.
[0005]
5. Decoding method, characterized in that it comprises the steps of: demultiplexing (S511), by the processing circuit, input coded data into low frequency coded data, coefficient information to obtain a coefficient and uniformity information regarding to standardization; decoding (S512), by the processing circuit, the low-frequency encoded data to generate a low-frequency signal; dividing (S513), by the processing circuit, the low frequency signal into a plurality of subbands to generate a low frequency subband signal for each of the subbands; calculate (S515), by the processing circuit, a resource amount based on the low frequency subband signals, wherein the resource amount includes low frequency subband energies of the low frequency subband signals ; subjecting (S516), by the processing circuit, the resource amount to smoothing by performing a weighted average on the resource amount of a predetermined number of continuous frames of the low frequency signal based on the smoothing information; and generating (S519), by the processing circuit, a high-frequency signal based on the coefficient obtained from the coefficient information, the amount of resource subjected to smoothing, and the low-frequency subband signals.
[0006]
6. Computer-readable storage medium, characterized in that it comprises computer-readable instructions which, when read by a computer, cause it to perform processing comprising the steps of: demultiplexing (S511) encoded input data into data low-frequency encoded coefficient information to obtain a coefficient and smoothing information in relation to the smoothing; decoding (S512) the low frequency encoded data to generate a low frequency signal; dividing (S513) the low frequency signal into a plurality of subbands to generate a low frequency subband signal for each of the subbands; calculating (S515) the resource amount based on the low frequency subband signals, wherein the resource amount includes low frequency subband energies of the low frequency subband signals; subjecting (S516) the resource amount to smoothing by performing a weighted average on the resource amount of a predetermined number of continuous frames of the low frequency signal based on the smoothing information; and generating (S519) a high frequency signal based on the coefficient obtained from the coefficient information, the amount of resource subject to smoothing and the low frequency subband signals.

类似技术:

---

公开号 | 公开日 | 专利标题

---

BR112013008490B1|2021-06-22|DECODING DEVICE AND METHOD, AND, COMPUTER-READABLE STORAGE MEANS

---

AU2019206091B2|2021-05-13|Frequency band extending device and method, encoding device and method, decoding device and method, and program

---

BR112012025570B1|2020-11-17|DECODER APPLIANCE AND DECODING METHOD

---

JP5652658B2|2015-01-14|Signal processing apparatus and method, encoding apparatus and method, decoding apparatus and method, and program

---

JP5609737B2|2014-10-22|Signal processing apparatus and method, encoding apparatus and method, decoding apparatus and method, and program

---

JP6508551B2|2019-05-08|Decryption apparatus and method, and program

---

JP6341306B2|2018-06-13|Signal processing apparatus and method, and program

同族专利:

---

公开号 | 公开日

---

BR112013008490A2|2016-08-09|

---

KR101835910B1|2018-03-07|

---

US20170076737A1|2017-03-16|

---

US9177563B2|2015-11-03|

---

RU2013115770A|2014-10-20|

---

US20130208902A1|2013-08-15|

---

US9536542B2|2017-01-03|

---

CA2811085A1|2012-04-19|

---

CN103155031B|2015-04-01|

---

AU2011314848A1|2013-03-28|

---

WO2012050023A1|2012-04-19|

---

CN103155031A|2013-06-12|

---

EP2608199A4|2016-04-20|

---

KR101979557B1|2019-05-16|

---

MX2013003952A|2013-06-03|

---

EP2608199A1|2013-06-26|

---

EP3579230A1|2019-12-11|

---

AR083365A1|2013-02-21|

---

RU2589293C2|2016-07-10|

---

TW201227719A|2012-07-01|

---

CA2811085C|2019-01-08|

---

RU2630384C1|2017-09-07|

---

US10236015B2|2019-03-19|

---

AU2011314848B2|2014-11-20|

---

JP2012083678A|2012-04-26|

---

KR20130141478A|2013-12-26|

---

TWI479481B|2015-04-01|

---

JP5707842B2|2015-04-30|

---

KR20180026791A|2018-03-13|

---

US20170352365A1|2017-12-07|

---

US9767824B2|2017-09-19|

---

US20160012829A1|2016-01-14|

---

EP2608199B1|2019-08-21|

引用文献:

---

公开号 | 申请日 | 公开日 | 申请人 | 专利标题

---

---

US4628529A|1985-07-01|1986-12-09|Motorola, Inc.|Noise suppression system|

---

JPH03254223A|1990-03-02|1991-11-13|Eastman Kodak Japan Kk|Analog data transmission system|

