JP2017212552A - Channel number converter and program thereof - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a channel number converter performing appropriate channel number conversion of each program along the content thereof.SOLUTION: A channel number converter 10 includes a weighting coefficient calculation unit 13 for receiving a multichannel sound signal and a reference signal corresponding to the content thereof, and calculating the weighting coefficient of each channel signal of the multichannel sound signal contained in the reference signal, a correction coefficient calculation unit 14 for calculating correction coefficients for correcting each channel signal of the multichannel sound signal based on the weighting coefficients, a correction coefficient application unit 15 for applying the correction coefficients to the multichannel sound signal, and a channel number conversion unit 17 for converting the multichannel sound signal, to which the correction coefficients are applied, into a reproduction channel signal of desired channel number, by a predetermined channel number conversion method.SELECTED DRAWING: Figure 1

Description

  The present invention relates to a channel number conversion device that converts the number of channels of a multi-channel audio signal and a program thereof.

  Currently, the practical use of multi-channel audio broadcasting (Non-Patent Document 1) such as 22.2ch is being promoted. With multi-channel audio broadcasting, it is possible to realize audio reproduction with high presence. However, in the case of a general home, for example, it is assumed that there are many cases where it is possible to reproduce only the number of channels smaller than 22.2 ch, such as 2 ch stereo. For this reason, in order to reproduce multi-channel audio broadcasting at home, it is necessary to convert the number of channels of audio signals in accordance with the number of channels that can be reproduced at home. In such a case, a technique is generally known in which downmixing or rendering is performed to convert the number of channels according to the reproduction environment (Non-Patent Documents 1 and 2).

  However, in a general channel number conversion method in which the number of channels is uniquely converted regardless of the contents of the program, there is a possibility that the audio signal after the channel number conversion does not conform to the intention of the program producer. On the other hand, when multi-channel audio broadcasting is performed, implementation of so-called simulcast where a program producer separately produces a program for 2ch stereo and simultaneously broadcasts multi-channel and 2ch stereo is being studied.

“Video Coding, Audio Coding and Multiplexing SPECIFICATIONS FOR DIGITAL BROADCASTING ARIB STANDARD ARIB STD-B32 Version 3.6” in Digital Broadcasting, March 25, 2016 Japan Radio Industry Association ISO / IEC 23008-3, “Information technology High efficiency coding and media delivery in heterogeneous environments”

  By the way, the number of speakers and their arrangement vary depending on the home environment, and there is a need to listen to broadcasting on the number of channels corresponding to the playback environment in each home. However, it is impossible to carry out simultaneous broadcasting corresponding to any number of channels and speaker arrangements. Further, when the number of channels is converted by a uniform channel number conversion method corresponding to the reproduction environment (number of speakers, speaker arrangement) in each home, it is difficult to convert for each program as intended by the program producer. For this reason, even if simulcasting is carried out, there is a high possibility that each home will not be able to reproduce the sound with the appropriate number of channels converted for each program.

  SUMMARY OF THE INVENTION Accordingly, an object of the present invention is to provide a channel number conversion device and a program thereof that can solve the above-described problems.

  According to an aspect of the present invention, the channel number conversion device receives a multi-channel audio signal and a reference signal corresponding to the multi-channel audio signal, and the multi-channel audio signal included in each channel of the reference signal. A weighting coefficient calculator for calculating a weighting coefficient corresponding to each channel signal, a correction coefficient calculator for calculating a correction coefficient to be multiplied to each channel signal of the multi-channel audio signal based on the weighting coefficient, and the correction coefficient A correction coefficient applying unit to be applied to the multi-channel audio signal; a channel number converting unit for converting the multi-channel audio signal to which the correction coefficient is applied into a reproduction channel signal having a desired number of channels by a predetermined channel number conversion method; .

  According to an aspect of the present invention, the weighting coefficient calculation unit may include a delay correction unit that corrects a delay between the reference signal and each channel signal of the multi-channel audio signal.

  According to an aspect of the present invention, the weighting coefficient calculator receives the multichannel audio signal and the reference signal, and analyzes a weighting ratio for each channel signal of the multichannel audio signal; May be provided.

  According to an aspect of the present invention, the weighting coefficient calculator calculates the weighting ratio for the energy of each channel signal of the reference signal and each channel signal of the multichannel audio signal corresponding to each channel of the reference signal. You may provide the weighting ratio correction | amendment part which correct | amends the said weighting ratio so that the sum of the energy of the multiplied signal may become equal.

  According to an aspect of the present invention, the weighting coefficient calculation unit groups each channel signal of the multi-channel audio signal based on the similarity of each channel signal, and based on the channel signal belonging to the group A grouping unit that generates a group signal representing a group may be further included, and the weighting coefficient calculation unit may calculate a weighting coefficient for the group signal.

  According to an aspect of the present invention, the grouping unit includes, based on the channel signals belonging to the group, an average of the channel signals, a channel signal at the centroid of the similarity of the channel signals, and the channel signals Any one of the channel signals having the maximum energy may be generated as the group signal.

  According to an aspect of the present invention, the weighting coefficient calculation unit includes one or more channel signals based on a cross-correlation coefficient between the channel signals of the multi-channel audio signal and the reference signal. A reference channel signal selection unit that selects the weighting factor, and the weighting factor calculation unit performs weighting with a constraint that the weighting factor of the selected channel signal is larger than the weighting factors of the other channel signals. A coefficient may be calculated.

  According to an aspect of the present invention, the weighting coefficient calculation unit includes one or more channel signals based on a cross-correlation coefficient between the channel signals of the multi-channel audio signal and the reference signal. A reference channel signal selection unit that selects the reference channel signal, and the weighting coefficient calculation unit may calculate the weighting coefficient only for the channel signal selected by the reference channel signal selection unit.

  According to an aspect of the present invention, the correction coefficient calculation unit is configured to calculate a signal obtained by applying the weighting coefficient to each channel signal of the multi-channel audio signal corresponding to the total energy of the reference signal or each channel of the reference signal. The correction coefficient may be calculated so that the sum of energy is equal to the sum of energy of signals obtained by applying the correction coefficient to each channel signal of the multi-channel audio signal.

