JP7132027B2 - Sound processing device and program - Google Patents

Description

The present invention relates to an audio processing apparatus and program for converting the number of channels and channel arrangement of multichannel audio signals.

In some cases, program audio consisting of audio signals of multiple audio channels is converted (down-mixed) to a number of channels that is smaller than the number of channels at the time of production, and listened to. At that time, the audio signals obtained by multiplying the down-mix coefficients (channel number conversion coefficients) are added to the original audio signals of each audio channel during production, and the converted audio signals of each audio channel during playback are added. Generate. As downmix coefficients, there are those specified by the standard specifications of a predetermined audio system (see, for example, Non-Patent Document 1) and those specified based on metadata added to the audio signal for each program (for example, , Non-Patent Document 2).

For example, Non-Patent Document 2 describes downmix processing from a 5.1ch surround sound signal to a stereo 2ch sound signal using the relationship shown in Equation (1).

In equation (1), Lt and Rt indicate left-channel and right-channel acoustic signals, respectively, of the two stereo channels that are conversion destinations. L, R, C, LFE, Ls, and Rs are the front left channel, front right channel, front center channel, low frequency effect sound channel, surround left channel, and surround right channel, respectively, of the 5.1ch surround that is the conversion source. Each acoustic signal is shown. Therefore, each element of the matrix of 2 rows and 6 columns in the first term of Equation (1) corresponds to the downmix coefficient. k represents a coefficient that defines the levels of the surround left channel Ls and the surround right channel Rs. k is 1/√2, 1/2, 0, for example. If no metadata is transmitted along with the audio signal, 1/√2 is used as k in the television receiver.

In addition, in recent years, a multi-channel sound system such as 22.2 multi-channel sound (hereinafter referred to as "22.2 ch sound") has been introduced in ultra-high definition television broadcasting such as 4K/8K broadcasting. On the other hand, recently, multi-channel sound, which adds upper 2ch or upper 4ch to 5.1ch surround or 7.1ch surround, has been proposed. .1ch (5.1+4ch), 11.1ch (7.1+4ch), etc. may be used for reproduction.

In addition, a method has been proposed in which a three-dimensional panning method (for example, see Non-Patent Document 3), which distributes amplitude to surrounding speakers by vector synthesis, is applied to channel number conversion, with an emphasis on reproducing the direction of arrival of sound. ing.

International Telecommunication Union, Recommendation ITU-BS.775-3, "Multichannel stereoscopic sound system with and without accompanying picture", 08/2012 Association of Radio Industries and Businesses, ARIB STD-B32 Version 3.6, "Video Coding, Audio Coding and Multiplexing Methods in Digital Broadcasting", March 25, 2016 V. Pulkki, "Virtual sound source positioning using vector base amplitude panning", J. Audio Eng. Soc., 45, 6, pp.456-466, 1997

However, at present, channel number conversion coefficients for converting from 22.2ch sound to 7.1ch (5.1+2ch), 9.1ch (5.1+4ch), 11.1ch (7.1+4ch), etc. have not been determined. In the future, sound systems with various numbers of channels and sound systems with different arrangements even with the same number of channels will be proposed. However, when converting from one channel arrangement to another, or when converting from a channel arrangement with a small number of channels to a channel arrangement with a large number of channels, it may be necessary to specify a channel number conversion factor. Determining the conversion coefficients is difficult.

In addition, in the channel number conversion that applies the method described in Non-Patent Document 3, if the number of conversion source channels at different playback positions from the conversion destination channel is large, the number of sounds expressed by the virtual sound source increases, and the overall sound becomes There is a problem that the timbre may deteriorate due to blurring.

It is an object of the present invention, which has been made in view of such circumstances, to produce an audio signal of program audio produced by an audio system having an arbitrary channel arrangement with a plurality of layers in the vertical direction, without significantly impairing the intention of the producer. An object of the present invention is to provide an audio processing device and a program capable of deriving a channel number conversion coefficient matrix for reproduction in an audio system having another channel arrangement.

In order to solve the above problems, an acoustic processing apparatus according to the present invention is an acoustic processing apparatus that converts the number of channels and channel arrangement of a multi-channel acoustic signal, wherein the reproduction positions of the converted multi-channel acoustic signal are set to a plurality of positions in the elevation direction. a conversion destination channel division unit that divides into layers; a conversion source channel reproduction position determination unit that determines a reproduction position of a conversion source multi-channel audio signal on each divided layer divided by the conversion destination channel division unit; A channel number conversion coefficient determination unit that determines a channel number conversion coefficient based on the reproduction position of the destination multi-channel acoustic signal and the reproduction position of the original multi-channel acoustic signal determined by the original channel reproduction position determination unit. and an output for generating the destination multi-channel acoustic signal from the source multi-channel acoustic signal using a channel number transform coefficient matrix whose elements are the channel number transform coefficients determined by the channel count transform coefficient determination unit and a signal generator.

Further, in the sound processing device according to the present invention, the conversion destination channel dividing unit can set the elevation angle at least to the foot (-90 degrees elevation), the floor layer (-90 to -45 degrees elevation), the lower layer (-45 degrees elevation) ~ 0 degrees), middle layer (0 degrees elevation), upper layer (0 to 45 degrees elevation), ceiling layer (45 degrees to 90 degrees elevation), overhead (90 degrees elevation), and these adjacent layers are divided into a plurality of layers in which the multi-channel acoustic signals to be converted are integrated in an arbitrary combination.

Further, in the sound processing apparatus according to the present invention, the destination channel division unit divides the reproduction position of the destination multi-channel acoustic signal into a plurality of layers in the elevation direction, and divides each divided layer in the azimuth direction. It is characterized by dividing into a plurality of ranges.

Further, in the sound processing device according to the present invention, the conversion destination channel dividing unit divides the reproduction position of the conversion destination multichannel sound signal into a plurality of ranges in the azimuth direction, dividing the conversion destination multichannel sound signal into For a playback position missing layer that includes a playback position missing range in which the playback position of The reproduction position of the converted multi-channel audio signal in the reproduction position missing range is characterized.

Further, in the sound processing device according to the present invention, the source channel reproduction position determining unit may determine two or more reference points of the reproduction position of the source multi-channel acoustic signal and the reproduction position of the destination multi-channel sound signal. Rotating and scaling the surround circle of the original multi-channel acoustic signal so that two or more reference points coincide, and determining the reproduction position of the original multi-channel acoustic signal existing between the two reference points. It is characterized by rearranging on the surround circle of the conversion target multi-channel audio signal.

Further, in the sound processing device according to the present invention, the conversion destination channel dividing unit divides each of the divided layers into four ranges of front, back, left, and right, and the conversion source channel reproduction position determination unit divides the conversion source multi-channel audio signal into four ranges. The reference points for the playback position of and the playback position of the conversion destination channel are set to the left front and right front two points corresponding to both ends of the screen divided by the conversion destination channel dividing unit, and the left front, right front, left rear, and right rear points that are divided into four. , or 6 points obtained by adding 90 degrees to the left and right sides to the four points in the four corners of the four ranges, respectively.

Further, in the sound processing apparatus according to the present invention, the conversion-source channel reproduction position determination unit sets the reference point to a predetermined position, a position specified by a user, a position specified by metadata, or a conversion-source channel playback position determination unit. It is characterized by determining based on the degree of similarity in the names or positions of the channel and the destination channel.

Further, in the sound processing device according to the present invention, the number-of-channels conversion coefficient determination unit reproduces the reproduction positions of two adjacent conversion target multi-channel acoustic signals and the conversion source channel sandwiched between the reproduction positions. The number-of-channels conversion coefficient is calculated based on the number of reproduction positions of the original multi-channel acoustic signal determined by the position determining unit.

Further, in the sound processing apparatus according to the present invention, the channel number conversion coefficient determining unit replaces the calculated channel number conversion coefficient with a predetermined representative value when the calculated channel number conversion coefficient is included in a predetermined range. do.

Further, in order to solve the above problems, a program according to the present invention causes a computer to function as the above sound processing device.

According to the present invention, sound produced in an audio system with an arbitrary channel arrangement having a plurality of layers in the vertical direction can be reproduced by a playback device for an audio system with a different channel arrangement without significantly impairing the intention of the producer. It is possible to generate a channel number transform coefficient matrix for listening.