---

JP2655485B2|1994-06-24|1997-09-17|日本電気株式会社|Voice cell coding device|

---

JP3498375B2|1994-07-20|2004-02-16|ソニー株式会社|Digital audio signal recording device|

---

JP3189598B2|1994-10-28|2001-07-16|松下電器産業株式会社|Signal combining method and signal combining apparatus|

---

JPH1020888A|1996-07-02|1998-01-23|Matsushita Electric Ind Co Ltd|Voice coding/decoding device|

---

US6073100A|1997-03-31|2000-06-06|Goodridge, Jr.; Alan G|Method and apparatus for synthesizing signals using transform-domain match-output extension|

---

SE512719C2|1997-06-10|2000-05-02|Lars Gustaf Liljeryd|A method and apparatus for reducing data flow based on harmonic bandwidth expansion|

---

US6415251B1|1997-07-11|2002-07-02|Sony Corporation|Subband coder or decoder band-limiting the overlap region between a processed subband and an adjacent non-processed one|

---

US6978236B1|1999-10-01|2005-12-20|Coding Technologies Ab|Efficient spectral envelope coding using variable time/frequency resolution and time/frequency switching|

---

SE9903553D0|1999-01-27|1999-10-01|Lars Liljeryd|Enhancing conceptual performance of SBR and related coding methods by adaptive noise addition and noise substitution limiting |

---

US6829360B1|1999-05-14|2004-12-07|Matsushita Electric Industrial Co., Ltd.|Method and apparatus for expanding band of audio signal|

---

JP3454206B2|1999-11-10|2003-10-06|三菱電機株式会社|Noise suppression device and noise suppression method|

---

CA2290037A1|1999-11-18|2001-05-18|Voiceage Corporation|Gain-smoothing amplifier device and method in codecs for wideband speech and audio signals|

---

SE0004163D0|2000-11-14|2000-11-14|Coding Technologies Sweden Ab|Enhancing perceptual performance or high frequency reconstruction coding methods by adaptive filtering|

---

JP2002268698A|2001-03-08|2002-09-20|Nec Corp|Voice recognition device, device and method for standard pattern generation, and program|

---

SE0101175D0|2001-04-02|2001-04-02|Coding Technologies Sweden Ab|Aliasing reduction using complex-exponential-modulated filter banks|

---

JP4231987B2|2001-06-15|2009-03-04|日本電気株式会社|Code conversion method between speech coding / decoding systems, apparatus, program, and storage medium|

---

CN1235192C|2001-06-28|2006-01-04|皇家菲利浦电子有限公司|Wideband signal transmission system|

---

MXPA03002115A|2001-07-13|2003-08-26|Matsushita Electric Ind Co Ltd|Audio signal decoding device and audio signal encoding device.|

---

US6988066B2|2001-10-04|2006-01-17|At&T Corp.|Method of bandwidth extension for narrow-band speech|

---

US6895375B2|2001-10-04|2005-05-17|At&T Corp.|System for bandwidth extension of Narrow-band speech|

---

JP3926726B2|2001-11-14|2007-06-06|松下電器産業株式会社|Encoding device and decoding device|

---

CN100395817C|2001-11-14|2008-06-18|松下电器产业株式会社|Encoding device and decoding device|

---

JP3870193B2|2001-11-29|2007-01-17|コーディングテクノロジーズアクチボラゲット|Encoder, decoder, method and computer program used for high frequency reconstruction|

---

EP1470550B1|2002-01-30|2008-09-03|Matsushita Electric Industrial Co., Ltd.|Audio encoding and decoding device and methods thereof|

---

JP2003255973A|2002-02-28|2003-09-10|Nec Corp|Speech band expansion system and method therefor|

---

US20030187663A1|2002-03-28|2003-10-02|Truman Michael Mead|Broadband frequency translation for high frequency regeneration|

---

JP2003316394A|2002-04-23|2003-11-07|Nec Corp|System, method, and program for decoding sound|

---

US7447631B2|2002-06-17|2008-11-04|Dolby Laboratories Licensing Corporation|Audio coding system using spectral hole filling|

---

DE60327039D1|2002-07-19|2009-05-20|Nec Corp|AUDIO DEODICATION DEVICE, DECODING METHOD AND PROGRAM|

---

JP4728568B2|2002-09-04|2011-07-20|マイクロソフトコーポレーション|Entropy coding to adapt coding between level mode and run length / level mode|

