JP5812842B2 - Audio equipment - Google Patents

Description

  The present invention relates to an audio apparatus that performs 5.1ch or 7.1ch audio reproduction.

  2. Description of the Related Art Conventionally, there has been known an audio apparatus that generates 7.1 ch audio reproduction by generating two sets of surround signals based on an LS signal and an RS signal when 5.1 ch audio data is input. (For example, refer to Patent Document 1). Further, Patent Documents 2 to 4 are known as conventional techniques for generating a surround signal by a similar method.

JP 2007-67463 A JP 2006-319694 A Japanese Patent No. 3682032 JP 2010-103768 A

  By using the method disclosed in Patent Document 1, it is possible to generate a 7.1ch audio signal from a 5.1ch audio signal. Therefore, by using a 7.1ch speaker or amplifier, 5.1ch And 7.1ch audio playback can be performed.

  On the other hand, audio contents recorded on a Blu-ray (registered trademark) disc include a 7.1ch format in addition to 2ch to 5.1ch.

  Further, in the case of a general audio device mounted on a vehicle, it is difficult to adopt a configuration using many speakers as disclosed in Patent Document 1. In general, four speakers are arranged in the front, rear, left and right of the vehicle interior. In order to output the 5.1ch and 7.1ch audio signals described above from these four speakers, these audio signals are combined (down). It is necessary to input to each speaker after mixing.

  By the way, in the 5.1ch or 7.1ch format recorded on the Blu-ray disc, the space defined by ITU-R (International Telecommunication Union Radiocommunication Division) (for example, about 2 m between each speaker and the listening position) Since the surround sound is optimized while monitoring the distance (which is assumed to be far away), the localization is limited in the passenger compartment where the listening position is set closer to the distance assumed in these formats. There is a problem that the reproduction sound is unbalanced and the playback sound is uncomfortable.

  In addition, when the number of audio signals is simply reduced by downmixing, there is a problem that it is difficult to make a difference between 5.1ch and 7.1ch audio reproduction. In general, regardless of 5.1ch or 7.1ch, when stereo playback is performed after down-mixing a large number of channel source signals, compared to the case where the left and right 2ch signals originally intended for stereo playback are played back as they are, It has been confirmed that the sound image corresponding to the speaker tends to be small. For example, when a sound source recorded in 2ch left and right is played back, the sound image of the dialogue of the movie becomes large, but when downmixing a sound source recorded in 5.1ch, the C component (center speaker) in which the dialogue is mainly recorded The sound of the component that is output from (1) may be smaller than the sound of the other components. In particular, when recorded in 7.1ch, since the left and right back channels are added, the relative proportion of the C component tends to further decrease.

  The present invention was created in view of the above points, and its purpose is to produce a sense of incongruity in the reproduced sound due to localization bias when audio reproduction is performed in a vehicle interior using speakers with fewer channels. An object of the present invention is to provide an audio device that can prevent this. Another object of the present invention is to provide an audio device capable of expressing a difference in audio reproduction according to an audio format when audio reproduction is performed in a vehicle interior using speakers smaller than the number of channels. It is in.

  In order to solve the above-described problems, an audio apparatus according to the present invention is an audio apparatus that performs audio reproduction using a smaller number of speakers than the number of channels, and each audio signal corresponding to a predetermined audio format is provided for each audio signal. Downmix processing means for generating a pair of left and right intermediate audio signals by performing a downmix process for synthesizing channel signals, and using an adaptive algorithm with a predetermined step size parameter for each of the pair of intermediate audio signals A surround signal generating means for generating a pair of low-correlated left and right surround signals and a highly correlated center signal by performing a decorrelation process, and combining the intermediate audio signal and the surround signal after weighting Process to generate a composite surround signal separately Combining means for separately processing the left and right signals after weighting the intermediate audio signal and the center signal to generate a left signal and a right signal, and a step size parameter corresponding to each of a plurality of audio formats And a control means for setting a coefficient value used for weighting by the synthesizing means, and a set of left and right synthesized surround signals synthesized by the synthesizing means and the left and right signals are output from separate speakers. .