1 is a block diagram showing the configuration of a sound processing device according to one embodiment of the present invention; FIG. 4 is a flowchart showing a procedure for deriving channel number conversion coefficients in the sound processing device according to the embodiment of the present invention; FIG. 4 is a diagram showing an example of division of conversion destination channels in the sound processing device according to the embodiment of the present invention; A first example of the playback position of the conversion source channel with an elevation angle of ±90 degrees when the playback position of the conversion destination channel does not exist at one point with an elevation angle of ±90 degrees in the sound processing apparatus according to the embodiment of the present invention is as follows. FIG. 4 is a diagram showing; A second example of the reproduction position of the conversion source channel with an elevation angle of ±90 degrees when the reproduction position of the conversion destination channel does not exist at one point with an elevation angle of ±90 degrees in the sound processing apparatus according to the embodiment of the present invention is shown below. FIG. 4 is a diagram showing; FIG. 4 is a diagram showing an example of a reproduction position of a conversion-source channel when an absolute position is held in the sound processing device according to one embodiment of the present invention; FIG. 5 is a diagram showing an example of correction of the reproduction position (lateral) of the conversion source channel based on the reference point in the sound processing device according to the embodiment of the present invention; FIG. 5 is a diagram showing an example of correction of the reproduction position (forward, backward) of the conversion source channel based on the reference point in the sound processing device according to the embodiment of the present invention; FIG. 4 is a diagram showing an example of a reproduction position of a conversion source channel when reference points are ±60 degrees of the conversion source channel and ±30 degrees of the conversion destination channel in the sound processing apparatus according to the embodiment of the present invention; FIG. 4 is a diagram showing an example of a reproduction position of a conversion-source channel when reference points are ±30 degrees of the conversion-source channel and ±30 degrees of the conversion-destination channel in the sound processing apparatus according to the embodiment of the present invention; FIG. 5 is a diagram showing an example of changing the reproduction position of the conversion source channel based on the angle ratio in the sound processing device according to the embodiment of the present invention; FIG. 4 is a diagram illustrating an example of defining coefficients according to the angle between the reproduction position of two conversion destination channels and the reproduction position of the conversion source channel in the sound processing device according to the embodiment of the present invention; FIG. 4 is a diagram for explaining an example of how to obtain channel distribution coefficients when coefficients are defined by the number of conversion source channels sandwiched between the reproduction positions of two conversion destination channels in the sound processing apparatus according to one embodiment of the present invention; be. FIG. 10 is a diagram showing an example of replacing the distribution coefficients of the acoustic signals so as to give clear numbers when expressed in decibels, in the sound processing device according to the embodiment of the present invention; 4 is a flow chart showing a procedure for deriving channel number conversion coefficients of the sound processing apparatus according to Example 1 of the present invention; FIG. 10 is a flow chart showing a procedure for deriving channel number conversion coefficients of the sound processing device according to the second embodiment of the present invention; FIG.

Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings.

FIG. 1 is a block diagram showing a configuration example of a sound processing device 1 according to this embodiment. The sound processing apparatus 1 shown in FIG. 1 includes a conversion destination channel division unit 11, a conversion source channel reproduction position determination unit 12, a channel number conversion coefficient determination unit 13, a channel number conversion coefficient matrix storage unit 14, and an output signal generation unit. a part 15;

The sound processing device 1 converts the number of channels and their arrangement from a conversion source multi-channel sound signal consisting of an arbitrary number of M channels to a conversion destination multi-channel sound signal consisting of an arbitrary number of N channels.

The conversion destination channel dividing unit 11 acquires conversion source/conversion destination channel position information including the number of conversion source channels and conversion destination channels. Further, the destination channel division unit 11 receives destination channel division information indicating how to divide the reproduction positions of the N destination multi-channel acoustic signals. Then, the conversion destination channel division unit 11 divides the reproduction position of the conversion destination channel based on the conversion destination channel division information. For example, the reproduction position of the conversion destination channel is divided into three in the elevation direction (height direction, vertical axis direction), and each of the three divided layers is further divided into four ranges in the azimuth direction. Details will be described later. Hereinafter, the "playback position of the conversion source multi-channel acoustic signal" will be referred to as the "source channel playback position", and the "playback position of the conversion destination multichannel acoustic signal" will be referred to as the "destination channel playback position".

A conversion source channel reproduction position determination unit 12 is a conversion position calculation that specifies whether the reproduction position of the conversion source channel is set to an absolute position (absolute position mode) or changed to a position relative to a certain reference position (relative position mode). Enter mode selection information. Based on the conversion position calculation mode selection information, the conversion source channel reproduction position determination unit 12 determines the reproduction position of the conversion source channel on each divided layer of the conversion destination channel divided by the conversion destination channel division unit 11. do. Details will be described later.

The number-of-channels conversion coefficient determination unit 13 defines the distribution coefficients according to the positional relationship between the reproduction positions of the two conversion destination channels and the reproduction positions of the conversion source channel (position mode), or the reproduction positions of the two conversion destination channels. Acquire distribution coefficient calculation mode selection information that specifies whether to define by the number of conversion source channels sandwiched between (number mode). After calculating the distribution coefficients based on the distribution coefficient calculation mode selection information, the channel number conversion coefficient determination unit 13 determines the channel number conversion coefficients and stores them in the channel number conversion coefficient matrix storage unit 14 .

The channel number conversion coefficient matrix storage unit 14 stores a channel number conversion coefficient matrix required for channel number conversion. For example, it stores a channel number conversion coefficient matrix for converting 22.2ch sound to 11.1ch sound (7.1+4ch), 9.1ch sound (5.1+4ch), and 7.1ch sound (5.1+2ch), which will be described later. Note that the channel number conversion coefficient matrix storage unit 14 is not an essential component. In that case, the number-of-channels transform coefficient matrix may be derived each time.

The output signal generator 15 inputs the conversion source multi-channel acoustic signal. Then, the output signal generating unit 15 uses a channel number transform coefficient matrix whose elements are the channel number transform coefficients determined by the channel number transform coefficient determining unit 13 to determine the number of channels of the source multichannel acoustic signal and their arrangement. is converted to generate a converted multi-channel audio signal and output to the outside.

Specifically, the output signal generation unit 15 reads out a channel number conversion matrix composed of N×M channel number conversion coefficients from the channel number conversion coefficient matrix storage unit 14, and according to equation (2), M conversion elements Multiplying the multi-channel sound signals by the channel number transform coefficients and adding them generate N multi-channel sound signals to be transformed.

FIG. 2 is a flow chart showing a procedure for deriving channel number conversion coefficients used when channel number conversion is performed from conversion source channels to conversion target channels in the sound processing apparatus 1 . When converting an arbitrary M-channel multi-channel acoustic signal into an N-channel multi-channel acoustic signal, first, the conversion destination channel division unit 11 divides the reproduction positions of the N conversion destination channels into a plurality of layers in the elevation direction. (step S11).

The speakers (actual playback positions of the conversion destination channels) installed in the sound reproduction space, which is assumed to be a cubic listening room, are categorized as floor, lower, middle, upper, and ceiling of the side walls, and the singular point of the feet (elevation angle -90 degrees), overhead (+90 degrees elevation), and a multi-layered structure including feet (-90 degrees elevation), floor layer (-90 to -45 degrees elevation), lower layer (-45 degrees to 0 degrees elevation) , middle layer (elevation angle 0 degrees: ear height), upper layer (elevation angle 0 to 45 degrees), ceiling layer (elevation angle 45 degrees to 90 degrees), overhead (elevation angle 90 degrees). The conversion destination channel dividing unit 11 classifies the elevation angle into at least seven layers, and divides the reproduction position of the conversion destination channel into a plurality of layers in which these adjacent layers are combined in arbitrary combinations. Since speakers are actually installed only in a limited range, for example, the upper layer and the ceiling layer may be integrated into one upper layer (0 to 90 degrees elevation angle). In the case of 22.2 ch, there are four layers: lower layer, middle layer, upper layer, and overhead. In the case of 11.1 ch, there are two layers, middle layer and upper layer. More than seven layers may be used if the layers are more precisely defined by elevation.

Next, the conversion destination channel dividing unit 11 divides each divided layer of the conversion destination channel into two ranges, left and right, in the azimuth direction (step S12), or divides it into four ranges, front, back, left, and right (cross plane, median plane). (step S13), or divided into six ranges of front, middle, back, left and right (step S14). The switching in steps S12 to S14 may be determined in advance by default settings, or may be determined by the user.

In steps S12 to S14, the reproduction position of the conversion destination channel may not exist in the range divided in the azimuth direction. Hereinafter, the range where the reproduction position of the conversion destination channel does not exist will be referred to as the reproduction position missing range, and the layer including the reproduction position missing range will be referred to as the playback position missing layer. The conversion destination channel dividing unit 11 virtually divides the reproduction position of the conversion destination channel existing in the range corresponding to the reproduction position missing range of the adjacent layer into the reproduction position missing layer for the playback position missing layer including the playback position missing range. It is set as the reproduction position of the conversion destination channel in the range (step S15). However, a layer composed of only one point such as overhead (90 degrees elevation angle) is excluded.

In FIG. 3, the playback positions of the conversion destination channels of 7.1ch (5ch in the middle layer and 2ch in the upper layer) and 11.1ch (7ch in the middle layer and 4ch in the upper layer) are divided into 2 layers, and each divided layer is divided into 4 parts in the front, back, left and right. example. In the 7.1ch sound shown in FIG. 3A, the upper layer is a reproduction position missing layer in which the reproduction position of the conversion destination channel does not exist in the rear range. Therefore, the rear channel of the adjacent middle layer is also used for reproduction of the upper layer.