---

JP3881943B2|2002-09-06|2007-02-14|松下電器産業株式会社|Acoustic encoding apparatus and acoustic encoding method|

---

SE0202770D0|2002-09-18|2002-09-18|Coding Technologies Sweden Ab|Method of reduction of aliasing is introduced by spectral envelope adjustment in real-valued filterbanks|

---

CN100492492C|2002-09-19|2009-05-27|松下电器产业株式会社|Audio decoding apparatus and method|

---

US7330812B2|2002-10-04|2008-02-12|National Research Council Of Canada|Method and apparatus for transmitting an audio stream having additional payload in a hidden sub-channel|

---

EP2665294A2|2003-03-04|2013-11-20|Core Wireless Licensing S.a.r.l.|Support of a multichannel audio extension|

---

US7318035B2|2003-05-08|2008-01-08|Dolby Laboratories Licensing Corporation|Audio coding systems and methods using spectral component coupling and spectral component regeneration|

---

US20050004793A1|2003-07-03|2005-01-06|Pasi Ojala|Signal adaptation for higher band coding in a codec utilizing band split coding|

---

KR20050027179A|2003-09-13|2005-03-18|삼성전자주식회사|Method and apparatus for decoding audio data|

---

US7844451B2|2003-09-16|2010-11-30|Panasonic Corporation|Spectrum coding/decoding apparatus and method for reducing distortion of two band spectrums|

---

EP1657710B1|2003-09-16|2009-05-27|Panasonic Corporation|Coding apparatus and decoding apparatus|

---

CN101556800B|2003-10-23|2012-05-23|松下电器产业株式会社|Acoustic spectrum coding method and apparatus, spectrum decoding method and apparatus, acoustic signal transmission apparatus, acoustic signal reception apparatus|

---

KR100587953B1|2003-12-26|2006-06-08|한국전자통신연구원|Packet loss concealment apparatus for high-band in split-band wideband speech codec, and system for decoding bit-stream using the same|

---

WO2005111568A1|2004-05-14|2005-11-24|Matsushita Electric Industrial Co., Ltd.|Encoding device, decoding device, and method thereof|

---

CN1954363B|2004-05-19|2011-10-12|松下电器产业株式会社|Encoding device and method thereof|

---

AT474310T|2004-05-28|2010-07-15|Nokia Corp|MULTI-CHANNEL AUDIO EXPANSION|

---

KR100608062B1|2004-08-04|2006-08-02|삼성전자주식회사|Method and apparatus for decoding high frequency of audio data|

---

US7716046B2|2004-10-26|2010-05-11|Qnx Software Systems , Inc.|Advanced periodic signal enhancement|

---

US20060106620A1|2004-10-28|2006-05-18|Thompson Jeffrey K|Audio spatial environment down-mixer|

---

SE0402651D0|2004-11-02|2004-11-02|Coding Tech Ab|Advanced methods for interpolation and parameter signaling|

---

EP2752849B1|2004-11-05|2020-06-03|Panasonic Intellectual Property Management Co., Ltd.|Encoder and encoding method|

---

BRPI0517780A2|2004-11-05|2011-04-19|Matsushita Electric Ind Co Ltd|scalable decoding device and scalable coding device|

---

KR100657916B1|2004-12-01|2006-12-14|삼성전자주식회사|Apparatus and method for processing audio signal using correlation between bands|

---

JP5224017B2|2005-01-11|2013-07-03|日本電気株式会社|Audio encoding apparatus, audio encoding method, and audio encoding program|

---

KR100707177B1|2005-01-19|2007-04-13|삼성전자주식회사|Method and apparatus for encoding and decoding of digital signals|

---

KR100708121B1|2005-01-22|2007-04-16|삼성전자주식회사|Method and apparatus for bandwidth extension of speech|

---

TWI319565B|2005-04-01|2010-01-11|Qualcomm Inc|Methods, and apparatus for generating highband excitation signal|

---

CN101138274B|2005-04-15|2011-07-06|杜比国际公司|Envelope shaping of decorrelated signals|

---

US20070005351A1|2005-06-30|2007-01-04|Sathyendra Harsha M|Method and system for bandwidth expansion for voice communications|

---

JP4899359B2|2005-07-11|2012-03-21|ソニー株式会社|Signal encoding apparatus and method, signal decoding apparatus and method, program, and recording medium|

---

KR100813259B1|2005-07-13|2008-03-13|삼성전자주식회사|Method and apparatus for encoding/decoding input signal|