  Specifically, the surround signal generation means described above uses a pair of left and right intermediate audio signals to extract a component highly correlated with the other signal in one signal and subtract it from the other signal. The left and right intermediate audio signals are interchanged to generate a pair of left and right surround signals, and the surround signal generating means updates the filter coefficient of the adaptive filter using an adaptive algorithm, thereby generating a highly correlated component. The center signal is generated by extraction, and the control means sets the value of the step size parameter used when updating the filter coefficient.

  By performing a downmix process on the audio signal corresponding to each audio format and then performing a decorrelation process to generate a set of left and right surround signals, audio is generated in the vehicle interior using a speaker having fewer channels. It is possible to eliminate the localization bias in the case of reproduction, and to prevent the reproduced sound from feeling uncomfortable. In addition, by setting the step size parameter value used in the adaptive algorithm and the weighting coefficient value used in the synthesis means to correspond to each audio format, it is possible to make a difference in audio playback depending on the audio format. It becomes.

  The plurality of audio formats described above preferably have different numbers of channels. This eliminates the localization bias for each audio format with a different number of channels, and makes it easy to make a difference in audio reproduction between them.

  The plurality of audio formats described above preferably include 5.1-channel and 7.1-channel audio formats. As a result, it is easy to eliminate the localization bias for each of the 5.1 channel and 7.1 channel audio formats and to make a difference in audio reproduction between them.

In addition, it is preferable that the step size parameter described above has a larger value corresponding to 7.1 channel audio data than a value corresponding to 5.1 channel audio data. In addition, the weight corresponding to the surround signal performed to generate the above-described composite surround signal is larger in the value corresponding to 7.1 channel audio data than in the value corresponding to 5.1 channel audio data. It is desirable. As a result, it is possible to perform audio reproduction suitable for the 7.1-channel audio format with many back channels, and to easily make a difference from the 5.1-channel audio format.

  In addition, it is desirable to set the weighting value corresponding to the center signal performed to generate the center signal described above to 0.5 or more. Thereby, in the case of 5.1 channel or 7.1 channel, it can prevent that the sound image corresponding to a center speaker becomes small.

  In addition, the weight corresponding to the center signal performed to generate the center signal described above is larger in the value corresponding to 7.1 channel audio data than in the value corresponding to 5.1 channel audio data. Is desirable. Thereby, when it changes to 7.1 channel with many channels, it can prevent that the sound image corresponding to a center speaker becomes small.

It is a figure which shows the whole audio apparatus structure of one Embodiment. It is a figure which shows schematic structure of a disc reader. It is a figure which shows the speaker arrangement prescribed | regulated by ITU-R. It is a figure which shows the detailed structure of a downmix process part. It is a figure which shows the detailed structure of a decorrelation process part. It is a figure which shows the detailed structure of an adaptive filter. It is a figure which shows the relationship between five types of signals output from a decorrelation process part, and each signal of the audio format of 5.1ch or 7.1ch. It is a figure which shows the detailed structure of a mix process part.

  An audio apparatus according to an embodiment to which the present invention is applied will be described below with reference to the drawings. FIG. 1 is a diagram illustrating an overall configuration of an audio apparatus according to an embodiment. This audio device is mounted on a vehicle. As shown in FIG. 1, the audio apparatus according to the present embodiment includes a disc reader 10, a decorrelation processing unit 20, a mix processing unit 30, an EQ processing unit 40, a control unit 50, a digital-analog converter (D / A). ) 60, an amplifier 62, and speakers 70, 72, 74, and 76.

The disc reader 10 reads a compressed audio signal of a predetermined format recorded on a Blu-ray disc as a disc type recording medium, performs a decoding process and a downmix process, and outputs a set of signals L ₀ and R ₀ . FIG. 2 is a diagram showing a schematic configuration of the disk reader 10. As shown in FIG. 2, the disk reading device 10 includes a disk reading unit 12, a decoding processing unit 14, and a downmix processing unit 16.

  The disc reader 12 optically reads a signal recorded on a Blu-ray disc, performs demodulation processing on the read signal, and outputs audio data encoded in a predetermined format. For example, in this embodiment, it is assumed that audio data corresponding to each format of 2ch, 5.1ch, or 7.1ch is included.