  According to an aspect of the present invention, the correction coefficient application unit includes a predetermined initial value of a correction coefficient corresponding to the number of channels of the multi-channel audio signal, and a correction coefficient calculated by the correction coefficient calculation unit, Based on the correction coefficient corresponding to the number of channels of the reference signal, the correction coefficient calculated by the correction coefficient calculation unit according to the number of channels of the reproduction channel signal may be corrected by an interpolation method.

  According to an aspect of the present invention, the channel number conversion device may further include a mono signal conversion unit that converts the reference signal into a mono signal by a predetermined channel number conversion method.

  According to one aspect of the present invention, there is provided a program for causing a computer to function as the channel number conversion device described in any one of the above.

  According to the channel number conversion apparatus of the present invention, when a multi-channel audio signal and an audio signal having a smaller number of channels are simultaneously provided, conversion of the number of channels according to the production intention can be realized for an arbitrary number of channels. Can do.

It is a block diagram which shows an example of the channel number converter in 1st embodiment which concerns on this invention. It is a block diagram which shows an example of the weighting coefficient calculation part in 1st embodiment which concerns on this invention. It is a flowchart which shows an example of the channel number conversion process in 1st embodiment which concerns on this invention. It is a block diagram which shows an example of the channel number converter in 2nd embodiment which concerns on this invention. It is a block diagram which shows an example of the weighting coefficient calculation part in 2nd embodiment which concerns on this invention. It is a flowchart which shows an example of the channel number conversion process in 2nd embodiment which concerns on this invention. It is a block diagram which shows an example of the channel number converter in 3rd embodiment which concerns on this invention. It is a block diagram which shows an example of the weighting coefficient calculation part in 3rd embodiment which concerns on this invention. It is a flowchart which shows an example of the channel number conversion process in 3rd embodiment which concerns on this invention. It is a block diagram which shows an example of the channel number converter in 4th embodiment which concerns on this invention. It is a flowchart which shows an example of the channel number conversion process in 4th embodiment which concerns on this invention.

<First embodiment>
Hereinafter, a channel number conversion apparatus according to a first embodiment of the present invention will be described with reference to FIGS.
FIG. 1 is a block diagram showing an example of a channel number conversion apparatus according to the first embodiment of the present invention. As shown in FIG. 1, the channel number conversion apparatus 10 includes a reference signal input unit 11, a multi-channel audio signal input unit 12, a weighting coefficient calculation unit 13, a correction coefficient calculation unit 14, a correction coefficient application unit 15, and a reproduction channel information acquisition unit. 16, a channel number conversion unit 17, and a storage unit 18.
FIG. 1 shows a state where a 22.2 ch multi-channel audio signal and a 2 ch reference signal are input to the channel number converter 10 and a 5.1 ch playback channel audio signal after the channel number conversion is output.
The channel number converter 10 refers to a predetermined multi-channel audio signal (n channel) with an audio signal (l channel) having a channel number smaller than the number of channels of the multi-channel audio signal (hereinafter referred to as a reference signal). Thus, it is a device that converts it into a reproduced audio signal (m channels) of a desired number of channels. In the following, description will be given by taking as an example the case where the predetermined multi-channel audio signal is a 22.2ch acoustic system for 8K SHV, the reference signal is 2ch stereo, and the reproduced audio signal is 5.1ch. However, the number of channels of the multi-channel audio signal, reference signal, and reproduced audio signal is not limited to the number of channels in this example. The channel number conversion device 10 is configured by a computer, and may be incorporated in a broadcast receiver such as a television or a media playback device such as a home theater, for example.

Hereinafter, the channel number conversion apparatus 10 will be described in detail.
The reference signal input unit 11 inputs a reference signal. For example, the reference signal input unit 11 inputs a 2ch stereo audio signal among 22.2 ch multichannel audio signals and 2ch stereo audio signals broadcast by simulcast.
The multichannel audio signal input unit 12 inputs a multichannel audio signal. For example, a 22.2 ch audio signal broadcast by simulcast is input.
A dedicated engineer may create audio signals corresponding to the number of channels broadcast by simulcast. For example, for 22.2ch audio signals, a 22.2ch dedicated engineer is suitable for the broadcast content depending on the playback environment such as the speaker layout corresponding to 22.2ch (the intention of the program creator was reflected). ) Create an audio signal for each channel so that a three-dimensional sound is reproduced. On the other hand, for 2ch audio signals, a 2ch dedicated engineer creates audio signals for each of the two channels so that audio suitable for the broadcast content is reproduced in a 2ch playback environment. Each of the 22.2ch and 2ch audio signals created at this time is created so as to suit the contents to be expressed by the program. For example, a documentary program in which narration is dominant and a music program have different characteristics of an audio signal created for each channel even if they are the same 2ch. Therefore, the relationship between the 22.2ch audio signal and the 2ch audio signal is not always a relationship that can be converted from the 22.2ch audio signal to the 2ch audio signal by a predetermined (one type) downmix coefficient. Absent.

The weighting coefficient calculator 13 calculates a weighting coefficient for each channel signal of the multichannel audio signal included in the signal of each channel of the reference signal. For example, reference signals (L, R), multi-channel audio signals (FC, FLc, FRc, FL, FR, SiL, SiR, BL, BR, BC, LFE1, LFE2, TpFC, TpFL, TpFR, TpSiL, TpSiR) , TpC, TpBL, TpBR, TpBC, BtFC, BtFL, BtFR), the weighting coefficient calculation unit 13 sets L to (a _1,1 × FC, a _2,1 × FLc,..., A _{24, 1} × BtFR), R a _{(a 1,2 × FC, a 2,2} × FLc, when expressed _{····, a 24,2} × BtFR) and, the coefficients _{(a 1,1} ~a _{1, 24,} to calculate the value of _{a 2,1} ~a _2,24). The same applies to configurations shown in weighting coefficient calculators 13a and 13b described later.

The correction coefficient calculation unit 14 uses the weighting coefficient calculated by the weighting coefficient calculation unit 13 to calculate a correction coefficient to be multiplied to each channel signal of the multichannel audio signal. More specifically, the correction coefficient calculation unit 14 uses the energy or loudness of a signal generated by applying a weighting coefficient to each channel signal of the reference signal and the multichannel audio signal as an index, and outputs each channel signal of the multichannel audio signal. Calculate the correction factor applied to.
Alternatively, the correction coefficient calculation unit 14 calculates a correction coefficient to be applied to each channel signal of the multichannel audio signal based on the square sum of the weighting coefficients of each channel signal of the multichannel audio signal with respect to each channel signal of the reference signal. .