In step S15, as layers adjacent to the reproduction position missing layer, the overhead layer composed of only one point with an elevation angle of 90 degrees and the foot layer composed of only one point with an elevation angle of -90 degrees are excluded. Further, the conversion destination channel dividing unit 11, when the reproduction position of the conversion destination channel exists in the range corresponding to the reproduction position missing range of both adjacent layers above and below the reproduction position missing layer, The reproduction position of the conversion destination channel in the corresponding range on the farther layer is assumed to be the reproduction position of the conversion destination channel in the missing range. For example, if the upper layer is the playback position missing layer, and the playback position of the conversion destination channel exists in the range corresponding to the playback position missing range of the ceiling layer and middle layer adjacent to the top and bottom of the upper layer, the elevation angle is away from 0 degrees. Let the reproduction position of the conversion destination channel in the ceiling layer on the other side be the reproduction position of the conversion destination channel in the reproduction position missing range. As a result, it is possible to maintain the up-and-down feeling at the time of production.

Here, if the conversion destination channel is stereo, etc., and the playback position of the conversion destination channel does not exist in the range corresponding to the playback position missing range of the layer adjacent to the playback position missing layer, the playback position missing layer The playback position of the destination channel may be virtually created by reversing the playback position of the destination channel back and forth within. Also, if the upper layer and ceiling layer are further divided into multiple layers, or if the elevation angle difference between layers adjacent to a certain layer is small, the playback position of the conversion destination channel on both adjacent layers will be changed to the corresponding layer. As the playback position, the original multi-channel acoustic signal may be distributed to both channels.

Although FIG. 3 shows an example in which each divided layer is divided into four layers, it may be divided into six layers such as left front, left horizontal, left rear, right front, right horizontal, and right rear. Only when there is no channel whose playback position exists at the front left, front right, rear left, or rear right, the channel whose playback position exists at 0 degrees or 180 degrees is equally distributed, for example, front left, front right (± 1 degree, etc.), or a channel in which the playback position exists in the rear left or rear right (±179 degrees). Also, a channel whose playback position exists at ±90 degrees laterally is normally regarded as a channel whose playback position exists in the range of rear left and rear right. may be regarded as a channel in which the playback position exists in the range of front left and front right.

In this way, by dividing the playback position of the conversion destination channel into multiple layers in the elevation direction, it is possible to reproduce the sound produced by an arbitrary channel arrangement audio system with multiple layers in the vertical direction according to the intention of the producer. It is possible to generate a number-of-channels transform coefficient matrix for listening on a playback device for an audio system with a different channel arrangement without significant loss.

Next, as shown in FIG. 2, the conversion-source channel playback position determining unit 12 determines the conversion-source channel at the elevation angle of ±90 degrees when the playback position of the conversion destination channel does not exist at one point of the elevation angle of ±90 degrees. is specified (step S16). If the reproduction position of the conversion destination channel exists at one point with an elevation angle of ±90 degrees, the reproduction position is not changed.

If the playback position of the conversion destination channel does not exist at one point with an elevation angle of ±90 degrees, in the first example, the playback position of the conversion source channel above the head (elevation angle of 90 degrees) or at the foot (elevation angle of −90 degrees) is set to the elevation angle of ±90 degrees. The playback position of the conversion destination channel closest to the center of each divided range on the layer adjacent to 90 degrees, and if there are multiple playback positions of the conversion destination channel that are equally distant from the center in the divided range, the multiple is the playback position of the conversion destination channel.

FIG. 4 shows a first example of the reproduction position of the conversion source channel with an elevation angle of ±90 degrees when the reproduction position of the conversion destination channel does not exist at one point with an elevation angle of ±90 degrees. When each divided layer is divided into four ranges, among the four ranges of conversion destination channels in the adjacent layers, the conversion destination channels whose playback positions are near the four corners (azimuth angles of 45 degrees and 135 degrees) are reproduced. Use one point for each position. For example, as shown in FIG. 4(a), if the conversion destination channel is 5ch, 4ch existing near the four corners indicated by the black circles is defined as the playback position, and as shown in FIG. 4(b), the conversion destination channel is 7ch, 4ch present near the four corners indicated by black circles is defined as the playback position. Also, if there are two conversion destination channels at playback positions that are equally distant from the four corners, such as azimuth angles of 30 degrees and 60 degrees, both are used. For example, as shown in FIG. 4(c), when the conversion destination channel is 10ch, both azimuth angles of 30 degrees and 60 degrees are used, and 6ch indicated by a black circle is defined as the playback position.

Once the reproduction position of the conversion source channel is defined, the level of the acoustic signal is defined according to the defined number of channels. When one point is selected from four ranges, the channel number conversion factor is √(1/4)=0.500, and when two points are selected from four ranges, the channel number conversion factor is √(1/8 )=0.354. In this example, it is assumed that energy is conserved. However, when an acoustic signal that should be reproduced in the upper layer is reproduced in the middle layer, etc., the sound signal originally reproduced in the middle layer may become difficult to hear. Therefore, a correction coefficient such as −1.5 dB may be applied to the overall energy. At this time, each element is 0.421 when one point is selected from four ranges, and each element is 0.297 when two points are selected from four ranges.

In the second example, the playback position of the conversion source channel above the head (90 degrees elevation angle) or at the feet (−90 degrees elevation angle) is converted to the center of the left and right divided range behind the layer adjacent to the ±90 degrees elevation angle, respectively. The playback position of the destination channel and the playback position of the conversion destination channel in the front center. If the reproduction position of the conversion destination channel does not exist in the front center, instead, the reproduction position of the conversion destination channel closest to the center of the front left and right divided ranges is set as the reproduction position of the conversion source channel.

FIG. 5 shows a second example of the reproduction position of the conversion source channel with an elevation angle of ±90 degrees when the reproduction position of the conversion destination channel does not exist at one point with an elevation angle of ±90 degrees. For example, as shown in FIG. 5(a), when the conversion destination channel is 5ch, 2ch existing near the rear two corners indicated by black circles and 1ch in the front center are defined as the playback position, and FIG. As shown, when the conversion destination channel is 7ch, 2ch near the rear two corners and 1ch in the front center indicated by the black circles are defined as the playback position. Also, if the conversion destination channel does not exist in the front center, the reproduction positions of the conversion destination channels that exist in the front left and right positions may be used. For example, as shown in FIG. 5(c), when the conversion destination channel is 4ch, 2ch near the rear two corners and the front left and right 2ch indicated by the black circles are defined as the playback positions.

When the two rear corners and the front center are the reproduction positions, each element of the channel number conversion coefficient matrix is √(1/3)=0.577. When the vicinity of the two rear corners and the front left and right are the reproduction positions, each element of the front left and right channel number conversion coefficient matrix is √(1/6)=0.408. In addition, as described above, in order to prevent the sound signal originally reproduced in the middle layer from being difficult to hear due to the sound signal that should be reproduced in the upper layer being reproduced in the middle layer, etc., the correction coefficient is multiplied and the overall energy may be reduced.

Next, as shown in FIG. 2, the reproduction positions of the conversion-source channel excluding the elevation angles of ±90 degrees are arranged on each divided layer of the conversion-destination channel. At this time, there is an absolute position mode (step S17) in which the reproduction position of the conversion source channel is arranged as it is, and a relative position mode in which a reference point is specified for the conversion source/conversion destination channel and arranged so that the relative relationship with respect to the reference point matches. (Step S18), the arrangement destination of the reproduction position of the conversion source channel is switched. The switching in steps S17 and S18 may be determined in advance by default settings, or may be determined by the user.

In the absolute position mode (step S17), the reproduction positions of the conversion source channels (M) are set as they are without being changed. In other words, the reproduction position defined in advance by the standard is used as it is.

FIG. 6 shows the playback positions of the conversion-source channel and the conversion-destination channel in the absolute position mode. Here, the inner white circle is the reproduction position of the conversion source channel, and the outer black circle is the reproduction position of the conversion destination channel. FIG. 6(a) is conversion from 10ch to 5ch, FIG. 6(b) is conversion from 10ch to 7ch, and FIGS. 6(c) and 6(d) are conversion from 8ch to 4ch. In (c) and (d) of FIG. 6, the arrangement of 4 channels of conversion destination channels is different. In the absolute position mode, the playback position of the conversion source channel is not changed regardless of the placement of the conversion destination channel.

In the relative position mode (step S18), the conversion-source channel reproduction position determining unit 12 changes the reproduction positions of the conversion-source channels (M) so as to maintain the relative positions with respect to the reference point. Specifically, the surround circle of the conversion-source channel is rotated and scaled (enlarged/scaled) so that the two reference points of the reproduction position of the conversion-source channel and the two reference points of the reproduction position of the conversion-destination channel are aligned. The reproduction position of the conversion-source channel located between the two reference points is rearranged on the surround circle of the conversion-destination channel to be the reproduction position of the conversion-source channel.