  The decoding processing unit 14 receives encoded audio data having a predetermined channel component such as 5.1ch from the disk reading unit 12, and decodes the audio data. For example, when 5.1ch audio data is input, the audio data is decoded to generate a signal C, a signal L, a signal R, a signal Ls, and a signal Rs. When 7.1ch audio data is input, the audio data is decoded to generate signal C, signal L, signal R, signal Ls, signal Rs, signal Lb, and signal Rb. The signal C, signal L, signal R, signal Ls, signal Rs, signal Lb, and signal Rb are respectively a center channel, a left channel, a right channel, a left surround channel, a right surround channel, a left back channel, and a right back channel. Corresponding to each of the. When 2ch audio data is input, the audio data is decoded to generate a signal L and a signal R.

  FIG. 3 is a diagram showing a speaker arrangement defined by ITU-R. The 5.1ch or 7.1ch surround sound recorded on a Blu-ray disc or the like is adjusted according to the monitor listening environment installed in the studio. The speaker arrangement at that time is the arrangement shown in FIG. In ITU-R, the ideal distance from the listening position to each speaker is 1.5 m (Small Room) or 4.0 m (Large Room), and sound is created under these conditions. However, it is impossible to maintain these distances in the vehicle interior, and instability remains in audio reproduction simply by arranging speakers according to the shape of the vehicle interior. This point is improved by the present invention.

Downmix processing unit 16 performs a down-mixing process for combining the signals of each channel to a predetermined channel number of the audio signal, and generates a stereo signal L _0, R ₀ as the left and right pair of intermediate audio signals. FIG. 4 is a diagram illustrating a detailed configuration of the downmix processing unit 16. The downmix processing unit 16 receives signals corresponding to 5.1ch or 7.1ch (the above-described signals C, L, R, etc.).

  As shown in FIG. 4, the downmix processing unit 16 includes nine multipliers 161 to 169 and six adders 171 to 176. The multiplier 161 multiplies the input signal Ls by a and outputs it. The multiplier 162 multiplies the input signal Rs by a and outputs it. The multiplier 163 multiplies the input signal Lb by b and outputs it. The multiplier 164 multiplies the input signal Rb by b and outputs it. The adder 171 outputs a signal Ls ′ obtained by adding the signal (a × Ls) output from the multiplier 161 and the signal (b × Lb) output from the multiplier 163. The adder 172 outputs a signal Rs ′ obtained by adding the signal (a × Rs) output from the multiplier 162 and the signal (b × Rb) output from the multiplier 164.

  The multiplier 165 multiplies the input signal C by 0.707 (−3 dB) and outputs the result. The multiplier 166 multiplies the signal Ls ′ output from the adder 171 by 0.707 (−3 dB) and outputs the result. The multiplier 167 multiplies the signal Rs ′ output from the adder 172 by 0.707 (−3 dB) and outputs the result. The adder 173 adds the signal L and the signal (0.707 × C) output from the multiplier 165. The adder 174 adds the signal R and the signal (0.707 × C) output from the multiplier 165. The adder 175 adds the signal (L + 0.707 × C) output from the adder 173 and the signal (0.707 × Ls ′) output from the multiplier 166. The adder 176 adds the signal (R + 0.707 × C) output from the adder 174 and the signal (0.707 × Rs ′) output from the multiplier 167.

The multiplier 168 outputs a signal L ₀ obtained by multiplying the signal output from the adder 175 by 0.707 (−3 dB). The multiplier 169 outputs a signal R ₀ obtained by multiplying the signal output from the adder 176 by 0.707 (−3 dB). The signals L ₀ and R ₀ thus generated are as follows.

L ₀ = 0.707 × ((L + 0.707 × C) + (0.707 × Ls ′))
= 0.707 x ((L + 0.707 x C)
+ (0.707 × (a × Ls + b × Lb))) (1)
R ₀ = 0.707 × ((R + 0.707 × C) + (0.707 × Rs ′))
= 0.707 x ((R + 0.707 x C)
+ (0.707 × (a × Rs + b × Rb))) (2)
By the way, different values are used for the multipliers a and b of the multipliers 161 to 164 for 5.1ch and 7.1ch. Specifically, in the case of 5.1ch, a = 1.0 and b = 0. In the case of 7.1 ch, a = 0.5 and b = 0.5. For 7.1ch, a = 0.5 and b = 0.5 are assumed on the assumption that the signals Ls and Rs and the signals Lb and Rb are combined equally. In some cases, the ratio may be varied. In this case, the values of a and b may be changed.