---

WO2007026821A1|2005-09-02|2007-03-08|Matsushita Electric Industrial Co., Ltd.|Energy shaping device and energy shaping method|

---

US8396717B2|2005-09-30|2013-03-12|Panasonic Corporation|Speech encoding apparatus and speech encoding method|

---

BRPI0617447A2|2005-10-14|2012-04-17|Matsushita Electric Ind Co Ltd|transform encoder and transform coding method|

---

KR100958144B1|2005-11-04|2010-05-18|노키아 코포레이션|Audio Compression|

---

EP1959433B1|2005-11-30|2011-10-19|Panasonic Corporation|Subband coding apparatus and method of coding subband|

---

JP4876574B2|2005-12-26|2012-02-15|ソニー株式会社|Signal encoding apparatus and method, signal decoding apparatus and method, program, and recording medium|

---

JP4863713B2|2005-12-29|2012-01-25|富士通株式会社|Noise suppression device, noise suppression method, and computer program|

---

RU2414009C2|2006-01-18|2011-03-10|ЭлДжи ЭЛЕКТРОНИКС ИНК.|Signal encoding and decoding device and method|

---

US7953604B2|2006-01-20|2011-05-31|Microsoft Corporation|Shape and scale parameters for extended-band frequency coding|

---

US7590523B2|2006-03-20|2009-09-15|Mindspeed Technologies, Inc.|Speech post-processing using MDCT coefficients|

---

JP4976381B2|2006-03-31|2012-07-18|パナソニック株式会社|Speech coding apparatus, speech decoding apparatus, and methods thereof|

---

EP2012305B1|2006-04-27|2011-03-09|Panasonic Corporation|Audio encoding device, audio decoding device, and their method|

---

EP2017830B9|2006-05-10|2011-02-23|Panasonic Corporation|Encoding device and encoding method|

---

JP2007316254A|2006-05-24|2007-12-06|Sony Corp|Audio signal interpolation method and audio signal interpolation device|

---

KR20070115637A|2006-06-03|2007-12-06|삼성전자주식회사|Method and apparatus for bandwidth extension encoding and decoding|

---

JP2007333785A|2006-06-12|2007-12-27|Matsushita Electric Ind Co Ltd|Audio signal encoding device and audio signal encoding method|

---

WO2007148925A1|2006-06-21|2007-12-27|Samsung Electronics Co., Ltd.|Method and apparatus for adaptively encoding and decoding high frequency band|

---

KR101244310B1|2006-06-21|2013-03-18|삼성전자주식회사|Method and apparatus for wideband encoding and decoding|

---

US8260609B2|2006-07-31|2012-09-04|Qualcomm Incorporated|Systems, methods, and apparatus for wideband encoding and decoding of inactive frames|

---

EP2063418A4|2006-09-15|2010-12-15|Panasonic Corp|Audio encoding device and audio encoding method|

---

JP4918841B2|2006-10-23|2012-04-18|富士通株式会社|Encoding system|

---

JP5141180B2|2006-11-09|2013-02-13|ソニー株式会社|Frequency band expanding apparatus, frequency band expanding method, reproducing apparatus and reproducing method, program, and recording medium|

---

US8295507B2|2006-11-09|2012-10-23|Sony Corporation|Frequency band extending apparatus, frequency band extending method, player apparatus, playing method, program and recording medium|

---

KR101565919B1|2006-11-17|2015-11-05|삼성전자주식회사|Method and apparatus for encoding and decoding high frequency signal|

---

JP4930320B2|2006-11-30|2012-05-16|ソニー株式会社|Reproduction method and apparatus, program, and recording medium|

---

JP5339919B2|2006-12-15|2013-11-13|パナソニック株式会社|Encoding device, decoding device and methods thereof|

---

JP4984983B2|2007-03-09|2012-07-25|富士通株式会社|Encoding apparatus and encoding method|

---

JP2008261978A|2007-04-11|2008-10-30|Toshiba Microelectronics Corp|Reproduction volume automatically adjustment method|

---

US8015368B2|2007-04-20|2011-09-06|Siport, Inc.|Processor extensions for accelerating spectral band replication|

---

KR101355376B1|2007-04-30|2014-01-23|삼성전자주식회사|Method and apparatus for encoding and decoding high frequency band|

---

WO2009001874A1|2007-06-27|2008-12-31|Nec Corporation|Audio encoding method, audio decoding method, audio encoding device, audio decoding device, program, and audio encoding/decoding system|