Here, the reference points are the left front and right front points corresponding to both ends of the screen, and the left front, left rear, right front and right rear four points (±45° and ±135 degrees), or the reproduction position of the conversion-source channel that exists closest to any of the six points obtained by adding 90 degrees to the left and right sides to the four points. If you want to specify the direction from which the sound is heard, you can increase the number of reference points to, for example, 8 points. If there are three or more reference points, select two adjacent reference points, and reproduce the conversion destination channel of the conversion source channel whose playback position is on the arc of the surround circle formed by the two reference points. Define the position. By repeating this operation, the arc formed by two adjacent reference points of the conversion source goes around the surround circle of the conversion source, and the reproduction positions of all the conversion source channels at the conversion destination are determined.

The two reference points are not limited to the playback positions of the conversion-source channel and the conversion-destination channel, and arbitrary positions can be specified. For example, by setting ±90 degrees right beside the listener as a reference point, it is possible to fix the direction of 90 degrees and change the ratio of changing the playback position in front of and behind the listener. Information on the position of this reference point may be included as metadata in the conversion position calculation mode selection information, may be determined in advance, or may be defined based on the similarity of channel names such as Left and Right. That is, the conversion-source channel reproduction position determining unit 12 sets the reference point to a predetermined position, a position specified by the user, a position specified by metadata, or a position corresponding to the names or positions of the conversion-source channel and the conversion-destination channel. Determined based on similarity.

FIG. 7 shows an example of changing the reproduction position of the conversion-source channel according to the reference point. Of the four reference positions on the front, back, left, and right, the reproduction position of the conversion-source channel on the side sandwiched between the left front and left rear points is shown in FIG. Here is an example of changing . In this example, it is assumed that the conversion destination reference points are 30 degrees and 110 degrees, and the conversion source reference points are 60 degrees and 135 degrees. Here, when converting the source 60 degrees (θ _in1 ) to the destination 30 degrees (θ _out1 ), and the source 135 degrees (θ _in2 ) to the destination 110 degrees (θ _out2 ), the source channel is converted to the reproduction position θ _out of the conversion destination channel shown _in equation (3).

Also, FIG. 8 shows an example of changing the reproduction positions of the front and rear conversion source channels with two points on the left and right front as reference points.

FIG. 9 shows an example of changing the playback position of the conversion source channel when the playback position of the conversion destination channel of ±30 degrees and the playback position of the conversion source channel of ±60 degrees are used as reference points (relative position of 60 degrees). indicates Here, the inner white circle is the reproduction position of the conversion source channel, the outer black circle is the reproduction position of the conversion destination channel, and the asterisk indicates the reference point. 9(a) and 9(b) are conversion from 10ch to 5ch, FIG. 9(c) is conversion from 10ch to 7ch, and FIGS. 9(d)(e)(f) are conversion from 8ch to 4ch. is a conversion. In FIGS. 9B, 9C, and 9E, a reference point is also added at 90 degrees to the side.

In FIG. 9, in addition to the two front channels, there are also reference points on the rear and sides. If one point from the four corners is used as a reference point, and the playback position of the conversion destination channel is 5.1 ch as shown in FIG. 9A, the side channels move slightly forward. Therefore, as shown in FIG. 9B, although there is no conversion destination channel at 90 degrees to the side, by using it as a reference point, the playback position of the side channel is fixed regardless of the playback position of the rear channel. be able to.

Also, in FIG. 10, when the playback position of the conversion destination channel of ±30 degrees and the playback position of the conversion source channel of ±30 degrees are used as reference points (relative position of 30 degrees), the playback position of the conversion source channel is changed. Here is an example of The numbers of conversion source channels and conversion destination channels are the same as in FIG.

After changing the playback position based on the reference point, the playback position of the conversion-source channel that exists near the target conversion-destination channel other than the conversion-source/conversion-destination channel corresponding to the reference point is changed to the same playback position as the conversion-destination channel. change to position. At this time, if there are a plurality of conversion-source channel candidates whose playback position is to be changed, either the angle-difference mode in which the conversion-source channel at a closer position is simply changed by the angle difference (step S19), or the conversion-source channel that is the most biased in terms of the angle ratio is selected. The angle ratio mode (step S20) for changing the conversion source channel is switched. Switching between steps S19 and S20 may be determined in advance by default settings, or may be determined by the user.

In the case of the angle ratio mode (step S20), specifically, the reproduction position of the conversion-source channel that exists at the position most skewed to the target conversion-destination channel is changed. First, the ratio of the angle between the reproduction position of each conversion-source channel and the reproduction positions of two conversion-destination channels adjacent to the conversion-source channel is obtained. This ratio indicates to what extent the conversion source channel of interest exists in a biased position to one side between two adjacent conversion destination channels. The reproduction position of the conversion source channel, which is largely biased toward one of the reproduction positions of the two adjacent conversion destination channels, is changed to the reproduction position of the adjacent conversion destination channel.

FIG. 11 shows an example of converting the reproduction position in the angle ratio mode. As shown in FIG. 11(a), when converting from 10ch to 5ch, the channels present at 30 and 110 degrees from the conversion destination channel correspond to the channels present at 90 degrees from the source channel. is adjacent to The ratio of angles between these channels is 20 degrees to 60 degrees. Channels existing at positions of 110 degrees and 250 degrees of the conversion destination channel are adjacent to the channel existing at the position of 135 degrees from the conversion source channel. The angle ratio between these channels is 25 degrees to 115 degrees. That is, the channel existing at the 135-degree position of the conversion-source channel has a greater bias between the two adjacent conversion-destination channels than the channel existing at the 90-degree position of the conversion-source channel. Therefore, the reproduction position of the channel existing at the position of ±135 degrees from the conversion source channel is converted to ±110 degrees. The remaining channels of the conversion source channel are not changed.

Similarly, when converting from 8ch to 4ch as shown in FIG. The playback position of the channel existing at the position of degrees is converted to ±110 degrees. The remaining channels of the conversion source channel are not changed.

In the above example, the reproduction position of the conversion-source channel is defined regardless of the difference between the reproduction position of the conversion-source channel and the reproduction position of the conversion-destination channel. If the difference in the playback position of the conversion destination channel is equal to or less than the threshold, the playback position of the conversion source channel may be changed to the playback position of the conversion destination channel. It should be noted that different numerical values may be used for this threshold in the four ranges of front, rear, left, and right. For example, if the front threshold is 15 degrees and the rear threshold is 25 degrees, the front channel of 45 degrees for 22.2 ch sound and the rear channels of 30 degrees, 90 degrees and 135 degrees for 5.1 ch sound are both It will be reproduced at 110 degrees of 5.1ch sound.

Next, as shown in FIG. 2, the distribution coefficients of the original multi-channel acoustic signals reproduced from the original channels are defined based on the changed reproduction positions of the original channels and the changed reproduction positions of the destination channels. A source channel sandwiched between reproduction positions of two destination channels distributes a source multi-channel audio signal to the two destination channels. The coefficient at that time is called the distribution coefficient. At this time, a position mode (step S21) that defines distribution coefficients according to the positional relationship (for example, angle or distance ratio) between the reproduction positions of the two conversion destination channels and the reproduction position of the conversion source channel, and the two conversion destination The operation mode is switched to the number mode (step S22) in which the distribution coefficient is defined by the number of conversion source channels sandwiched between the reproduction positions of the channels. Switching between steps S21 and S22 may be determined in advance by default settings, or may be determined by the user.

FIG. 12 is a diagram illustrating an example of how to obtain distribution coefficients in the position mode (step S21). In the position mode, for example, the coefficients to be distributed to the conversion destination channels are calculated from the angle (2θ ₀ ) formed by the two conversion destination channels and the angle (θ ₁ ) formed by the conversion source channel from its center. Methods for calculating coefficients include methods such as Tan's rule and VBAP described in Non-Patent Document 3, but any method may be used. For example, the distribution coefficients w ₀₁ and w ₀₂ are determined by equation (4).

On the other hand, in the case of the number mode (step S22), when the number of conversion source channels sandwiched between the reproduction positions of two conversion destination channels is n, the denominator is n+1 and the numerator is the distribution destination conversion channel. Let k be the order of the conversion source channels counted from the conversion destination channel on the reverse side, and the square root of this fraction √(k/(n+1)) be the distribution coefficient.

FIG. 13 is a diagram for explaining an example of how to obtain channel distribution coefficients in the case of the number mode. As shown in FIG. 13(a), when the number of source channels sandwiched between the playback positions of two destination channels O ₁ and O ₂ is one, the source channel I ₁ is the destination channel O ₁ , O ₂ are each allocated with a partition coefficient of √(1/2)=0.707.

Here, if the conversion source channel I ₁ is extremely biased toward either one of the conversion destination channels O ₁ and O ₂ , 1 is between the conversion source channel I ₁ and the conversion destination channel that is not biased. By adding 1 or more dummy channels, the coefficients may be the same as when there are 2 or more conversion source channels.