The correction coefficient application unit 15 applies the correction coefficient calculated by the correction coefficient calculation unit 14 by any of the methods described above to the multichannel audio signal. At this time, the correction coefficient application unit 15 corrects the correction coefficient according to the number of channels of the audio signal to be reproduced, and applies the corrected correction coefficient to the multi-channel audio signal.
The playback channel information acquisition unit 16 acquires, for example, information on the number of channels of the playback channel audio signal (number of playback channels) as information on the playback channel audio signal.

The channel number conversion unit 17 receives the multi-channel audio signal after the correction coefficient application unit 15 applies the correction coefficient, and converts the multi-channel audio signal according to the number of reproduction channels by, for example, channel number conversion processing described later. Convert the number of channels.
The storage unit 18 stores various data necessary for the channel number conversion process.

Next, the weighting coefficient calculator 13 will be described in detail with reference to FIG.
FIG. 2 is a block diagram showing an example of the weighting coefficient calculator in the first embodiment according to the present invention.
As shown in FIG. 2, the weighting coefficient calculation unit 13 includes a delay correction unit 131, a weighting ratio analysis unit 132, and a weighting ratio correction unit 133.
The delay correction unit 131 corrects the delay of each channel signal of the multi-channel audio signal with respect to each channel signal of the reference signal. The time lag between the two signals may be advanced or delayed, and various cases are considered. These are collectively referred to as a delay. The delay correction unit 131 determines the delay of the signal corresponding to each channel signal of the multichannel audio signal included in each channel of the reference signal with respect to the channel signal constituting the multichannel audio signal, for example, by a cross-correlation function. calculate. This is because, in the reference signal, for example, in order to express a three-dimensional sound expressed by a multi-channel audio signal, the sound related to a certain channel signal included in the multi-channel audio signal is shifted in the time axis direction. This is because there are cases in which the sound is output from the rear in the multi-channel audio signal and the sound output from the front in the reference signal, for example, so that the sound from the rear is slightly delayed in the reference signal. . The delay correction unit 131 corrects the delay of each channel signal of the multi-channel audio signal by the calculated delay amount, and corrects the phase of the signal corresponding to the multi-channel audio signal included in each channel signal of the reference signal, and the multi-channel audio. Align the phase of the signal with the channel signal.

The weighting ratio analysis unit 132 analyzes the weighting ratio of the multichannel audio signal with respect to each channel signal by using any multivariate analysis method such as multiple regression analysis or canonical correlation analysis. Alternatively, the weighting ratio analysis unit 132 may analyze the weighting ratio by machine learning such as a genetic algorithm or deep learning.
The weighting ratio correction unit 133 corrects the weighting ratio analyzed by the weighting ratio analysis unit 132 based on the energy of each channel signal constituting the reference signal, and outputs the corrected value as a weighting coefficient. Specifically, the weighting ratio correction unit 133 multiplies each channel of the multichannel audio signal by the energy of each channel signal of the reference signal and the weighting ratio analyzed by the weighting ratio analysis unit 132, and obtains the pseudo reference signal obtained by multiplication. The weighting ratio is corrected so that the energy of each channel signal becomes equal.

Next, details of the channel number conversion process will be described with reference to FIG.
FIG. 3 is a flowchart showing an example of the channel number conversion process in the first embodiment according to the present invention.
As a premise, for a certain program, a multi-channel audio signal and a reference signal created according to the number of channels are broadcast simultaneously, and the channel number conversion device 10 inputs both signals.
First, in step S11, the multichannel audio signal input unit 12 inputs a multichannel audio signal. The multichannel audio signal input unit 12 outputs the multichannel audio signal to the weighting coefficient calculation unit 13. In parallel with step S11, in step S12, the reference signal input unit 11 inputs a reference signal. The reference signal input unit 11 separates the input 2ch stereo reference signal into LR channel signals (p _j , 1 ≦ j ≦ 2, l = 2). Here, p _j is each channel signal of the reference signal after separation. The reference signal input unit 11 outputs each channel signal of the reference signal to the weighting coefficient calculation unit 13.

  Next, in step S13, in the weighting coefficient calculation unit 13, the delay correction unit 131 inputs the reference signal and the multichannel audio signal, and corrects the delay of the multichannel audio signal. For example, the delay correction unit 131 calculates a delay corresponding to each channel signal of the multichannel audio signal for each channel signal of the reference signal, for example, by calculating a cross correlation function, and corrects each channel of the multichannel audio signal. To do. The delay correction unit 131 outputs the multichannel audio signal after delay correction to the weighting ratio analysis unit 132. The reason for correcting the delay is to calculate the weighting ratio more accurately.

Next, in step S14, the weighting ratio analysis unit 132 inputs the multichannel audio signal after delay correction and the reference signal, and calculates the weighting ratio of each channel signal of the multichannel audio signal to each channel signal of the reference signal. Calculate using multiple regression analysis. Specifically, the weighting ratio analysis unit 132 is configured to appropriately configure each channel signal of 2ch (reference signal), and 22.2ch (multichannel audio signal) audio signal (q _i , 1 ≦ i ≦ 24). , N = 24), the weighting ratio between channels (a _ij , 1 ≦ i ≦ 24, 1 ≦ j ≦ 2) is calculated. Here, the channel signal p ^ _j of the pseudo reference signal can be expressed by the following equation (1).

From equation (1), the weighting ratio “a _ij ” corresponds to the level ratio of each channel of the 22.2 ch audio signal included in each channel signal of 2 ch. The weighting ratio analysis unit 132 outputs the calculated weighting ratio to the weighting ratio correction unit 133.

Next, in step S15, the weighting ratio correction unit 133 inputs the weighting ratio and the reference signal analyzed by the weighting ratio analysis unit 132, and corrects the weighting ratio based on energy. For example, the Lch energy (E _Lch ) of the 2ch reference signal can be expressed by the following equation (2).