---

WO2009004727A1|2007-07-04|2009-01-08|Fujitsu Limited|Encoding apparatus, encoding method and encoding program|

---

JP5045295B2|2007-07-30|2012-10-10|ソニー株式会社|Signal processing apparatus and method, and program|

---

US8041577B2|2007-08-13|2011-10-18|Mitsubishi Electric Research Laboratories, Inc.|Method for expanding audio signal bandwidth|

---

MX2010001394A|2007-08-27|2010-03-10|Ericsson Telefon Ab L M|Adaptive transition frequency between noise fill and bandwidth extension.|

---

DK3401907T3|2007-08-27|2020-03-02|Ericsson Telefon Ab L M|Method and apparatus for perceptual spectral decoding of an audio signal comprising filling in spectral holes|

---

CA2697920C|2007-08-27|2018-01-02|Telefonaktiebolaget L M Ericsson |Transient detector and method for supporting encoding of an audio signal|

---

EP2209116B8|2007-10-23|2014-08-06|Clarion Co., Ltd.|Device and method for high-frequency range interpolation of an audio signal|

---

KR101373004B1|2007-10-30|2014-03-26|삼성전자주식회사|Apparatus and method for encoding and decoding high frequency signal|

---

JP4733727B2|2007-10-30|2011-07-27|日本電信電話株式会社|Voice musical tone pseudo-wideband device, voice musical tone pseudo-bandwidth method, program thereof, and recording medium thereof|

---

EP2214163A4|2007-11-01|2011-10-05|Panasonic Corp|Encoding device, decoding device, and method thereof|

---

RU2449386C2|2007-11-02|2012-04-27|Хуавэй Текнолоджиз Ко., Лтд.|Audio decoding method and apparatus|

---

US20090132238A1|2007-11-02|2009-05-21|Sudhakar B|Efficient method for reusing scale factors to improve the efficiency of an audio encoder|

---

WO2009059631A1|2007-11-06|2009-05-14|Nokia Corporation|Audio coding apparatus and method thereof|

---

JP2009116275A|2007-11-09|2009-05-28|Toshiba Corp|Method and device for noise suppression, speech spectrum smoothing, speech feature extraction, speech recognition and speech model training|

---

US8504377B2|2007-11-21|2013-08-06|Lg Electronics Inc.|Method and an apparatus for processing a signal using length-adjusted window|

---

US8688441B2|2007-11-29|2014-04-01|Motorola Mobility Llc|Method and apparatus to facilitate provision and use of an energy value to determine a spectral envelope shape for out-of-signal bandwidth content|

---

ES2629453T3|2007-12-21|2017-08-09|Iii Holdings 12, Llc|Encoder, decoder and coding procedure|

---

US20100280833A1|2007-12-27|2010-11-04|Panasonic Corporation|Encoding device, decoding device, and method thereof|

---

AT518224T|2008-01-04|2011-08-15|Dolby Int Ab|AUDIO CODERS AND DECODERS|

---

JP5448850B2|2008-01-25|2014-03-19|パナソニック株式会社|Encoding device, decoding device and methods thereof|

---

KR101413968B1|2008-01-29|2014-07-01|삼성전자주식회사|Method and apparatus for encoding audio signal, and method and apparatus for decoding audio signal|

---

US8433582B2|2008-02-01|2013-04-30|Motorola Mobility Llc|Method and apparatus for estimating high-band energy in a bandwidth extension system|

---

US20090201983A1|2008-02-07|2009-08-13|Motorola, Inc.|Method and apparatus for estimating high-band energy in a bandwidth extension system|

---

AU2009220321B2|2008-03-03|2011-09-22|Intellectual Discovery Co., Ltd.|Method and apparatus for processing audio signal|

---

KR101449434B1|2008-03-04|2014-10-13|삼성전자주식회사|Method and apparatus for encoding/decoding multi-channel audio using plurality of variable length code tables|

---

EP2104096B1|2008-03-20|2020-05-06|Fraunhofer-Gesellschaft zur Förderung der angewandten Forschung e.V.|Apparatus and method for converting an audio signal into a parameterized representation, apparatus and method for modifying a parameterized representation, apparatus and method for synthesizing a parameterized representation of an audio signal|

---

KR20090122142A|2008-05-23|2009-11-26|엘지전자 주식회사|A method and apparatus for processing an audio signal|

---

EP2294770B1|2008-06-20|2013-08-07|Rambus, Inc.|Frequency responsive bus coding|

---

MX2011000372A|2008-07-11|2011-05-19|Fraunhofer Ges Forschung|Audio signal synthesizer and audio signal encoder.|

