JP2006157210A - Multichannel sound field processing apparatus - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a multichannel sound field processing apparatus for providing the presence of a hall space adaptively to actual speaker arrangement. <P>SOLUTION: The multichannel sound field processing apparatus 20 configured such that the processing apparatus 20 obtains sound field processing coefficient data Dist1, Dist2, Gain1, Gain 2, and Delay from a multichannel audio input signal with a distribution of an arrival direction, an amplitude and a delay time of an initial reflecting sound in response to a speaker arrangement angle on the basis of acoustic characteristic data obtained by simulating a geometrical response of the amplitude, the angle and the arrival time of the initial reflecting sound by using a prescribed hall space for an object by means of a prescribed algorithm, and a digital signal processing apparatus DAP carries out multichannel sound field processing and includes a sound field processing coefficient data calculation means that calculates the sound field processing coefficient data by means of the prescribed algorithm on the basis of sound characteristic data stored in advance in a memory device 7 and positional information of actual speakers of each channel received externally at the start of the sound field processing by the DAP. <P>COPYRIGHT: (C)2006,JPO&NCIPI

Description

  The present invention relates to a multi-channel sound field processing technique, and in particular, various dimensions that exist in a user's actual audio reproduction space for multi-channel audio signals of three or more channels such as DVD videos in which movies are recorded. The present invention relates to a multi-channel sound field processing apparatus that generates a multi-channel audio signal that has been subjected to sound field processing so as to give a realistic feeling to a shaped hole.

  In order to reproduce the acoustic characteristics that give a sense of presence as if listening to a concert hall while staying at home from a two-channel audio input signal based on CD (Compact Disc) or FM broadcasting, etc. A so-called digital acoustic processor (hereinafter abbreviated as Digital Acoustic Processor, abbreviated as DAP) for adding a sound signal simulating a direct sound, reflected sound and reverberation sound to the sound input signal and reproducing it in a multi-channel is referred to as a high-grade audio device. (The DAP is a kind of digital signal processor (DSP)).

  Recently, the audio output of packaged media such as movies typified by DVD has become multi-channel, and multi-channel compatible with multi-channel audio input signals of more than 3 channels such as 5.1ch or 7.1ch. The sound field processing device has been put to practical use.

  FIG. 4 is a block diagram for explaining a signal generation algorithm of the sound field processing apparatus 10 corresponding to the multi-channel audio input signal (in the case of 7.1 channel). 5 (a), 5 (b), and 5 (c), the ideal two channels (stereo), 5.1 channel, and 7.1 channel in the audio playback space on the user side (such as a living room at home) are shown. The speaker arrangement for the listener M is shown.

  In the 7.1-channel multi-channel sound field processing apparatus 10 of FIG. 4, 7.1 channels (L, R, and R) for multi-channel audio input signals (L, R, Sl, Sr, C, Bl, Br) by the DAP are used. Sl, Sr, C, LFE (subwoofer), Bl, Br) signal processing uses actual hole drawing data, and multi-channel sound sources are appropriately arranged at various locations in the hall (generally at the hall). Calculate the geometric response (impulse response) of the direct sound, early reflections, and reverberation to the listener at the center or slightly behind it by simulation to obtain the acoustic characteristic data of the hall. Field processing coefficient data is a predetermined algorithm on the premise of an ideal speaker arrangement in the audio reproduction space on the user side (see FIG. 5C). In stored all been calculated in the external memory device 9, the signal processing is performed by the control of the signal processing unit from CPU8 of the DAP based on this.

  That is, the sound coming into the ear of the listener M is roughly divided into three elements: (a) direct sound, (b) early reflection sound, and (c) reverberation sound, and independent calculation is performed for each channel's sound input. I'm doing it for the signal. The following is an overview.