Substituting these values into equations (1) and (2),
In the case of 5 and 1ch, it is as follows.

L ₀ = 0.707 × (L + 0.707 × C + 0.707 × Ls) (3)
R ₀ = 0.707 × (R + 0.707 × C + 0.707 × Rs) (4)
In the case of 7.1ch, it is as follows.

L ₀ = 0.707 × (L + 0.707 × C + 0.707 × (Ls + Lb) / 2)
... (5)
R ₀ = 0.707 × (R + 0.707 × C + 0.707 × (Rs + Rb) / 2)
... (6)
In the case of 2ch, only the signal L and the signal R input from the decode processing unit 14 are output, and the other components are not added, and each of these signals L and R is multiplied by a multiplier 168 or a multiplier. The signal is multiplied by 0.707 (−3 dB) at 169 and output as the signal L ₀ or the signal R ₀ .

The decorrelation processing unit 20 performs decorrelation processing using a set of signals L ₀ and R ₀ output from the downmix processing unit 16. FIG. 5 is a diagram illustrating a detailed configuration of the decorrelation processing unit 20. As illustrated in FIG. 5, the decorrelation processing unit 20 includes a surround L generation unit 210, a surround R generation unit 220, and a C generation unit 230.

The surround L generation unit 210 generates a surround L signal (signal SL) using the signals L ₀ and R ₀ output from the downmix processing unit 16. For this purpose, the surround L generation unit 210 includes an FIR filter 211, an adaptive filter (ADF) 212, an adder 213, and an LMS algorithm processing unit 214. The FIR filter 211 is used as a delay circuit (delay unit), and delays the input signal L ₀ by a time corresponding to the number of taps and outputs the delayed signal. The adaptive filter 212 has the same configuration as the FIR filter, and multiplies the input signal R ₀ by a predetermined tap coefficient W and outputs the result. The adder 213 is addition means, and subtracts the signal output from the adaptive filter 212 from the signal L ₀ output from the FIR filter 211 to output an error signal e _L. The LMS algorithm processing unit 214 is LMS algorithm processing means, and varies the filter coefficient of the adaptive filter 212 so that the power of the error signal e _L output from the adder 213 is minimized using the LMS algorithm. The error signal e _L output from the adder 213 is taken out as it is as the signal SL.

  FIG. 6 is a diagram illustrating a detailed configuration of the adaptive filter 212. As shown in FIG. 6, the adaptive filter 212 includes a plurality of delay elements 215, multipliers 216 that multiply the signals held in the delay elements 215 by variable filter coefficients, and the multipliers 216. And an adder 217 for adding the outputs. The values of the filter coefficients (multipliers) of the plurality of multipliers 216 are updated by the LMS algorithm processing unit 214.

The LMS algorithm processing unit 214 updates the value of the filter coefficient of the adaptive filter 212 so that the power of the error signal e _L output from the adder 213 is minimized, and the adaptive filter 212 receives the input signal R _0. Among these components, the value of the filter coefficient is updated so as to extract a component having a high correlation with the signal L ₀ . That is, the signal R ₀ and the error signal e _L output from the adder 213 are input to the LMS algorithm processing unit 214, and the signal R ₀ and the error signal e _L are processed by the LMS algorithm. The LMS algorithm processing unit 214 outputs a filter coefficient update command to each multiplier 216 in the adaptive filter 212, and the value of the filter coefficient superimposed on the signal held in each delay element 215 is changed.

Thus, the adaptive filter 212 extracts a component having a high correlation with the LS signal in the signal R ₀ , and this component is subtracted from the signal L ₀ by the adder 213. Therefore, the error signal e _L output from the adder 213 includes only a component that is not highly correlated with the signal R ₀ in the signal L ₀ , and this is used as the surround L signal.