As shown in FIG. 13(b), when the number of source channels sandwiched between the playback positions of the two destination channels O ₁ and O ₂ is two, the source channel I ₃ is the destination channel O ₁ , are allocated to O ₂ by factors of √(2/3)=0.816 (-1.76 dB) and √(1/3)=0.577 (-4.77 dB), respectively. Note that when the positions of the conversion destination channel _I2 and the conversion source channel O1 _match , as in the case of the reproduction position of the conversion destination channel I2 and the conversion source channel O1, the distribution coefficient is set to 1.00, and the conversion Allocate to the destination channel.

In either case, the coefficients are defined so that the sum of the squares of the distribution coefficients of the acoustic signals is 1 in order to conserve energy. Correction coefficient for total energy due to the angle of conversion, correction coefficient for total energy due to different layers of channel playback positions of source multichannel sound signal and destination multichannel sound signal, source multichannel sound signal and destination multichannel sound At least one of the total energy correction coefficient based on the difference between the signal reproduction positions before and after the signal reproduction position and the total energy correction coefficient based on the similarity between the original multi-channel acoustic signals may be multiplied. The similarity between the conversion source multi-channel acoustic signals in this embodiment is the correlation between the conversion source channel A acoustic signal and the conversion source channel B acoustic signal distributed to the same conversion destination channel C. The sum of the energies of the A and B acoustic signals is multiplied by a correction coefficient so that the energies of the A and B acoustic signals included in the converted C acoustic signal are equal.

For example, depending on the playback positions of the conversion destination channels, such as when the playback positions of the conversion destination channels are close to each other, the listening impression may be preserved if the energy is not preserved. Therefore, the correction coefficient is -1.5 dB (coefficient of 0.841) when the conversion source channels are equally distributed, and -0.75 dB (coefficient of 0.917) when distributed largely to one side. can be multiplied by At this time, the distribution coefficient for equal distribution is −4.5 dB (coefficient 0.595 (0.707×0.841)), and when a large allocation is made to conversion destination channel O ₁ , −2 for each of O ₁ and O ₂ . 0.5 dB (coefficient 0.749 (0.816×0.917)) and −5.5 dB (coefficient 0.530 (0.577×0.917)). In addition, when the conversion source channels in the upper and lower layers are reproduced by the conversion destination channels in the middle layer, when the rear conversion source channels are reproduced by the front conversion destination channels, the total energy is -1.5 dB or -3. A correction coefficient such as 0 dB may be multiplied. Also, when integrating multiple conversion source channels that are likely to contain highly correlated signals such as LFE (low frequency effect channel) into the same conversion destination channel, -3.0 dB, -4.5 dB, - A correction factor such as 6.0 dB may be multiplied.

Finally, as shown in FIG. 2, when each channel number conversion coefficient is a numerical value included in a certain range, rounding is performed to replace it with a representative value (step S23). The output signal generator 15 performs channel number conversion processing using the channel number conversion coefficient matrix composed of the channel number conversion coefficients thus derived. A channel number conversion coefficient is obtained by multiplying or rounding a distribution coefficient by the correction coefficient described above.

FIG. 14 shows an example of replacement so that the distribution coefficients of the acoustic signal are expressed in decibels so as to give a clear number. For example, √(1/2)=√(3/6)=0.707 is -3.0 dB when rounded in 0.1 dB steps, and √(1/4)=0.500 is -6.0 dB. corresponds to By replacing with the representative value in this way, accurate directional reproduction is lost, but the calculation cost is reduced, and versatility is increased by using the same numerical value even if the reproduction position of the conversion destination channel is slightly different.

Below, from 22.2ch sound (middle layer 10ch, upper layer 8ch, ceiling 1ch, lower layer 3ch), 11.1ch (middle layer 7ch, upper layer 4ch), 9.1ch (middle layer 5ch, upper layer 4ch), 7.1ch sound (middle layer 5ch) , upper layer 2 ch). An example of 22.2ch audio playback positions is as shown in Table 1, and the channel order of conversion source channels is as shown in Equation (5) (M=24).

Tables 2, 3 and 4 show examples of the playback position of the destination channel. Since there is no upper layer rear channel for 7.1ch sound, the middle layer rear channel is used. Also, since there are no ceiling channels in any of the destination channel arrangements, the four corner channels on the top layer are used. Also, since there is no lower layer channel, the middle layer channel is used as it is for reproducing the lower layer.

The conversion destination channel order of 7.1ch sound is as shown in formula (6) (N=8), the conversion destination channel order of 9.1ch sound is as shown in formula (7) (N=10), The conversion destination channel order of 11.1ch sound is as shown in Equation (8) (N=12).

Next, the reproduction position of the conversion source channel is converted. When the playback position of the conversion source channel is used as it is, it is as shown in Table 1 above. Here, an example of changing the playback position in accordance with the reference point will be described. Here, consider an example in which the positions of four speakers are matched.

Table 5 shows an example of converting 22.2ch sound to 7.1ch (5.1+2ch). There are cases where the middle layer of 60 degrees of the conversion source is converted to the conversion destination of 30 degrees, and the middle layer of 30 degrees is converted to the conversion destination of 30 degrees. Also, the upper 45 degrees of the conversion source may be converted to the conversion destination of 30 degrees, 45 degrees, and 90 degrees. Since there is no conversion destination channel behind the upper layer, the reproduction position 110 degrees of the conversion destination channel behind the middle layer is used. However, if the upper layer channel of the conversion destination is 90 degrees to the side, the playback position of 90 degrees is used for both forward and backward.

Table 6 shows an example of converting 22.2ch sound to 9.1ch (5.1+4ch). There are cases where the middle layer of 60 degrees of the conversion source is converted to the conversion destination of 30 degrees, and the middle layer of 30 degrees is converted to the conversion destination of 30 degrees. Further, when the upper layer of 45 degrees of the conversion source is converted to the destination of 30 degrees and 45 degrees, the upper layer of 135 degrees of the conversion source may be converted to the upper layers of 110 degrees and 135 degrees of the conversion destination.

Table 7 shows an example of converting 22.2ch sound to 11.1ch (7.1+4ch). There are cases where the middle layer of 60 degrees of the conversion source is converted to the conversion destination of 30 degrees, and the middle layer of 30 degrees is converted to the conversion destination of 30 degrees. Further, when the upper layer of 45 degrees of the conversion source is converted to the destination of 30 degrees and 45 degrees, the upper layer of 135 degrees of the conversion source may be converted to the upper layers of 110 degrees and 135 degrees of the conversion destination.

Table 8 shows an example of conversion from 22.2 ch sound to 7.1 ch sound when 6 points are used as reference points with 90 degrees on the side as reference points in addition to the four reference points in all four directions. Table 9 shows an example of conversion from 22.2ch sound to 9.1ch sound, and Table 10 shows an example of conversion from 22.2ch sound to 11.1ch sound.

Next, each element of the channel number conversion coefficient matrix for absolute position and relative position is derived based on the number of conversion source channels and based on the angle between the reproduction position of the conversion destination channel and the reproduction position of the conversion source channel. Here, the angle-based coefficient is an example calculated using the Tan rule.

Tables 11 and 12 show examples of channel number conversion coefficients when converting from 22.2ch sound to 7.1ch sound. Here, the upper part of each column is the channel number of the 22.2 ch sound that is the conversion source, and the left end of each row is the number of the 7.1 ch sound that is the conversion destination.

Tables 13 to 16 show examples of channel number conversion coefficients when converting from 22.2ch sound to 7.1ch sound. Here, O means absolute position, R1 means relative position 30 degrees to 30 degrees, 135 degrees to 110 degrees, R2 means relative position 30 degrees to 30 degrees, 90 degrees fixed, 135 degrees R3 means relative position 60 degrees to 30 degrees, 135 degrees to 110 degrees, R4 means relative position 60 degrees to 30 degrees, 90 degrees fixed, 135 degrees to 110 degrees means. Also, for the upper layer, _A means 30 degrees and 110 degrees, OA means absolute position, _RA 1 means relative position 45 degrees to 30 degrees, 135 degrees to 110 degrees, and _RA 2 means relative position 45 degrees to 30 degrees, 90 degrees fixed, 135 degrees to 110 degrees, B means 45 degrees and 110 degrees, O _B absolute position, R _B 1 relative position 45 degrees To 45 degrees, 135 degrees to 110 degrees, R _B 2 means relative position 45 degrees to 45 degrees, 90 degrees fixed, 135 degrees to 110 degrees, C means 90 degrees, O _C means an absolute position, R _C 1 means a relative position of 45 degrees to 90 degrees and 135 degrees to 90 degrees, N means the number of conversion source channels, and P means a reproduction position. Also, for example, "O, N" means that when determining the channel number conversion coefficient, the positions of the conversion source and the conversion destination are determined by absolute positions, and the distribution coefficient is determined by the number of channels of the conversion source. .

Table 17 shows an example of channel number conversion coefficients when converting from 22.2ch sound to 9.1ch sound. The upper part of each column is the channel number of the 22.2 ch sound that is the conversion source, and the left end of each row is the conversion destination.