---

KR101706009B1|2008-07-11|2017-02-22|프라운호퍼 게젤샤프트 쭈르 푀르데룽 데어 안겐반텐 포르슝 에. 베.|Audio encoder, audio decoder, method for encoding and decoding an audio signal. audio stream and computer program|

---

JP5203077B2|2008-07-14|2013-06-05|株式会社エヌ・ティ・ティ・ドコモ|Speech coding apparatus and method, speech decoding apparatus and method, and speech bandwidth extension apparatus and method|

---

JP5419876B2|2008-08-08|2014-02-19|パナソニック株式会社|Spectrum smoothing device, coding device, decoding device, communication terminal device, base station device, and spectrum smoothing method|

---

JP2010079275A|2008-08-29|2010-04-08|Sony Corp|Device and method for expanding frequency band, device and method for encoding, device and method for decoding, and program|

---

WO2010028292A1|2008-09-06|2010-03-11|Huawei Technologies Co., Ltd.|Adaptive frequency prediction|

---

US8352279B2|2008-09-06|2013-01-08|Huawei Technologies Co., Ltd.|Efficient temporal envelope coding approach by prediction between low band signal and high band signal|

---

WO2010028299A1|2008-09-06|2010-03-11|Huawei Technologies Co., Ltd.|Noise-feedback for spectral envelope quantization|

---

US8798776B2|2008-09-30|2014-08-05|Dolby International Ab|Transcoding of audio metadata|

---

GB2466201B|2008-12-10|2012-07-11|Skype Ltd|Regeneration of wideband speech|

---

GB0822537D0|2008-12-10|2009-01-14|Skype Ltd|Regeneration of wideband speech|

---

CN101770776B|2008-12-29|2011-06-08|华为技术有限公司|Coding method and device, decoding method and device for instantaneous signal and processing system|

---

SG172976A1|2009-01-16|2011-08-29|Dolby Int Ab|Cross product enhanced harmonic transposition|

---

US8457975B2|2009-01-28|2013-06-04|Fraunhofer-Gesellschaft Zur Foerderung Der Angewandten Forschung E.V.|Audio decoder, audio encoder, methods for decoding and encoding an audio signal and computer program|

---

JP4945586B2|2009-02-02|2012-06-06|株式会社東芝|Signal band expander|

---

US8463599B2|2009-02-04|2013-06-11|Motorola Mobility Llc|Bandwidth extension method and apparatus for a modified discrete cosine transform audio coder|

---

EP2402940B9|2009-02-26|2019-10-30|Panasonic Intellectual Property Corporation of America|Encoder, decoder, and method therefor|

---

JP5564803B2|2009-03-06|2014-08-06|ソニー株式会社|Acoustic device and acoustic processing method|

---

CN101853663B|2009-03-30|2012-05-23|华为技术有限公司|Bit allocation method, encoding device and decoding device|

---

CO6440537A2|2009-04-09|2012-05-15|Fraunhofer Ges Forschung|APPARATUS AND METHOD TO GENERATE A SYNTHESIS AUDIO SIGNAL AND TO CODIFY AN AUDIO SIGNAL|

---

EP2239732A1|2009-04-09|2010-10-13|Fraunhofer-Gesellschaft zur Förderung der Angewandten Forschung e.V.|Apparatus and method for generating a synthesis audio signal and for encoding an audio signal|

---

JP5223786B2|2009-06-10|2013-06-26|富士通株式会社|Voice band extending apparatus, voice band extending method, voice band extending computer program, and telephone|

---

US8515768B2|2009-08-31|2013-08-20|Apple Inc.|Enhanced audio decoder|

---

JP5928539B2|2009-10-07|2016-06-01|ソニー株式会社|Encoding apparatus and method, and program|

---

JP5754899B2|2009-10-07|2015-07-29|ソニー株式会社|Decoding apparatus and method, and program|

---

CN102081927B|2009-11-27|2012-07-18|中兴通讯股份有限公司|Layering audio coding and decoding method and system|

---

US8600749B2|2009-12-08|2013-12-03|At&T Intellectual Property I, L.P.|System and method for training adaptation-specific acoustic models for automatic speech recognition|

---

US8447617B2|2009-12-21|2013-05-21|Mindspeed Technologies, Inc.|Method and system for speech bandwidth extension|