  Incidentally, the LMS algorithm is an algorithm using an instantaneous square error as an evaluation amount, and the LMS algorithm processing unit 214 updates the value of the filter coefficient W according to the following equation.

W (n + 1) = W (n) +2 μ · e _L (n) · R ₀ (n) (7)
Here, μ is a step size parameter. When this value is set larger, the convergence of the filter coefficient W becomes faster. Conversely, when this value is set smaller, the convergence of the filter coefficient W becomes slower.

Similarly, the surround R generating unit 220 generates a surround R signal (signal SR) using the signals L ₀ and R ₀ output from the downmix processing unit 16. For this purpose, the surround R generation unit 220 includes an FIR filter 221, an adaptive filter (ADF) 222, an adder 223, and an LMS algorithm processing unit 224. The FIR filter 221 is used as a delay circuit (delay means), and delays the input signal R ₀ by a time corresponding to the number of taps and outputs it. The adaptive filter 222 has the same configuration as the FIR filter, and multiplies the input signal L ₀ by a predetermined tap coefficient W and outputs the result. The adder 223 is addition means, and subtracts the signal output from the adaptive filter 222 from the signal R ₀ output from the FIR filter 221 to output an error signal e _R. The LMS algorithm processing unit 224 is LMS algorithm processing means, and varies the filter coefficient of the adaptive filter 222 so that the power of the error signal e _R output from the adder 223 is minimized using the LMS algorithm. Further, the error signal e _R output from the adder 223 is extracted as a signal SR as it is.

The LMS algorithm processing unit 224 updates the value of the filter coefficient of the adaptive filter 222 so that the power of the error signal e _R output from the adder 223 is minimized, and the adaptive filter 222 receives the input signal L _0. Among these components, the value of the filter coefficient is updated so as to extract a component having a high correlation with the signal R ₀ . That is, the LMS algorithm processing unit 224 receives the signal L ₀ and the error signal e _R output from the adder 223, and these signals L ₀ and the error signal e _R are processed by the LMS algorithm. As a result, a filter coefficient update command is output from the LMS algorithm processing unit 224 to each multiplier in the adaptive filter 222, and the value of the filter coefficient superimposed on the signal held in each delay element is changed.

Thus, the adaptive filter 222 extracts a component having a high correlation with the signal R ₀ in the signal L ₀ , and this component is subtracted from the signal L ₀ by the adder 223. Therefore, the error signal e _R output from the adder 223 includes only a component that is not highly correlated with the signal L ₀ in the signal R ₀ , and this is used as the surround R signal.

  The LMS algorithm is an algorithm using the instantaneous square error as an evaluation amount, and the LMS algorithm processing unit 224 updates the value of the filter coefficient W according to the following equation.

W (n + 1) = W (n) +2 μ · e _R (n) · L ₀ (n) (8)
Here, μ is a step size parameter. When this value is set larger, the convergence of the filter coefficient W becomes faster. Conversely, when this value is set smaller, the convergence of the filter coefficient W becomes slower.

The C generation unit 230 generates a center signal (signal C ₀ ) using each output signal of the adaptive filter 212 in the surround L signal generation unit 210 and the adaptive filter 222 in the surround R signal generation unit 220. For this purpose, the C generation unit 230 includes multipliers 231 and 232 and an adder 233. The multiplier of the multiplier 231 is set to 0.5, and the signal output from the adaptive filter 212 in the surround L generation unit 210 is multiplied by 0.5. The multiplier of the multiplier 232 is set to 0.5, and the signal output from the adaptive filter 222 in the surround R generator 220 is multiplied by 0.5. The adder 233 adds the output signals of the two multipliers 231 and 232 and outputs the result. This output signal is taken as the center signal C _0.

FIG. 7 shows five types of signals C ₀ , L ₀ , R ₀ , SL, RL and 5.1 ch or 7.1 ch audio format signals output from the decorrelation processing unit 20 (from the decoding processing unit 14). It is a figure which shows the relationship with the 5 types or 7 types of signal which are output and input into the downmix process part 16. FIG.