The channel number conversion coefficients from middle and lower layers of 22.2 ch sound to 9.1 ch sound are the same as in the case of 7.1 ch (Tables 13 and 16). Table 18 shows an example of channel number conversion coefficients when converting from the upper layer of 22.2ch sound to 9.1ch sound. where A means 30 degrees and 110 degrees, O _A means absolute position, R _A 1 means relative position 45 degrees to 30 degrees, 135 degrees to 110 degrees, and R _A 2 is Relative position 45 degrees to 30 degrees, 90 degrees fixed, 135 degrees to 110 degrees, B means 45 degrees and 110 degrees, O _B means absolute position, R _B 1 means relative position 45 45 degrees to 45 degrees, 135 degrees to 110 degrees, R _B 2 means relative position 45 degrees to 45 degrees, 90 degrees fixed, 135 degrees to 110 degrees, D is 45 degrees and 135 degrees OD means absolute position=relative position 45 degrees to 45 degrees, 135 degrees to 135 degrees, N means the number of conversion source channels, and _P means the playback position.

Table 19 shows an example of channel number conversion coefficients from 22.2ch sound to 11.1ch sound. Here, the upper part of each column is the channel number of the 22.2 ch sound that is the conversion source, and the left end of each row is the number of the 11.1 ch sound that is the conversion destination.

The channel number conversion coefficient from the upper layer of 22.2ch sound to 11.1ch sound is the same as in the case of 9.1ch (Table 18). Here, O means absolute position, R5 means relative position 30 degrees to 30 degrees, R6 means relative position 30 degrees to 30 degrees, 90 degrees fixed, R7 means relative position 60 degrees 30 degrees, R8 means that the relative position is 60 degrees to 30 degrees and fixed at 90 degrees, N means the number of conversion source channels, and P means the playback position. The subscript _Bt means that only the lower layer (bottom layer) is extracted.

When performing a channel number transform, simplification of coefficients may be desired rather than preserving precise angles. Therefore, it is possible to simplify the conversion of the number of channels by replacing coefficients in a certain range with representative values. Therefore, Table 22 shows an example of deriving channel number conversion coefficients step by step in units of 0.5 dB. When multiplied by the channel number conversion factor in units of 0.5 dB (top), the channel number conversion factor that needs to be multiplied to conserve energy in the corresponding channel is shown (bottom).

Table 23 shows channel number conversion coefficients when a representative value is specified with a good numerical value such as 1.5 dB. This numerical value is close to the numerical value obtained by quantizing the representative value in the central five steps when a certain space is divided into seven.

When the representative values of the channel number conversion coefficients in units of 1.5 dB are specified, the values of the channel number conversion coefficients shown in Tables 13-16, 18, 20 and 21 are as shown in Tables 24-30, respectively.

Here, an example of five levels of representative values is shown, but the representative values may be subdivided as shown in Table 31.

Also, in this embodiment, the signal is distributed so as to conserve energy, but when the acoustic signal is distributed, it may be perceived as being louder than the original signal, so the energy may be reduced. For example, assume that the energy is -1.5 dB when distributed equally, and returns to the original energy when approaching one side. In that case, the representative values of Table 23 are as shown in Table 32, for example.

(Example 1)
Next, in FIG. 2, select step S13 for dividing each divided layer of the conversion destination channel into four ranges of front, rear, left, and right. Select step S18 (relative position mode) for arranging so that the relative relationships match, and further select step S22 (number mode) for prescribing coefficients according to the number of conversion source channels sandwiched between the reproduction positions of two conversion destination channels. The operation of the sound processing device 1 according to the first embodiment will be described by taking the case of the first embodiment as the first embodiment.

FIG. 15 is a flow chart showing a procedure for deriving channel number conversion coefficients by the sound processing device 1 according to the first embodiment.

First, the number of channels and speaker arrangement (playback position) of the conversion source and the conversion destination are compared (step S101), and it is determined whether or not there is a transmitted channel number conversion coefficient (step S102). If there is a transmitted channel number conversion coefficient, this channel number conversion coefficient is used, so the process ends. Next, it is determined whether or not there is a channel number conversion coefficient value stored inside the system of the sound processing device 1 (step S103). This process is also terminated if there are channel number transform coefficient values stored inside the system.

That is, the channel number conversion coefficient is used when there is no other coefficient value obtained in advance, and the channel number conversion coefficient received together with the conversion source multichannel sound signal, the internal If there are channel number conversion coefficients stored in , their values are used for the channel number conversion.

To derive the channel number conversion coefficient, first, the reproduction position of the conversion destination channel is divided into a plurality of layers in the elevation direction (step S104). For example, for 22.2 ch sound, there are four layers: lower layer, middle layer, upper layer, and overhead (elevation angle 90 degrees). In audio mode 2/0/2-3/0/2) and 11.1 ch audio (audio mode 2/0/2-3/2/2), there are two layers, the middle layer and the upper layer.

After that, each divided layer of the conversion source channel is divided into four ranges in the front, rear, left, and right (cross plane, median plane) (step S105). Here, if the reproduction position of the conversion destination channel does not exist in the four-divided range, the reproduction position of the conversion destination channel in the adjacent layer is determined as the reproduction position of the layer (step S106). However, a layer composed of only one point such as overhead (90 degrees elevation angle) is excluded. For example, in 7.1ch sound, since there is no channel in the upper rear layer of the conversion destination channel arrangement, the rear channel in the middle layer is also used for reproduction of the upper layer. Also, since 7.1ch sound A, 9.1ch sound, and 11.1ch sound do not have a lower layer, if the conversion source channel has a lower layer channel, the middle layer is used for reproduction of those channels.

Next, the reproduction position of the conversion source channel is arranged on each divided layer of the conversion destination channel (step S107).

For a layer with only one point such as overhead (90 degrees elevation angle), if there is no playback position in the conversion destination channel, the four corners (azimuth angles 45 degrees and 135 degrees) of the four conversion destination channels in the front, back, left and right One point for each reproduction position of the conversion destination channel is used as a reference point for the conversion destination channel (step S108). For example, TpC located above the 22.2ch sound (90 degrees elevation angle) does not have a reproduction position in the conversion destination 7.1ch sound A, 9.1ch sound, or 11.1ch sound. Therefore, the playback positions of the conversion destination channels near the four corners (azimuth angles of 45 degrees and 135 degrees) of the conversion destination are used. The sound A is distributed to two channels in the upper layer and two channels in the rear middle layer.

As for the method of determining the reference point, first, if a channel exists at the same position of the conversion source and the conversion destination, the channel is determined as the reference point (step S109). For example, when converting from 22.2 ch sound to 7.1 ch sound A or 9.1 ch sound, the 7.1 ch sound A and 9.1 ch sound, which are conversion destinations, are converted to azimuth angles of 0 degrees and ±30 degrees in the middle layer. A certain channel is defined as a reference point. Since there are no reference points in the upper 4 ranges and the lower 2 ranges in the middle layer, the remaining reference points of the conversion source and conversion destination channels are determined.

If there is no channel at the same position in each of the four regions of each divided layer except for the channels in the transverse plane and the median plane, the front two ranges are the front (0 < playback position ≤ 60 degrees) and the rear For range 2, a channel close to the side (90 degrees<playback position≦135 degrees) is set as a reference point (step S109). For example, in the source 22.2 ch sound, the reference point is the channel at the position of 45 degrees in the front of the upper layer and the channel at the position of 135 degrees in the rear. On the other hand, in the conversion destination channel, the reference point for the upper front two ranges (0 degree < playback position ≤ 60 degrees, the point closest to 45 degrees) is the position of 30 degrees, and the reference point for the upper and middle layers rear two ranges ( The point closest to 135 degrees (90 degrees<playback position≦135 degrees) corresponds to the channel at the position of 110 degrees.

If a plurality of channels exist in the same proximity such as 30 degrees and 60 degrees to 45 degrees, the channel with the smaller original angle (30 degrees) is set as the reference point (step S109). It should be noted that the method of taking the reference points of the front as the front and the side as the rear is based on the characteristics of human hearing, and the concept of 5.1ch sound and 22.2ch sound, in which the speaker placement was determined in consideration of these characteristics. Based on

When determining the reference point, if there is forward reference point designation information, the channels corresponding to the reference points in the two front ranges are switched from the reference point obtained in step S109 to another reference point (step S110). For example, if the information indicated by the channels corresponding to both ends of the screen is used as the front reference point designation information, the channels corresponding to both ends of the conversion source and conversion destination screens are used as the reference points of the two front ranges, respectively, regardless of the position information. replace the points. When the conversion source is 22.2 ch sound, it is mainly considered that the channel of ±60 degrees in the middle layer is desired to be used as the reference point.

When the reference point is determined, the reproduction position of the conversion-source reference point is changed so that the reproduction positions of the determined conversion-source and conversion-destination reference points match (step S111). For example, the playback positions of the channels that are the reference points of the 22.2ch sound are changed from 45 degrees to 30 degrees and from 135 degrees to 110 degrees. If the reference point in the middle layer front of the 22.2 ch sound, which is the conversion source, has been changed to a channel of ±60 degrees by the front reference point switching information, 60 degrees is changed to 30 degrees, and the total of 30 degrees of the 22.2 ch sound is changed. Change the playback position of the channel that was at the same time to 15 degrees.