---

EP2357649B1|2010-01-21|2012-12-19|Electronics and Telecommunications Research Institute|Method and apparatus for decoding audio signal|

---

TWI447709B|2010-02-11|2014-08-01|Dolby Lab Licensing Corp|System and method for non-destructively normalizing loudness of audio signals within portable devices|

---

JP5375683B2|2010-03-10|2013-12-25|富士通株式会社|Communication apparatus and power correction method|

---

EP2555188B1|2010-03-31|2014-05-14|Fujitsu Limited|Bandwidth extension apparatuses and methods|

---

JP5850216B2|2010-04-13|2016-02-03|ソニー株式会社|Signal processing apparatus and method, encoding apparatus and method, decoding apparatus and method, and program|

---

JP5609737B2|2010-04-13|2014-10-22|ソニー株式会社|Signal processing apparatus and method, encoding apparatus and method, decoding apparatus and method, and program|

---

JP5652658B2|2010-04-13|2015-01-14|ソニー株式会社|Signal processing apparatus and method, encoding apparatus and method, decoding apparatus and method, and program|

---

US8793126B2|2010-04-14|2014-07-29|Huawei Technologies Co., Ltd.|Time/frequency two dimension post-processing|

---

PL3544008T3|2010-07-19|2020-08-24|Dolby International Ab|Processing of audio signals during high frequency reconstruction|

---

US9047875B2|2010-07-19|2015-06-02|Futurewei Technologies, Inc.|Spectrum flatness control for bandwidth extension|

---

US8560330B2|2010-07-19|2013-10-15|Futurewei Technologies, Inc.|Energy envelope perceptual correction for high band coding|

---

JP6075743B2|2010-08-03|2017-02-08|ソニー株式会社|Signal processing apparatus and method, and program|

---

JP2012058358A|2010-09-07|2012-03-22|Sony Corp|Noise suppression apparatus, noise suppression method and program|

---

JP5707842B2|2010-10-15|2015-04-30|ソニー株式会社|Encoding apparatus and method, decoding apparatus and method, and program|

---

WO2012052802A1|2010-10-18|2012-04-26|Nokia Corporation|An audio encoder/decoder apparatus|

---

JP5743137B2|2011-01-14|2015-07-01|ソニー株式会社|Signal processing apparatus and method, and program|

---

JP5704397B2|2011-03-31|2015-04-22|ソニー株式会社|Encoding apparatus and method, and program|

---

JP6024077B2|2011-07-01|2016-11-09|ヤマハ株式会社|Signal transmitting apparatus and signal processing apparatus|

---

JP5975243B2|2011-08-24|2016-08-23|ソニー株式会社|Encoding apparatus and method, and program|

---

JP6037156B2|2011-08-24|2016-11-30|ソニー株式会社|Encoding apparatus and method, and program|

---

JP5942358B2|2011-08-24|2016-06-29|ソニー株式会社|Encoding apparatus and method, decoding apparatus and method, and program|

---

JP5845760B2|2011-09-15|2016-01-20|ソニー株式会社|Audio processing apparatus and method, and program|

---

CN103918030B|2011-09-29|2016-08-17|杜比国际公司|High quality detection in the FM stereo radio signal of telecommunication|

---

JPWO2013154027A1|2012-04-13|2015-12-17|ソニー株式会社|Decoding device and method, audio signal processing device and method, and program|

---

JP5997592B2|2012-04-27|2016-09-28|株式会社Ｎｔｔドコモ|Speech decoder|

---

TWI517142B|2012-07-02|2016-01-11|Sony Corp|Audio decoding apparatus and method, audio coding apparatus and method, and program|

---

US9437198B2|2012-07-02|2016-09-06|Sony Corporation|Decoding device, decoding method, encoding device, encoding method, and program|

---

AU2013284705B2|2012-07-02|2018-11-29|Sony Corporation|Decoding device and method, encoding device and method, and program|

---

RU2649944C2|2012-07-02|2018-04-05|Сони Корпорейшн|Decoding device, decoding method, coding device, coding method and program|

---

JP2014123011A|2012-12-21|2014-07-03|Sony Corp|Noise detector, method, and program|

---

EP3048609A4|2013-09-19|2017-05-03|Sony Corporation|Encoding device and method, decoding device and method, and program|JP5754899B2|2009-10-07|2015-07-29|ソニー株式会社|Decoding apparatus and method, and program|

---

JP5652658B2|2010-04-13|2015-01-14|ソニー株式会社|Signal processing apparatus and method, encoding apparatus and method, decoding apparatus and method, and program|