  (A) Direct sound is a sound that reaches the listener M directly from each multi-channel sound source, and has a delay and an attenuation according to the distance D and direction. FIG. 6A shows the direction and distance of the sound ray projected on the horizontal plane of the direct sound from the 5.1 sound sources PA Lch, PA Rch, PA Slch, PA Srch, PA Cch, and PA LFE in the hall space H. (Delay) is shown, and (b) shows the relationship between the distance from the ideal 7.1ch speaker arrangement in the audio reproduction space on the user side and the direction of the sound ray.

  (B) The Early Reflection is a sound that reaches the listener M when sound waves bounce off the wall surfaces (the left and right wall surfaces, the front and back wall surfaces, and the upper and lower surfaces of the hall), and this is also the same as the direct sound. There is delay and attenuation. These sounds are called imaginary sounds because the listener M perceives that a sound source (imaginary sound source IMS) is present at a position symmetrical to the sound source with respect to the wall surface. As shown in the example of the primary reflected sound with PARch as a sound source in FIG. 7, this imaginary sound is imaginary to the listener M after being delayed by the distance of the path A + B and the air attenuation, or by the sound absorption of the wall surface. It is perceived that there is a sound source at the position of the sound source IMS. This primary reflection to tertiary reflection (reflected sound that arrives at the listener M after reflecting the wall three times) is called the initial reflected sound. After the fourth reflection, it is regarded as a sound that circulates around the wall, and the following reverberant sound: Perceived as attenuation of acoustic energy. FIG. 8 (a) shows an example of the sound ray of the primary reflected sound of the sound source of each channel reflected by the left and right wall surfaces in the hall space H, and FIG. 8 (b) shows an ideal speaker arrangement in the multi-channel audio reproduction space on the user side. And the direction of the sound ray direction of the primary reflected sound.

  (C) A reverberation sound is a sound in which sound emitted in space attenuates while reflecting off a wall surface. For reverberation signal processing, a reverberation sound generation filter called a reverberation filter that adds reverberation sound is used, and the reverberation sound signal generated by the reverberation sound generation filter is added to the audio signal. . The signal processing of the reverberation sound generation filter is a comb filter that can adjust the reverberation time, and is effective in increasing the density of reflected sound because all frequencies are in the passband. An algorithm devised by Schroeder that uses a combination with an allpass filter that is also effective in preventing the generation of colored sound is used (see [Patent Document 1] for details).

  Next, the calculation algorithm for obtaining the sound field processing coefficient data relating to the direct sound and the initial reflected sound will be described below.

  FIG. 4 shows an algorithm for generating the coefficient of the initial reflected sound pattern in the multi-channel DAP. Using the drawing data of several existing holes, the sound source is arranged in various places, and the geometry up to the listener is shown. Based on the acoustic characteristic data (initial reflected sound amplitude (energy), angle (angle), arrival time amplitude (del)) obtained by simulation of the response, Source (input signal arrival direction), Delay Sound field processing consisting of (arrival time of reflected sound (delay time)), Dist1, Dist2 (direction of reflected sound (distribution according to angle)), Gainl, Gain2 (amplitude of reflected sound (distribution according to angle)) Coefficient data is created and stored in the external memory device 9.

  At present, the sound field processing coefficient data (six bytes per reflected sound) of the initial reflected sound of several hundreds per hole are all stored in the external memory device 9 (nonvolatile memory IC such as ROM) outside the DAP. The sound field processing coefficient data for a predetermined hole designated by the user at the time of DAP activation is read.

  Here, considering how the sound reaches the listener M in the actual hall with respect to the sound from the sound source of the front PARch, first, the listener M receives the direct sound D as shown in FIG. To reach. When this direct sound D is reproduced by a multi-channel speaker arranged in an audio reproduction space (such as a living room) in the home, an appropriate level distribution amplitude and delay are added according to the angle information from the R speaker and the SR speaker. Thus, it can be perceived as a sound source coming from that direction. These output levels gR and gSR are assumed to have an amplitude of arrival gfr1,

で求められる。これに、距離Ｄなる行路分の遅延を加えることで直接音の再生がなされる。同様に初期反射音Ａ＋Ｂについても、その振幅、到来方向の角度、到来時間（遅延時間）の情報を使い音場処理係数データが算出される。

Is required. By directly adding a delay corresponding to the path of the distance D, the sound is directly reproduced. Similarly, for the initial reflected sound A + B, sound field processing coefficient data is calculated using information on its amplitude, angle of arrival direction, and arrival time (delay time).