  Further, the Lch energy of the pseudo reference signal expressed using the weighting ratio can be expressed by the following equation (3).

The weighting ratio correction unit 133 calculates a constant c ₁ such that the Lch energy of the reference signal and the Lch energy of the pseudo reference signal are equal to each other using Expression (4).

Next, the weighting ratio correction unit 133 corrects the weighting ratio by multiplying each of the weighting ratios a _ij (1 ≦ i ≦ 24, 1 ≦ j ≦ 2) by the square root of the constant c _j . The corrected weighting ratio is a weighting coefficient. In step S <b> 16, the weighting coefficient calculation unit 13 outputs the weighting coefficient obtained by the correction by the weighting ratio correction unit 133 to the correction coefficient calculation unit 14.

Note that the weighting ratio correction unit 133 may use loudness, amplitude, or the like in addition to energy as an index serving as a correction reference. For example, the weighting ratio correction unit 133 may calculate a constant c _j such that the sum of the loudness of each channel signal of the reference signal is equal to the sum of the loudness of each channel signal of the pseudo reference signal.

Next, in step S17, the correction coefficient calculator 14 calculates a correction coefficient. Specifically, the correction coefficient calculation unit 14 corrects the correction coefficient b _i (1 ≦ i ≦ 24) for equalizing the sum of the energy of the entire 2ch stereo signal and the total energy of the 22.2ch audio signal to which the correction coefficient is applied. ) Is calculated by the following equation (5).

For example, the following formula (6) can be derived from the formula 5.
(B ₁ ) ² = c ₁ × (a ₁₁ ) ² + c ₂ × (a ₂₁ ) ² (6)
Other values of b _{2 to} b ₂₄ can be similarly derived from the equation (5).

Note that the correction coefficient calculation unit 14 may use loudness, amplitude, etc. in addition to energy as an index serving as a reference for calculating the correction coefficient. For example, the correction coefficient calculation unit 14, the total loudness of a multichannel audio signal according to the total loudness and the correction coefficient of the reference signal may calculate the correction coefficient b _i as equal. Alternatively, the correction coefficient calculator 14 corrects the correction coefficient b _i so that the sum of squares of the amplitudes of all the channel signals of the reference signal is equal to the sum of squares of the amplitudes of all the channel signals of the multichannel audio signal to which the correction coefficient is applied. May be calculated. The correction coefficient calculation unit 14 outputs the calculated correction coefficients b _{1 to} b ₂₄ to the correction coefficient application unit 15.

Next, in step S18, the correction coefficient application unit 15 corrects the correction coefficient using, for example, linear interpolation according to the number of reproduction channels. The correction coefficient b _i calculated through the process of steps S13 to S17 is obtained from the 22.2 ch multi-channel audio signal and the 2 ch reference signal, so to speak, a correction suitable for converting 22.2 ch into 2 ch. A coefficient (a correction coefficient corresponding to the number of channels of the reference signal). When a 22.2 ch multi-channel audio signal is downmixed to a playback channel audio signal (5.1 ch) using a predetermined channel number conversion method, for example, it is mechanically converted by a downmix coefficient defined by a standard or the like. Will be. For this reason, there is a possibility that the 5.1 channel playback channel audio signal after conversion is different from the audio signal intended as the audio expressed by the program producer using 5.1 channel. In the present embodiment, a correction signal b _i for approximating a 22.2 ch multi-channel audio signal to the reference signal is calculated using the reference signal reflecting the intention of the program producer with the number of channels smaller than 5.1 ch as teacher data. To do. However, if the correction coefficient obtained in the above step is applied to the 22.2 ch as it is, the multi-channel audio signal after application has a shift that is assumed to be converted to 2 ch. 15 performs a correction to alleviate the shift to 2ch. Then, the correction coefficient application unit 15, the correction coefficient b _i, is modified to become more correction coefficients suitable for 5.1ch (correction coefficient corresponding to the number of channels of the reproduction channel signal).

For example, the correction coefficient when converting 22.2 ch (n channel) to 22.2 ch (n channel) is “1.0”, and the correction coefficient when converting 22.2 ch (n channel) to 2 ch (l channel) When the coefficient is “b _i ”, when converting the number of channels of 5.1 ch (m channel) between the two, 22.2 ch (n channel) is converted to 5.1 ch (m channel). It is assumed that the correction coefficient b _i is an intermediate value between the two correction coefficients (for example, b _i = (b _i +1) / 2). Here, b _i ≦ b _i ≦ 1.0. Then, the following equation (6) can be derived from the relationship of linear interpolation.
(B ^ _i −b _i ) ÷ (1.0−b _i ) = (m−l) ÷ (n−l)
... (6)
By transforming this, the following equation (7) can be derived.
b _i = b _i + (1−b _i ) × (m−l) ÷ (n−l) (7)
In the above description of linear interpolation, it was assumed that l <m <n. However, Equation (7) can be applied to both m> l and m <l. There are no restrictions on the relationship. In addition, here, the case where the correction coefficient is corrected for 5.1ch using linear interpolation has been described as an example, but the correction coefficient may be corrected using another interpolation method (polynomial interpolation or the like). .
Next, in step S19, the correction coefficient application unit 15 applies the correction coefficient b _i suitable for 5.1ch to the 22.2ch multi-channel audio signal. Specifically, the correction coefficient application unit 15 multiplies b ^ _i by q _i . Then, the correction coefficient application unit 15 outputs b ^ _i × q _i (1 ≦ i ≦ 24) after application to the channel number conversion unit 17.

In the above example, the correction coefficient when converting 22.2 ch to 22.2 ch is assumed to be “1.0”, but the present invention is not limited to this. For example, an initial value (an initial value of a correction coefficient corresponding to the number of channels of the multichannel audio signal) can be set according to the importance of each channel signal q _i of the multichannel audio signal. For example, the initial value of the correction coefficient is recorded for each channel signal q _i (1 ≦ i ≦ 24) of the multi-channel audio signal in the storage unit 18, and the correction coefficient application unit 15 receives the channel signal q _i from the storage unit 18. B ^ _i may be calculated by reading the initial value of the correction coefficient for each (1 ≦ i ≦ 24). For example, the storage unit 18, the initial value is "1.0" in the correction factor _{q 1,} the initial value is "0.9" in the correction factor _{q 2,} · · _·, the initial value of the correction factor _{q 24} May be recorded as “0.8” or the like.