---

JP5609737B2|2010-04-13|2014-10-22|ソニー株式会社|Signal processing apparatus and method, encoding apparatus and method, decoding apparatus and method, and program|

---

JP5850216B2|2010-04-13|2016-02-03|ソニー株式会社|Signal processing apparatus and method, encoding apparatus and method, decoding apparatus and method, and program|

---

JP6075743B2|2010-08-03|2017-02-08|ソニー株式会社|Signal processing apparatus and method, and program|

---

JP5707842B2|2010-10-15|2015-04-30|ソニー株式会社|Encoding apparatus and method, decoding apparatus and method, and program|

---

JP5743137B2|2011-01-14|2015-07-01|ソニー株式会社|Signal processing apparatus and method, and program|

---

JP6037156B2|2011-08-24|2016-11-30|ソニー株式会社|Encoding apparatus and method, and program|

---

JP5975243B2|2011-08-24|2016-08-23|ソニー株式会社|Encoding apparatus and method, and program|

---

JP5942358B2|2011-08-24|2016-06-29|ソニー株式会社|Encoding apparatus and method, decoding apparatus and method, and program|

---

CN104718572B|2012-06-04|2018-07-31|三星电子株式会社|Audio coding method and device, audio-frequency decoding method and device and the multimedia device using this method and device|

---

RU2649944C2|2012-07-02|2018-04-05|Сони Корпорейшн|Decoding device, decoding method, coding device, coding method and program|

---

JP6305694B2|2013-05-31|2018-04-04|クラリオン株式会社|Signal processing apparatus and signal processing method|

---

CN111477245A|2013-06-11|2020-07-31|弗朗霍弗应用研究促进协会|Speech signal decoding device and speech signal encoding device|

---

JP6531103B2|2013-09-12|2019-06-12|ドルビー・インターナショナル・アーベー|QMF based processing data time alignment|

---

US10469969B2|2013-09-17|2019-11-05|Wilus Institute Of Standards And Technology Inc.|Method and apparatus for processing multimedia signals|

---

EP3048609A4|2013-09-19|2017-05-03|Sony Corporation|Encoding device and method, decoding device and method, and program|

---

EP3062535B1|2013-10-22|2019-07-03|Industry-Academic Cooperation Foundation, Yonsei University|Method and apparatus for processing audio signal|

---

US20150170655A1|2013-12-15|2015-06-18|Qualcomm Incorporated|Systems and methods of blind bandwidth extension|

---

KR20210094125A|2013-12-23|2021-07-28|주식회사 윌러스표준기술연구소|Method for generating filter for audio signal, and parameterization device for same|

---

AU2014371411A1|2013-12-27|2016-06-23|Sony Corporation|Decoding device, method, and program|

---

EP3122073A4|2014-03-19|2017-10-18|Wilus Institute of Standards and Technology Inc.|Audio signal processing method and apparatus|

---

KR101856540B1|2014-04-02|2018-05-11|주식회사 윌러스표준기술연구소|Audio signal processing method and device|

---

JP2016038435A|2014-08-06|2016-03-22|ソニー株式会社|Encoding device and method, decoding device and method, and program|

---

EP3818528A1|2018-07-03|2021-05-12|Soclip!|Beat decomposition to facilitate automatic video editing|

法律状态:
2018-12-26| B06F| Objections, documents and/or translations needed after an examination request according [chapter 6.6 patent gazette]|

---

2019-09-10| B06U| Preliminary requirement: requests with searches performed by other patent offices: procedure suspended [chapter 6.21 patent gazette]|

---

2020-11-17| B06A| Notification to applicant to reply to the report for non-patentability or inadequacy of the application [chapter 6.1 patent gazette]|

---

2021-03-30| B09A| Decision: intention to grant [chapter 9.1 patent gazette]|

---

2021-06-22| B16A| Patent or certificate of addition of invention granted|Free format text: PRAZO DE VALIDADE: 20 (VINTE) ANOS CONTADOS A PARTIR DE 05/10/2011, OBSERVADAS AS CONDICOES LEGAIS. |

优先权:

---

申请号 | 申请日 | 专利标题

---

JP2010-232106|2010-10-15|

---

JP2010232106A|JP5707842B2|2010-10-15|2010-10-15|Encoding apparatus and method, decoding apparatus and method, and program|

---

PCT/JP2011/072957|WO2012050023A1|2010-10-15|2011-10-05|Encoding device and method, decoding device and method, and program|

---