The mix processing unit 30 performs mixing processing after weighting each signal output from the decorrelation processing unit 20, and combines the four signals L-out, R-out, Ls-out, and Rs. -out is generated. The signals Ls-out and Rs-out correspond to the composite surround signal, the signal L-out corresponds to the left signal, and the signal R-out corresponds to the right signal. FIG. 8 is a diagram illustrating a detailed configuration of the mix processing unit 30. As shown in FIG. 8, the mix processing unit 30 includes seven multipliers 31, 32, 33, 34, 37, 38, 39 and four adders 35, 36, 41, 42. The multiplier 31, the multiplier (weighting coefficient) is set to g1, the signal L _0, which is outputted from the surround L generation unit 210 of the decorrelation processing section 20 g1 times and output. The multiplier 32 has a multiplier (weighting coefficient) set to g1, and multiplies the signal _R0 output from the surround R generation unit 220 in the decorrelation processing unit 20 by g1 and outputs the signal. The multiplier 33 has a multiplier (weighting coefficient) set to g2, and multiplies the signal SL output from the surround L generation unit 210 in the decorrelation processing unit 20 by g2 and outputs the signal SL. The multiplier 34 has a multiplier (weighting coefficient) set to g2, and multiplies the signal SR output from the surround R generation unit 220 in the decorrelation processing unit 20 by g2 and outputs it. The adder 35 outputs a signal Ls-out (= g1 × L ₀ + g2 × SL) obtained by adding the output signals of the multipliers 31 and 33. The adder 36 outputs a signal Rs-out (= g1 × R ₀ + g2 × SR) obtained by adding the output signals of the multipliers 32 and 34.

Also, the multiplier 37, the multiplier (weighting coefficient) is set to g3, the signal L _0, which is outputted from the surround L generation unit 210 of the decorrelation processing section 20 g3 times and output. The multiplier 38 has a multiplier (weighting coefficient) set to g3, and multiplies the signal _R0 output from the surround R generation unit 220 in the decorrelation processing unit 20 by g3 and outputs it. The multiplier 39, the multiplier (weighting coefficient) is set to g4, the signal C ₀ which is output from the C generation unit 230 of the decorrelation processing section 20 outputs g4 multiplied by. The adder 41 outputs a signal L-out (= g3 × L ₀ + g4 × C ₀ ) obtained by adding the output signals of the multipliers 37 and 39. The adder 42 outputs a signal R-out (= g3 × R ₀ + g4 × C ₀ ) obtained by adding the output signals of the multipliers 38 and 39. The multipliers g1, g2, g3, and g4 described above are variably set according to the format of the input audio data.

  The EQ processing unit 40 performs equalizing processing as sound quality control processing on the four types of signals L-out, R-out, Ls-out, and Rs-out output from the mix processing unit 30. Thereby, for example, bass emphasis is performed. The decorrelation processing unit 20, the mix processing unit 30, and the EQ processing unit 40 of the present embodiment can be easily realized by a DSP (digital signal processor).

  The control unit 50 controls the operation of the entire audio apparatus. In addition, when an identification signal indicating an audio format such as 5.1ch, 7.1ch, or 2ch is input to the control unit 50, according to the audio format, the surround L generation unit 210 in the decorrelation processing unit 20 and An instruction is sent to the surround R generation unit 220, the value of the step size parameter μ for updating the value of the filter coefficient W is changed by the LMS algorithm processing units 214 and 224, and an instruction is given to the mix processing unit 30 To change the values of the multipliers g1, g2, g3, g4 of the multipliers 31-34, 37-39.

  The digital-analog converter 60 converts the signals L-out, R-out, Ls-out, and Rs-out after being equalized by the EQ processing unit 40 into analog signals. These signals are amplified by the amplifier 62 and then output from the four speakers 70, 72, 74, 76. The speaker 70 is installed on the front left side in the passenger compartment and outputs an audio sound corresponding to the signal L-out. The speaker 72 is installed on the front right side in the passenger compartment and outputs an audio sound corresponding to the signal R-out. The speaker 74 is installed on the rear left side in the passenger compartment and outputs an audio sound corresponding to the signal Ls-out. The speaker 76 is installed on the right rear side in the vehicle interior and outputs an audio sound corresponding to the signal Rs-out.