  Next, in step S20, the channel number conversion unit 17 converts the input multi-channel audio signal after application of the correction coefficient by the predetermined channel number conversion method (general channel number conversion method) described later, Outputs playback channel audio signals.

  According to the present embodiment, when a multi-channel audio signal and an audio signal (reference signal) having a smaller number of channels are simultaneously provided, a channel that is more in line with the production intention with reference to the reference signal having a smaller number of channels. Number conversion can be realized.

<Second embodiment>
Hereinafter, the weighting coefficient calculator according to the second embodiment of the present invention will be described with reference to FIGS.
FIG. 4 is a block diagram showing an example of the channel number conversion apparatus according to the second embodiment of the present invention.
As shown in FIG. 4, the channel number conversion apparatus 10a includes a reference signal input unit 11, a multichannel audio signal input unit 12, a weighting coefficient calculation unit 13a, a correction coefficient calculation unit 14, a correction coefficient application unit 15, and a reproduction channel information acquisition unit. 16, a channel number conversion unit 17, and a storage unit 18. As described above, the channel number conversion device 10a according to the second embodiment includes the weighting coefficient calculation unit 13a instead of the weighting coefficient calculation unit 13 of the first embodiment. In addition, the other structure of the channel number converter 10a by 2nd embodiment is the same as that of the channel number converter 10 of 1st embodiment.

FIG. 5 is a block diagram showing an example of a weighting coefficient calculator in the second embodiment according to the present invention.
As shown in FIG. 5, the weighting coefficient calculation unit 13 a according to the present embodiment includes a delay correction unit 131, a weighting ratio analysis unit 132, a weighting ratio correction unit 133, and a grouping unit 134.
The grouping unit 134 groups each channel signal of the multi-channel audio signal based on the similarity of each channel signal. For example, the grouping unit 134 calculates a cross-correlation coefficient between each channel audio signal, and groups channels having a large cross-correlation coefficient. Further, the grouping unit 134 may group channels having similar audio signal properties using methods such as principal component analysis and cluster analysis, or both methods. Further, the grouping unit 134 generates a group signal representing channel audio signals belonging to the same group. In this embodiment, the weighting coefficient calculator 13a calculates a weighting coefficient for this group signal. There are the following methods for generating a group signal. For example, the grouping unit 134 may generate an average of channel signals belonging to the same group and treat the generated signal as a group signal. Further, for example, the grouping unit 134 may select a signal at the center of gravity of the similarity of channel signals belonging to the same group, and handle the selected signal as a group signal. Further, for example, the grouping unit 134 may select a channel audio signal having the maximum energy among the channel audio signals belonging to the same group and treat the selected signal as a group signal.

The functions of the delay correction unit 131, the weighting ratio analysis unit 132, and the weighting ratio correction unit 133 are the same as in the first embodiment, and correct the delay of each group signal of the multichannel audio signal with respect to each channel signal of the reference signal. . The weighting ratio analysis unit 132 analyzes the weighting ratio of each group signal of the multichannel audio signal in each channel signal of the reference signal by multivariate analysis. The weighting ratio correction unit 133 energizes each channel signal of the pseudo reference signal obtained by multiplying each group signal of the multichannel audio signal by the energy of each channel signal of the reference signal and the weighting ratio analyzed by the weighting ratio analysis unit 132. A constant c _j is calculated such that the energy of the channel of the pseudo reference signal corresponding to the channel of the reference signal is equal. Each channel audio signal belonging to a group is given a value obtained by equally dividing the weighting ratio given to the group signal according to the number of channels belonging to the group.

Next, the channel number conversion process in the second embodiment will be described with reference to FIG. FIG. 6 is a flowchart showing an example of channel number conversion processing in the second embodiment according to the present invention.
A process similar to the process described with reference to FIG. 3 will be briefly described. First, in step S11, the multichannel audio signal input unit 12 inputs a multichannel audio signal. The multichannel audio signal input unit 12 outputs the multichannel audio signal to the weighting coefficient calculator 13a. In step S12, the reference signal input unit 11 inputs a reference signal. The reference signal input unit 11 outputs each channel signal of the reference signal to the weighting coefficient calculation unit 13a. Next, in step S121, in the weighting coefficient calculator 13a, the grouping unit 134 groups each channel signal of the multi-channel audio signal using a method such as principal component analysis. The grouping unit 134 outputs the group signal after grouping to the delay correction unit 131. Next, in step S13, the delay correction unit 131 inputs the reference signal and the multichannel audio signal, and corrects the delay of the multichannel audio signal. The delay correction unit 131 outputs the multichannel audio signal (each group signal) after delay correction to the weighting ratio analysis unit 132.

Next, in step S14, the weighting ratio analysis unit 132 inputs the multichannel audio signal after delay correction and the reference signal, and calculates the weighting ratio of each group signal of the multichannel audio signal to each channel signal of the reference signal. Calculate using multiple regression analysis. The weighting ratio analysis unit 132 outputs the analyzed weighting ratio to the weighting ratio correction unit 133. Next, in step S15, the weighting ratio correction unit 133 inputs the weighting ratio analyzed by the weighting ratio analysis unit 132 and the reference signal, and corrects the weighting ratio based on energy. In the second embodiment, the weighting ratio correction unit 133 equals the energy of each channel signal of the reference signal and the energy of each channel of the pseudo reference signal obtained by multiplying each group signal of the multichannel audio signal by the weighting ratio. A constant c _j is calculated, and a weighting coefficient obtained by multiplying each weighting ratio a _ij by the square root of c _j is calculated. At this time, the weighting ratio correction unit 133 assigns the same weighting coefficient to each channel signal belonging to the same group signal. In step S <b> 16, the weighting coefficient calculation unit 13 outputs the weighting coefficient calculated by the weighting ratio correction unit 133 to the correction coefficient calculation unit 14.