In addition, the reference point was changed so that the original distance ratio between the speakers was maintained even for the conversion source channel that did not become the reference point in the four ranges in each divided layer except for the channel on the cross section and the median plane. Similarly change the playback position according to the ratio. All channels of the 7.1ch audio A, 9.1ch audio, and 11.1ch audio of the conversion destination are reference points, and there is no channel whose reproduction position is changed in this step. In this way, in most of the currently proposed formats, almost all channels serve as reference points, and it is thought that there are few cases where the playback position is changed at this stage.

Next, among the four regions within each divided layer excluding the channels of the transverse plane and median plane, the reproduction position of the conversion source channel with the smallest angle difference is changed (locked) for each of the reproduction positions of the conversion destination channel (step S112). Since the playback positions of the channels set as the reference points already match, the playback positions are changed here only for the conversion source channels that are within the four ranges and that did not become the reference points. .

After the channel to be represented by the virtual sound source is determined, specific coefficient values are specified. In order to increase the identifiability of each channel, the coefficient values are determined according to the number of conversion source channels existing between the conversion destination channels to be reproduced, and are evenly re-arranged between the conversion destination channels to be reproduced (step S113). If the coefficients are defined so that the sum of the squares of the distribution coefficients of the acoustic signals is 1 so that the energy of the source channel and the energy of the destination channel are matched, the coefficients are as shown in Table 33 below.

The antilogarithmic value obtained by this method may be rounded so that it becomes a good value when expressed in decibels (step S114). For example, when transmitting 22.2 ch audio signals in MPEG-4 AAC, values are used in increments of 1.5 dB, so 1/√2 is -3 dB, 1/√3 is -4.5 dB, 3) rounds to -1.5 dB. Depending on the degree of rounding when expressed in decibels, it is conceivable that the exact numerical value obtained by the calculation of this method or the value of the rounding example shown in the example may not match completely, but within a range of about ±0.05, for example. A certain amount of deviation in numerical values such as is included in the relationship defined in this formula.

Next, the audio mode is multi-channel stereo (3/3/3-5/2/3-3/0/0.2), and in multi-channel stereo beyond multi-channel stereo (3/2.1) Tables 35 to 46 show the values of the conversion coefficients for the number of channels to be multiplied by the decoded multi-channel stereo sound signal for output, derived according to the procedure shown in FIG. Table 34 shows the values that fall within symbols a through j in Tables 35-46. dmix_pos_adj_idx is an example of the front reference point designation information described above, and Table 34 shows changes in channel number conversion coefficients due to switching of reference points according to this front reference point designation information. Use a value similar to dmix_pos_adj_idx=0 until dmix_pos_adj_idx is sent for the first time.

Table 35 shows multi-channel stereo (2/0/0-3/0/2-0.1), sound system C according to ITU-R BS.2051, Channel configuration 14 according to ISO/IEC 23008-3 This is the channel number conversion coefficient when outputting with .

Table 36 shows channel number conversion when outputting with multichannel stereo (3/2/2.1), sound system I described in ITU-R BS.2051, and Channel configuration 12 described in ISO/IEC 23008-3. is the coefficient.

Table 37 shows channel number conversion coefficients when outputting in multichannel stereo (3/3.1), Channel configuration 11 described in ISO/IEC 23008-3.

Table 38 shows channel number conversion coefficients when outputting in multichannel stereo (5/2.1), Channel configuration 7 described in ISO/IEC 23008-3.

Table 39 shows channel number conversion coefficients when outputting with sound system D described in ITU-R Recommendation BS.2051 and Channel configuration 16 described in ISO/IEC 23008-3.

Table 40 shows channel number conversion coefficients when outputting with sound system E described in ITU-R recommendation BS.2051.

Table 41 shows channel number conversion coefficients when outputting with sound system F described in ITU-R Recommendation BS.2051 and Channel configuration 15 described in ISO/IEC 23008-3.

Table 42 shows channel number conversion when outputting with sound system G described in ITU-R recommendation BS.2051 and Channel configuration 20 described in ISO/IEC 23008-3 (when the display is inside LR) is the coefficient.

Table 43 shows channel number conversion coefficients when outputting with sound system J described in ITU-R Recommendation BS.2051 and Channel configuration 19 described in ISO/IEC 23008-3.

Table 44 shows channel number conversion coefficients when outputting with Channel configuration 17 described in ISO/IEC 23008-3.

Table 45 shows channel number conversion coefficients when outputting with Channel configuration 18 described in ISO/IEC 23008-3.

Table 46 shows channel number conversion coefficients when outputting in multichannel stereo (2/0/0-3/2/2-0.1).

(Example 2)
Next, in FIG. 2, select step S13 for dividing each divided layer of the conversion destination channel into four ranges of front, back, left, and right, specify FL/FR as the reference point, and then Select step S18 (relative position mode) for changing the reproduction position of the conversion source channel, which is located at the most biased position to the target conversion destination channel other than the channel, to the same position as the conversion destination channel (relative position mode), and select two conversion destinations. The operation of the sound processing apparatus 1 according to the second embodiment will be described by taking as the second embodiment a case where step S22 (number mode) in which coefficients are defined by the number of conversion source channels sandwiched between channel reproduction positions is selected.

FIG. 16 is a flow chart showing a procedure for deriving channel number conversion coefficients by the sound processing device 1 according to the second embodiment. Steps S201-208, 212, and 213 are the same as steps S101-108, 113, and 114 described in the first embodiment, respectively, so description thereof will be omitted, and only steps S209-211 will be described.

In step S209, if there is front reference point designation information, the front end reference points are designated by this information. In step S210, the reproduction position of the conversion-source reference point is changed so that the reproduction positions of the conversion-source and conversion-destination reference points match. In other words, the playback positions of the front end channels of the conversion source are changed to the playback positions of the channels at the front ends of the conversion destination. change. For example, when converting from 22.2ch sound to 7.1ch sound, both front ends of 22.2ch sound are FL and FR at azimuth angles of ±60 degrees, and 7.1ch sound is L and R at ±30 degrees. is. In this case, the playback positions of FL and FR are changed to 30 degrees, and the playback positions of FLc and FRc between them are changed while maintaining the distance ratio, and repositioned to the position of 15 degrees.

In step S211, the reproduction position of the conversion-source channel that is most biased in angle ratio to the reproduction position of the conversion-destination channel in each divided layer is changed (locked). For example, when converting from 22.2ch sound to 7.1ch sound (5.1+2ch) or 9.1ch sound (5.1+4ch), in the conversion destination 7.1ch sound and 9.1ch sound, A channel is located at a position of 110 degrees with the front of the listener being 0 degrees, and the 22.2ch sound that is the conversion source has channels at a position of 90 degrees and 135 degrees in the vicinity thereof. Here, it is determined which of the 90-degree and 135-degree channels of the conversion source is to be changed to the 110-degree position of the conversion destination. The two destination channels adjacent to the source channel positioned at 90 degrees are positioned at 30 degrees and 110 degrees, and the angles formed with the source channel are 60 degrees and 20 degrees, respectively. The ratio is 3:1, and it can be said that the 90-degree source channel is biased by 3:1 between the 30-degree and 110-degree destination channels. Similarly, the 135-degree channel can be said to be skewed 4.8:1 between the adjacent 110-degree and 250-degree destination channels. As a result, the 135-degree channel is determined as the conversion-source channel whose reproduction position is to be changed because the conversion destination channel is biased toward the 110-degree conversion destination channel. At this time, the 90-degree speaker is physically closer to 110 degrees. A virtual sound source at a position that is greatly biased in the conversion destination channel, which is wide open at 135 degrees, is expressed, and the sound image and localization of the 135-degree channel can be greatly degraded. The reason why the channel with less bias is selected as the channel for changing the reproduction position is to suppress deterioration of the overall impression of each channel.

When the audio mode is multi-channel stereo (3/3/3-5/2/3-3/0/0.2) and output in multi-channel stereo exceeding multi-channel stereo (3/2.1) , the value obtained by deriving the number-of-channels conversion coefficient by which the decoded multichannel stereo sound signal is multiplied by the procedure shown in FIG. It is often the same as the value.