  The downmix processing unit 16 described above corresponds to the downmix processing unit, the decorrelation processing unit 20 corresponds to the surround signal generation unit, the mix processing unit 30 corresponds to the synthesis unit, and the control unit 50 corresponds to the control unit.

The audio apparatus according to the present embodiment has such a configuration. Next, a setting operation corresponding to each format of 5.1ch and 7.1ch will be described. As shown in FIG. 7, focusing on the signals SL and SR generated by the decorrelation processing unit 20, the 7.1-channel format signal L ₀ , R ₀ , R ₀ , as compared to the 5.1-format signal L ₀ and R ₀ , Since R ₀ contains more back components (signals Lb and Rb), it can be seen that the correlation between the signals SL and SR becomes lower. In the case of the 7.1ch format, since the expression of sound is more diversified, the signal Ls and the signal Lb (or the signals Rs and Rb) are not always output at the same level, and the 7.1ch format is used. In this case, it is conceivable that the output of the signal SL or the signal SR is smaller than that in the 5.1ch format.

In addition, regarding the C component (signal C ₀ ), the level of the C component tends to be lower at the time of the downmix process than the non-downmix stereo sound source. In addition, when the speaker arrangement (reproduction sound field) shown in FIG. 3 is considered, the signal level per channel becomes smaller as the number of channels is larger when the sum of sounds is taken as one reference. Considering these, in the present embodiment, a large proportion (multiplier g4) combining signals C ₀ to the signal L ₀ and the signal R _0, in particular to set the value of g4 0.5 or more. In addition, since the tendency of the C component sound to be smaller than the other components is larger in the case of 7.1 ch than in the case of 5.1 ch, the value of the multiplier g4 in the case of 7.1 ch is set to 5.1 ch. You may make it enlarge compared with the case of.

Considering the above points, in order to clarify the difference between the 5.1 format and the 7.1cn format, the following settings are performed in the present embodiment.
(1) In the mix process after the decorrelation process, the level (weighting coefficient g2) for mixing the signals SL and RS is increased in the 7.1ch format as compared with the 5.1 format.
(2) The level at which the signal C ₀ is mixed (weighting coefficient g4) is increased as compared with the 2ch stereo sound source.
(3) In order to increase the adaptation speed of the decorrelation processing, the value of the step size parameter μ is increased in the 7.1ch format as compared with the 5.1 format.

  For example, in the 5.1ch format, each value is set as follows.

μ = 0.0003
g1 = 0.5 (−6 dB)
g2 = 0.5 (-6 dB)
g3 = 0.707 (-3 dB)
g4 = 0.5 to 0.707
On the other hand, in the 7.1ch format, each value is set as follows.

μ = 0.0005
g1 = 0.5 (−6 dB)
g2 = 0.707 (-3 dB)
g3 = 0.707 (-3 dB)
g4 = 0.707 (-3 dB)
For comparison, each value in the 2ch format is shown as follows.

μ = 0.0003
g1 = 0.5 (−6 dB)
g2 = 0.5 (-6 dB)
g3 = 1.0 (0 dB)
g4 = 0.0
Each value mentioned above is an example, and about g1-g4, a value can be changed in the range of 0 or more and 1 or less. Further, μ is a very small value of 1 or less and can be changed.

  As described above, in the audio apparatus according to the present embodiment, when performing audio reproduction in 5.1ch format or 7.1ch format, decorrelation is performed after downmix processing is performed on the original signal corresponding to each format. Processing is performed to generate the signals Ls-out and Rs-out to be input to the rear speakers 74 and 76, so that the localization bias can be eliminated and it is possible to prevent the reproduced sound from feeling uncomfortable. . In particular, by increasing the level (weighting coefficient g2) for mixing the signals SL and SR when performing the mixing process in the 7.1ch format as compared with the 5.1 format, the back characteristic peculiar to the 7.1ch format is achieved. The localization bias due to the components (signals Lb and Rb) can be adjusted.