  Next, in step S17, the correction coefficient calculation unit 14 calculates a correction coefficient in the same manner as in the first embodiment. Next, in step S18, the correction coefficient application unit 15 corrects the correction coefficient according to the number of reproduction channels by linear interpolation or the like. In step S19, the correction coefficient application unit 15 applies the corrected correction coefficient to the multichannel audio signal. The correction coefficient application unit 15 outputs the multi-channel audio signal after applying the correction coefficient to the channel number conversion unit 17. Next, in step S20, the channel number conversion unit 17 converts the number of channels of the multichannel audio signal after applying the correction coefficient by a predetermined method. The channel number converter 10a outputs the playback channel audio signal after the channel number conversion to a playback device or the like.

  According to the second embodiment, the same effect as the first embodiment can be obtained. In addition, according to the second embodiment, among channel audio signals included in a multi-channel audio signal, channel audio signals having similar audio signal characteristics are grouped to perform delay correction processing and weighting coefficient calculation processing. Therefore, the amount of calculation can be suppressed as compared with the first embodiment.

<Third embodiment>
Hereinafter, the weighting coefficient calculator according to the third embodiment of the present invention will be described with reference to FIGS.
FIG. 7 is a block diagram showing an example of the channel number conversion apparatus according to the third embodiment of the present invention.
As shown in FIG. 7, the channel number conversion device 10b includes a reference signal input unit 11, a multichannel audio signal input unit 12, a weighting coefficient calculation unit 13b, a correction coefficient calculation unit 14, a correction coefficient application unit 15, and a reproduction channel information acquisition unit. 16, a channel number conversion unit 17, and a storage unit 18. As described above, the channel number conversion device 10a according to the second embodiment includes the weighting coefficient calculation unit 13b instead of the weighting coefficient calculation unit 13 of the first embodiment. In addition, the other structure of the channel number converter 10b by 2nd embodiment is the same as that of the channel number converter 10 of 1st embodiment.

FIG. 8 is a block diagram showing an example of the weighting coefficient calculator in the third embodiment according to the present invention.
As shown in FIG. 8, the weighting coefficient calculation unit 13b according to the present embodiment includes a delay correction unit 131b, a weighting ratio analysis unit 132b, a weighting ratio correction unit 133, a reference channel signal selection unit 135, and a channel classification unit 136. And including.
The reference channel signal selection unit 135 selects a channel signal (reference channel signal) having the largest cross-correlation coefficient with the channel signal of the reference signal among the channel signals included in the multi-channel audio signal. The reference channel signal selection unit 135 outputs information on the selected reference channel.
The channel classification unit 136 classifies the channel signals included in the multi-channel audio signal into a reference channel audio signal and other (non-reference channel signal).

The delay correction unit 131b corrects the delay of at least one of the standard channel signal and the non-standard channel signal group with respect to the channel signal of the reference signal. The delay of each channel audio signal included in the non-reference channel signal group may be individually corrected.
The weighting ratio analysis unit 132b calculates the weighting ratio after imposing a constraint that the weighting ratio of the reference channel signal is larger than any weighting coefficient of the non-reference channel signal group.

The function of the weighting ratio correction unit 133 is the same as that of the first embodiment, and the channel signal of the multichannel audio signal is multiplied by the energy of each channel signal of the reference signal and the weighting ratio analyzed by the weighting ratio analysis unit 132b. A constant c _j is calculated such that the energy of each channel of the obtained pseudo reference signal is equal to the energy of the channel of the pseudo reference signal corresponding to the channel of the reference signal.

Next, the channel number conversion process in the third embodiment will be described with reference to FIG. FIG. 9 is a flowchart showing an example of channel number conversion processing in the third embodiment according to the present invention.
A process similar to the process described with reference to FIGS. 3 and 6 will be briefly described. First, in step S11, the multichannel audio signal input unit 12 inputs a multichannel audio signal. The multichannel audio signal input unit 12 outputs the multichannel audio signal to the weighting coefficient calculation unit 13b. In step S12, the reference signal input unit 11 inputs a reference signal. The reference signal input unit 11 outputs each channel signal of the reference signal to the weighting coefficient calculation unit 13b. Next, in step S122, in the weighting coefficient calculator 13b, the reference channel signal selector 135 selects a reference channel signal. For example, when the program is a news report program, it is considered that the dialog audio signal is dominant in the reference signal of 2ch. The reference channel signal selection unit 135 selects a channel (a channel signal having the largest cross-correlation coefficient) containing a lot of dialog audio signals as a reference channel signal among 22.2ch channel signals. The reference channel signal selection unit 135 outputs information on the selected reference channel signal (which channel has been selected) to the channel classification unit 136. In step S123, the channel classification unit 136 receives the reference channel signal information and the multi-channel audio signal, and classifies the channel signal included in the multi-channel audio signal into a reference channel signal and a non-reference channel signal. The channel classification unit 136 outputs the classified reference channel signal and non-reference channel signal to the delay correction unit 131b. Next, in step S124, the delay correction unit 131b corrects the delay of the reference channel signal with respect to each channel signal of the reference signal. Next, in step S125, the delay correction unit 131b corrects the delay of the non-reference channel signal with respect to each channel signal of the reference signal. At this time, the delay correction unit 131b may correct the delay for each of the non-reference channel signals. Note that it is preferable to perform both of the processes of steps S124 to S125, but either one may be performed, for example, only the process of correcting the delay of the reference channel signal (step S124). The delay correction unit 131b outputs the reference channel signal and the non-reference channel signal to the weighting ratio analysis unit 132b. Next, in step S141, the weighting ratio analysis unit 132b performs multiple regression analysis on the weighting ratio under the constraint that the weighting ratio of the reference channel signal is larger than any weighting ratio of the channel signal included in the non-reference channel signal. Calculate according to The weighting ratio analysis unit 132b outputs the analyzed weighting ratio to the weighting ratio correction unit 133. Next, in step S15, the weighting ratio correction unit 133 inputs the weighting ratio analyzed by the weighting ratio analysis unit 132 and the reference signal, and corrects the weighting ratio based on energy. In step S <b> 16, the weighting coefficient calculation unit 13 outputs the weighting coefficient corrected by the weighting ratio correction unit 133 to the correction coefficient calculation unit 14.