When outputting in multi-channel stereo (2/0/0-3/0/2-0.1), sound system C according to ITU-R BS.2051, and Channel configuration 14 according to ISO/IEC 23008-3 is as shown in Table 35. When outputting with multichannel stereo (3/2/2.1), sound system I according to ITU-R BS.2051, and Channel configuration 12 according to ISO/IEC 23008-3, the conversion factors for the number of channels are shown in the table below. 36. Table 38 shows the channel number conversion coefficients when outputting in multichannel stereo (5/2.1) and Channel configuration 7 described in ISO/IEC 23008-3. Table 39 shows the channel number conversion coefficients when outputting with sound system D described in ITU-R Recommendation BS.2051 and Channel configuration 16 described in ISO/IEC 23008-3. Table 40 shows the channel number conversion coefficients when outputting with sound system E described in ITU-R Recommendation BS.2051. Table 41 shows the channel number conversion coefficients when outputting with sound system F described in ITU-R Recommendation BS.2051 and Channel configuration 15 described in ISO/IEC 23008-3. The channel number conversion factor when outputting with sound system G described in ITU-R recommendation BS.2051 and Channel configuration 20 described in ISO/IEC 23008-3 (when the display is inside LR) is shown in the table below. 42. Table 43 shows the channel number conversion coefficients when outputting with sound system J described in ITU-R Recommendation BS.2051 and Channel configuration 19 described in ISO/IEC 23008-3. Table 44 shows the channel number conversion coefficients when outputting with Channel configuration 17 described in ISO/IEC 23008-3. Table 46 shows the channel number conversion coefficients when outputting in multichannel stereo (2/0/0-3/2/2-0.1).

Table 47 shows the channel number conversion coefficients when outputting in multichannel stereo (3/3.1), Channel configuration 11 described in ISO/IEC 23008-3, because they differ from those in Example 1. The values entered into the symbols in Table 47 are as shown in Table 34.

Table 48 also shows the channel number conversion coefficients when outputting in Channel configuration 18 described in ISO/IEC 23008-3, because they differ from those in the first embodiment. The values entered for the symbols in Table 48 are as shown in Table 34.

A computer can be suitably used to function as the sound processing device 1 described above. , and the program is read and executed by the CPU of the computer. This program can be recorded on a computer-readable recording medium.

Also, the program may be recorded on a computer-readable medium. It can be installed on a computer using a computer readable medium. Here, the computer-readable medium on which the program is recorded may be a non-transitory recording medium. The non-transitory recording medium is not particularly limited, but may be, for example, a recording medium such as CD-ROM or DVD-ROM.

The method of deriving the number of channels conversion coefficients of the present invention has the following effects. First, it can correspond to any number of channels and any channel arrangement. In other words, in this derivation method, by appropriately changing the playback position of the conversion source channel, the listening direction of the characteristic channels in the left and right front and left and right rear of the conversion source channel is specified, and then by using the virtual sound source It is possible to suppress the deterioration of tone and sound image, and reduce the change in impression as a whole.

Secondly, it is possible to suppress the deterioration of the impression (timbre) of the conversion source format. In other words, if there is a conversion destination channel that does not have a conversion source channel with a matching playback position at a position other than the front left, right, or rear left, right, the playback position of the conversion source channel is moved to the playback position of the nearest conversion destination channel. By changing it, it is possible to suppress deterioration of the overall impression due to an increase in the number of virtual sound sources.

Thirdly, when the relative position mode is adopted and the reference points at the four corners are designated, the combination of the reference front left, right, rear left, and right channels can be maintained. In other words, the front left, right, rear left, right channels of the conversion source and conversion destination channels are adjusted so that the combination of the front left, right, rear left, and right channels, which are the main channels of each format, does not change due to the difference in speaker density between the conversion source and conversion destination formats. , and based on this reference point, the reproduction position of the conversion source channel can be changed. Also, the channels corresponding to the reference points of the front two ranges may be switchable. This makes it possible to select the playback position according to the viewing environment for the channel in front of the listener, which is the main channel in many formats.

Fourthly, by adopting the relative position mode, the playback position of the conversion-source channel that exists at the most biased position in the target conversion-destination channel other than the conversion-source/conversion-destination channel corresponding to the reference point will be the same as the conversion-destination channel. When the position is changed, it is possible to suppress the blurring of the sound image and the accompanying deterioration of the sense of direction that may occur when the virtual sound source is placed at a position biased to either side between the wide-open speakers.

Fifthly, the number of channels can be converted without greatly impairing the expression/performance of the conversion source format. In other words, among the conversion source channels at the same height (layer), only one channel is changed to the playback position of each conversion destination channel. It is possible to avoid greatly impairing the expression and performance unique to sound. Further, by evenly rearranging the conversion source channels represented by the virtual sound sources between the two conversion destination channels sandwiching the reproduction position thereof, it becomes easy to identify the direction of each conversion source channel.

Sixth, it is possible to use numerical values that are easy to use for the channel number conversion coefficients. In other words, when actually using the channel number conversion coefficients calculated by this method, the distribution coefficients derived from its simplicity can be replaced in advance so that they are decibel-expressed values.

Although the above embodiments have been described as representative examples, it will be apparent to those skilled in the art that many modifications and substitutions may be made within the spirit and scope of the invention. Therefore, the present invention should not be construed as limited by the embodiments described above, and various modifications and changes are possible without departing from the scope of the appended claims. For example, it is possible to combine a plurality of configuration blocks described in the configuration diagrams of the embodiments into one, or divide one configuration block.

REFERENCE SIGNS LIST 1 sound processing device 11 conversion target channel division unit 12 conversion source channel reproduction position determination unit 13 channel number conversion coefficient determination unit 14 channel number conversion coefficient matrix storage unit 15 output signal generation unit

Claims (10)

An audio processing device that converts the number of channels and channel arrangement of a multichannel audio signal,
a conversion destination channel division unit that divides the reproduction position of the conversion destination multi-channel acoustic signal into multiple layers in the elevation direction;
a source channel playback position determination unit that determines a playback position of the source multichannel audio signal on each divided layer divided by the destination channel division unit;
A channel number conversion factor for determining a channel number conversion factor based on the reproduction position of the conversion destination multichannel acoustic signal and the reproduction position of the conversion source multichannel sound signal determined by the conversion source channel reproduction position determination unit. a decision unit;
Output signal generation for generating the destination multi-channel acoustic signal from the source multi-channel acoustic signal using a channel number conversion coefficient matrix having each element the channel number conversion coefficient determined by the channel count conversion coefficient determination unit. Department and
A sound processing device comprising: The conversion destination channel dividing unit has at least elevation angles of feet (-90 degrees of elevation), floor layers (-90 to -45 degrees of elevation), lower layers (-45 to 0 degrees of elevation), and middle layers (0 degrees of elevation). , upper layer (0 to 45 degrees elevation), ceiling layer (45 to 90 degrees elevation), and overhead (90 degrees elevation). 2. The sound processing apparatus according to claim 1, wherein a reproduction position of said converted multi-channel sound signal is divided. The conversion destination channel division unit divides the reproduction position of the conversion destination multi-channel acoustic signal into a plurality of layers in the elevation direction, and further divides each divided layer into a plurality of ranges in the azimuth direction. , The sound processing device according to claim 1 or 2. The conversion destination channel dividing unit divides the reproduction position of the conversion destination multi-channel acoustic signal into a plurality of ranges in the azimuth direction, and divides the reproduction position missing range in which the reproduction position of the conversion destination multi-channel acoustic signal does not exist. For the playback position missing layer including 4. The sound processing device according to claim 3, wherein the reproduction position is set to . The conversion-source channel reproduction position determining unit adjusts two or more reference points of the reproduction position of the conversion-source multi-channel acoustic signal and two or more reference points of the reproduction position of the conversion-destination multi-channel acoustic signal to match each other. , rotating and scaling the surround circle of the conversion-source multi-channel sound signal, and rearranging the reproduction position of the conversion-source multi-channel sound signal existing between the reference points on the surround circle of the conversion-destination multi-channel sound signal; The sound processing device according to any one of claims 1 to 4, characterized in that: The conversion destination channel dividing unit divides each divided layer into four ranges of front, back, left, and right,
The conversion source channel reproduction position determining unit sets the reference points of the reproduction position of the conversion source multi-channel audio signal and the reproduction position of the conversion destination channel to the front left and front right corresponding to both ends of the screen divided by the conversion destination channel dividing unit. 2 points, 4 points at the four corners of the 4 ranges divided into 4, left front, left rear, right front, and right rear, and 6 points obtained by adding 90 degrees to the left and right sides to the 4 points. 6. The sound processing device according to claim 5, wherein the reproduction position of the original multi-channel sound signal is set. The conversion-source channel playback position determination unit sets the reference point to a predetermined position, a position specified by a user, a position specified by metadata, or a similar name or position of the conversion-source channel and the conversion-destination channel. 7. The sound processing device according to claim 5 or 6, characterized in that the decision is based on degrees. The number-of-channels conversion coefficient determination unit determines the reproduction positions of two adjacent destination multi-channel acoustic signals and the source multi-channel sound signals sandwiched between the reproduction positions determined by the source channel reproduction position determination unit. 8. The sound processing device according to any one of claims 1 to 7, wherein said number-of-channels conversion coefficient is calculated based on the number of reproduction positions of . 9. The channel number conversion coefficient determining unit replaces the calculated channel number conversion coefficient with a predetermined representative value when the calculated channel number conversion coefficient is included in a predetermined range. The acoustic processing device according to the item. A program for causing a computer to function as the sound processing device according to any one of claims 1 to 9.

Images