  Further, by making the value of the step size parameter μ when performing the decorrelation processing after performing the downmix processing on the original signal according to each format between the 5.1ch format and the 7.1ch format, Differences in audio playback according to the number of channels can be expressed. In particular, by setting the step size parameter μ when performing 7.1ch format audio playback to a larger value, the adaptation speed becomes faster and the time for reflecting many channel component signals in the playback sound is shortened. It is possible to cope with diversified sound expressions.

Further, by setting the level for mixing the signal C ₀ (value of the weighting coefficient g4) to 0.5 or more, it is possible to prevent the sound image corresponding to the center speaker from becoming small in the case of 5.1ch or 7.1ch. be able to. Further, when the level for mixing the signal C ₀ (value of the weighting coefficient g4) is changed to 7.1 channel having a larger number of channels by increasing 7.1 channel than 5.1 channel, the center is changed. It is possible to prevent the sound image corresponding to the speaker from becoming small.

  In addition, this invention is not limited to the said embodiment, A various deformation | transformation implementation is possible within the range of the summary of this invention. In the above-described embodiment, the 5.1ch and 7.1ch audio formats have been mainly described. However, the present invention can also be applied to other audio formats.

  As described above, according to the present invention, by performing a downmix process on an audio signal corresponding to each audio format and then performing a decorrelation process to generate a set of left and right surround signals, It is possible to eliminate the localization bias when audio reproduction is performed in the vehicle interior using fewer speakers, and it is possible to prevent the playback sound from feeling uncomfortable.

DESCRIPTION OF SYMBOLS 10 Disc reader 20 Decorrelation process part 30 Mix process part 40 EQ process part 50 Control part 12 Disk read part 14 Decode process part 16 Down mix process part 210 Surround L production | generation part 220 Surround R production | generation part 230 C production | generation part 211, 221 FIR filter 212, 222 Adaptive filter (ADF)

Claims (8)

An audio device that performs audio playback using a smaller number of speakers than the number of channels,
Downmix processing means for performing downmix processing for synthesizing the signals of the respective channels with respect to an audio signal corresponding to a predetermined audio format to generate a set of left and right intermediate audio signals;
A pair of low-correlated left and right surround signals and a highly correlated center signal are generated by performing decorrelation processing using an adaptive algorithm with a predetermined step size parameter for each of the pair of intermediate audio signals. Surround signal generating means for
The intermediate audio signal and the surround signal are weighted and then combined to generate a combined surround signal separately, and the intermediate audio signal and the center signal are weighted and combined. And a combining means for separately performing left and right processing for generating the left signal and the right signal,
Corresponding to each of a plurality of audio formats, a control means for setting a value of the step size parameter and a value of a coefficient used for weighting by the synthesizing means,
An audio apparatus comprising: a pair of left and right synthesized surround signals synthesized by the synthesizing means, and a left signal and a right signal output from separate speakers. In claim 1,
The surround signal generating means performs a process of extracting a component having a high correlation with the other signal in one signal by using the pair of left and right intermediate audio signals and subtracting the component from the other signal. By changing the signals, the left and right sets of the surround signal are generated,
The surround signal generating means extracts the highly correlated component by updating the filter coefficient of the adaptive filter using the adaptive algorithm, and generates the center signal.
The audio apparatus according to claim 1, wherein the control means sets a value of the step size parameter used when the filter coefficient is updated. In claim 1 or 2,
The audio apparatus according to claim 1, wherein the plurality of audio formats have different numbers of channels. In claim 3,
The audio apparatus according to claim 5, wherein the plurality of audio formats include 5.1 channel and 7.1 channel audio formats. In claim 4,
The audio apparatus according to claim 1, wherein the step size parameter has a larger value corresponding to 7.1 channel audio data than a value corresponding to 5.1 channel audio data. In claim 4 or 5,
The weighting corresponding to the surround signal performed to generate the synthesized surround signal is greater in the value corresponding to 7.1 channel audio data than the value corresponding to 5.1 channel audio data. Features audio equipment. In any one of Claims 4-6,
An audio apparatus, wherein a weighting value corresponding to the center signal performed to generate the center signal is set to 0.5 or more. In any one of Claims 4-7,
The weighting corresponding to the center signal performed to generate the center signal has a value corresponding to 7.1 channel audio data larger than a value corresponding to 5.1 channel audio data. An audio device.

Images