Next, in step S17, the correction coefficient calculation unit 14 calculates a correction coefficient (b _i ). Next, in step S18, the correction coefficient application unit 15 corrects the correction coefficient according to the number of reproduction channels. In step S 19, the correction coefficient applying unit 15 applies the corrected correction coefficient (b _i ) to the multichannel audio signal, and outputs the multichannel audio signal after applying the correction coefficient to the channel number converting unit 17. Next, in step S20, the channel number conversion unit 17 converts the number of channels of the multichannel audio signal by a predetermined method. The channel number converter 10b outputs the playback channel audio signal after the channel number conversion to a playback device or the like.

  According to the third embodiment, the same effect as that of the first embodiment can be obtained. Further, according to the third embodiment, only the channel signal (reference channel signal) having similar audio signal characteristics is selected from the channel signals included in the multi-channel audio signal, and the delay correction process and the weighting coefficient are selected. Since the calculation process is performed, the amount of calculation can be reduced compared to the first embodiment. The third embodiment can be used in a situation where, for example, a multi-channel audio signal of a program in which a dialog audio signal is dominant, such as a TV talk program, is converted into a reproduction channel audio signal.

Further, the following processing (weighting coefficient calculation, etc.) is performed only for the reference channel signal, and processing is not performed for the non-reference channel signal (the intensity of each signal of the non-reference channel signal group is “0”). May be used as an embodiment). The reference channel signal selection unit 135 not only selects the channel signal having the largest cross-correlation coefficient with the channel signal of the reference signal, but also selects the channel signal having a large cross-correlation coefficient. A plurality of selected channel signals may be selected as a reference channel signal by selecting a plurality of channels in descending order of correlation coefficient or selecting a channel signal having a cross-correlation coefficient equal to or greater than a predetermined threshold.
Note that the sound included in the dialog sound signal is not necessarily limited to the sound of the dialog (dialog). An audio signal mainly composed of a human voice may be used as the dialog audio signal.

<Fourth embodiment>
Hereinafter, the weighting coefficient calculation unit according to the fourth embodiment of the present invention will be described with reference to FIGS.
FIG. 10 is a block diagram showing an example of the channel number conversion apparatus in the fourth embodiment according to the present invention.
As shown in FIG. 10, the channel number conversion device 10c includes a reference signal input unit 11, a multi-channel audio signal input unit 12, a weighting coefficient calculation unit 13, a correction coefficient calculation unit 14, a correction coefficient application unit 15, and a reproduction channel information acquisition unit. 16, a channel number conversion unit 17, a storage unit 18, and a mono signal conversion unit 19. As described above, the channel number converter 10c according to the fourth embodiment includes the mono signal converter 19 in addition to the configuration of the first embodiment. In addition, the other structure of the channel number converter 10c by 4th embodiment is the same as that of the channel number converter 10 of 1st embodiment. Further, the configuration of the weighting coefficient calculator 13 is the same as that described with reference to FIG.

  The mono signal converter 19 downmixes the reference signal into a mono signal by a predetermined method. For example, the mono signal converter 19 converts a 2ch audio signal input by the reference signal input unit 11 into a 1ch audio signal. A known downmix method may be used for conversion from a 2ch audio signal to a 1ch audio signal.

  In the fourth embodiment, since a weighting ratio is calculated for a mono signal or a correction coefficient is calculated, there are portions different from the calculation formulas of those values described in the first embodiment. For example, the weighting ratio correction unit 133 calculates the constant c from the following equation (8).

The correction coefficient calculator 14 calculates the correction coefficient using the weighting coefficient calculated by the weighting coefficient calculator 13 as in the first embodiment. In the fourth embodiment, the correction coefficient calculator 14c Only the weighting coefficient for the mono signal converted from the reference signal is input. Therefore, unlike the first embodiment, it is not necessary to integrate the weighting coefficients for 2ch. Specifically, the correction coefficient calculation unit 14 derives the correction coefficient b ₁ by the following relational expression (9) calculated from the above expression (8).
(B ₁ ) ² = c × (a ₁ ) ² (9)
The correction coefficient calculation unit 14 similarly calculates the values of the other correction coefficients b _{2 to} b ₂₄ .

Next, the channel number conversion process in the fourth embodiment will be described with reference to FIG. FIG. 11 is a flowchart showing an example of channel number conversion processing in the fourth embodiment according to the present invention.
A process similar to the process described with reference to FIG. 3 will be briefly described. First, in step S11, the multichannel audio signal input unit 12 inputs a multichannel audio signal. In step S12, the reference signal input unit 11 inputs a reference signal. The reference signal input unit 11 outputs the reference signal to the mono signal conversion unit 19. Next, in step S126, the mono signal conversion unit 19 converts the reference signal into a 1ch mono signal, and outputs the mono signal to the weighting coefficient calculation unit 13. Next, in the weighting coefficient calculation unit 13, the delay correction unit 131 inputs a mono signal and a multi-channel audio signal. In step S13, the delay correction unit 131 corrects the delay of the multi-channel audio signal with respect to the mono signal. Next, in step S14, the weighting ratio analysis unit 132 inputs the multichannel audio signal and the mono signal after delay correction, and performs a multiple regression analysis or the like on the weighting ratio of each channel signal of the multichannel audio signal to the mono signal. Calculate using. The weighting ratio analysis unit 132 outputs the analyzed weighting ratio to the weighting ratio correction unit 133. Next, in step S15, the weighting ratio correction unit 133 inputs the weighting ratio and the mono signal analyzed by the weighting ratio analysis unit 132, and corrects the weighting ratio based on energy. In the fourth embodiment, the weighting ratio correction unit 133 calculates a constant c such that the mono signal energy is equal to the pseudo mono signal energy obtained by multiplying each channel signal of the multichannel audio signal by the weighting ratio. Then, a weighting coefficient obtained by multiplying each weighting ratio a _i by the square root of c is calculated. In step S <b> 16, the weighting coefficient calculator 13 outputs the weighting coefficient (the same value as the correction coefficient) to the correction coefficient calculator 14. The correction coefficient calculation unit 14 outputs the correction coefficient to the correction coefficient application unit 15.

  Next, in step S18, the correction coefficient application unit 15 corrects the correction coefficient according to the number of reproduction channels. In step S19, the correction coefficient application unit 15 applies the corrected correction coefficient to the multichannel audio signal. Next, in step S20, the channel number conversion unit 17 inputs the multichannel audio signal after applying the correction coefficient by the predetermined rendering method from the correction coefficient application unit 15, and converts the number of channels. The channel number conversion device 10c outputs the playback channel audio signal after the channel number conversion to a playback device or the like.

  According to the fourth embodiment, the same effect as that of the first embodiment can be obtained. Further, since the processing such as calculation of the correction coefficient is performed with reference to the mono signal, the amount of calculation can be reduced as compared with the first embodiment. Note that the fourth embodiment may be applied to the second embodiment or the third embodiment.

  Note that the above-described channel number conversion apparatuses 10, 10a, 10b, and 10c have a computer system therein. The operation process of the channel number conversion device 10 and the like is stored in a computer-readable recording medium in the form of a program, and the above processing is performed by the computer system reading and executing this program. The computer system here includes a CPU, various memories, an OS, and hardware such as peripheral devices.

Further, the “computer system” includes a homepage providing environment (or display environment) if a WWW system is used.
The “computer-readable recording medium” refers to a storage device such as a flexible medium, a magneto-optical disk, a portable medium such as a ROM and a CD-ROM, and a hard disk incorporated in a computer system. Furthermore, the “computer-readable recording medium” dynamically holds a program for a short time like a communication line when transmitting a program via a network such as the Internet or a communication line such as a telephone line. In this case, a volatile memory in a computer system serving as a server or a client in that case, and a program that holds a program for a certain period of time are also included. The program may be a program for realizing a part of the functions described above, and may be a program capable of realizing the functions described above in combination with a program already recorded in a computer system.

  In addition, it is possible to appropriately replace the components in the above-described embodiments with known components without departing from the spirit of the present invention. The technical scope of the present invention is not limited to the above-described embodiment, and various modifications can be made without departing from the spirit of the present invention.

10, 10a, 10b, 10c ... Channel number conversion device 11 ... Reference signal input unit 12 ... Multi-channel audio signal input unit 13, 13a, 13b ... Weighting coefficient calculation unit 131, 131b ... Delay correction unit 132, 132b ... Weighting ratio analysis unit 133 ... Weighting ratio correction unit 134 ... Grouping unit 135 ... Reference channel signal selection unit 136 ... Channel classification unit 14 ... Correction coefficient calculation 15 ... Correction coefficient application unit 16 ... Playback channel information acquisition unit 17 ... Channel number conversion unit 18 ... Storage unit 19 ... Mono signal conversion unit

Claims (12)

Weighting coefficient calculation for inputting a multichannel audio signal and a reference signal corresponding to the multichannel audio signal, and calculating a weighting coefficient corresponding to each channel signal of the multichannel audio signal included in each channel of the reference signal And
A correction coefficient calculator that calculates a correction coefficient to be multiplied to each channel signal of the multi-channel audio signal based on the weighting coefficient;
A correction coefficient applying unit that applies the correction coefficient to the multi-channel audio signal;
A channel number conversion unit that converts the multichannel audio signal to which the correction coefficient is applied into a reproduction channel signal having a desired number of channels by a predetermined channel number conversion method;
A channel number conversion device comprising: A delay correction unit that corrects a delay between the reference signal and each channel signal of the multi-channel audio signal;
The channel number conversion apparatus according to claim 1, further comprising: The weighting coefficient calculation unit inputs the multichannel audio signal and the reference signal, and analyzes a weighting ratio for each channel signal of the multichannel audio signal;
The channel number conversion apparatus according to claim 1 or 2, further comprising: The weighting coefficient calculation unit includes a sum of energy of each channel signal of the reference signal and a signal energy obtained by multiplying each channel signal of the multichannel audio signal corresponding to each channel of the reference signal by the weighting ratio. A weighting ratio correction unit for correcting the weighting ratio to be equal;
The channel number conversion device according to claim 3, further comprising: The weighting factor calculator is
A grouping unit for grouping each channel signal of the multi-channel audio signal based on the similarity of each channel signal, and generating a group signal representing the group based on the channel signal belonging to the group;
The weighting coefficient calculator calculates a weighting coefficient for the group signal;
5. The channel number conversion device according to claim 2, wherein the number of channels is converted. The grouping unit may be any one of an average of the channel signals, a channel signal at the center of gravity of the similarity of the channel signals, and a channel signal having the maximum energy among the channel signals based on the channel signals belonging to the group. Or as the group signal,
The number-of-channels conversion device according to claim 5. The weighting factor calculator is
A reference channel signal selection unit that selects one or a plurality of channel signals based on a cross-correlation coefficient with the reference signal from each channel signal of the multi-channel audio signal;
The weighting coefficient calculation unit calculates a weighting coefficient on the condition that the weighting coefficient of the selected channel signal is larger than the weighting coefficients of other channel signals.
5. The channel number conversion device according to claim 2, wherein the number of channels is converted. The weighting factor calculator is
A reference channel signal selection unit that selects one or a plurality of channel signals based on a cross-correlation coefficient with the reference signal from each channel signal of the multi-channel audio signal;
The weighting coefficient calculator calculates a weighting coefficient only for the channel signal selected by the reference channel signal selector.
5. The channel number conversion device according to claim 2, wherein the number of channels is converted. The correction coefficient calculation unit includes the sum of the energy of the signal obtained by applying the weighting coefficient to each channel signal of the multichannel audio signal corresponding to the total energy of the reference signal or each channel of the reference signal, and the multichannel audio Calculating the correction coefficient so that the sum of energy of signals obtained by applying the correction coefficient to each channel signal of the signal is equal;
9. The channel number converter according to claim 1, wherein the number of channels is converted. The correction coefficient application unit corrects the predetermined initial value of the correction coefficient corresponding to the number of channels of the multi-channel audio signal and the correction coefficient calculated by the correction coefficient calculation unit and corresponding to the number of channels of the reference signal. And correcting the correction coefficient calculated by the correction coefficient calculation unit according to the number of channels of the reproduction channel signal by interpolation based on the coefficient,
10. The channel number conversion apparatus according to claim 1, wherein 11. The channel number conversion apparatus according to claim 1, further comprising: a mono signal conversion unit that converts the reference signal into a mono signal by a predetermined channel number conversion method.   A program for causing a computer to function as the channel number conversion device according to any one of claims 1 to 11.

Images