  In the simulation, a speaker to be reproduced is selected from these three parameters (reflected sound arrival time Delay, reflected sound directions Dist1 and Dist2, reflected sound amplitudes Gainl and Gain2), and calculated from the formula [Equation 1]. The coefficient is used. Since the speakers reproduced here are always two adjacent speakers, the algorithm shown in FIG. 4 is obtained. To construct one reflected sound, it is realized by one delayed signal of a channel, two amplitude coefficients, and information on which two speakers reproduce the signal multiplied by this amplitude. Yes.

  In the signal processing, independent calculation of direct sound, early reflection sound, and reverberation sound (reverb) is performed on the sound input signal of each channel. The initial reflected sound is obtained from the simulation by the sound ray method described above. FIG. 9 shows an example of a geometric response diagram.

  The sound field mode (DAP mode) of the audio device (typically an AV amplifier) developed by the present applicant equipped with the multi-channel sound field processing device 10 provides 10 types of hall spaces shown in [Table 1] below. ing.

現在は、ＤＡＰの処理能力向上により、ディスクリート７ｃｈのディレイラインを設定することができるようになり、音場再生時に約４５０本の初期反射音を信号処理する（たたみ込める）までになっている。

Currently, with the improvement of DAP processing capability, a discrete 7-channel delay line can be set, and about 450 initial reflected sounds are processed (convolved) during sound field reproduction.

  Regarding the sound field processing of the reverberant sound by the DAP, the following [Patent Document 1] filed by the applicant of the present application and published below describes a new reverberant signal generation method and apparatus and reverberant signal addition apparatus. Details of the invention are disclosed.

  In other words, for the purpose of realizing a reverberation generator that generates a reverberation sound with a long delay time and a high reflected sound density with a hardware configuration with a small memory capacity, a predetermined phase below a mid-frequency with respect to a high-frequency. A delay unit in which a plurality of all-pass filters having a delay amount are vertically connected; a signal acquisition unit that acquires a predetermined amount of the signal of the delay unit; an acquired unit that adds the acquired signal and the supplied acoustic signal; The reverberation signal is obtained by cyclically performing the all-pass filter processing so that the added signal is supplied again to the all-pass filter means.

JP 2001-350487 A

  As described above, the multi-channel sound field processing apparatus 10 using the DAP has 10 types of multi-channel sound field processing modes (referred to as DAP modes) that provide a sense of realism of the actual mounted hall as described above. Coefficient data of early reflection sound (arrival of reflection sound) of several hundred (for example, 450) for a total of 4 patterns of 2 listening points assumed in the case of 5 channels and 7 channels in the mode (see Table 1) Time Delay, reflected sound directions Dist1, Dist2 and reflected sound amplitudes Gainl, Gain2) are all stored in the external memory device 9, and the coefficient data of the corresponding listening position of the corresponding channel number of the corresponding mode selected by the user at the time of DAP activation. Was being read.

  Thus, a huge amount of sound field processing coefficient data of four patterns of initial reflected sounds (450 in each example in the previous example) must be stored in an external memory device (ROM, etc.) for all modes. That's right.

  However, the memory capacity that can be used is naturally limited, and this has become the bottleneck for the realization of more precise multi-channel sound field processing. It is desirable to realize the ability.

  Conventionally, the sound field processing coefficient data of the initial reflected sound is calculated in advance on the premise of a multi-channel audio reproduction space having an ideal speaker arrangement on the user side. That is, the arrival direction (angle), amplitude, and delay time of the initial reflected sound are calculated on the premise of an ideal multi-channel speaker arrangement in the audio reproduction space as shown in FIGS. 5B and 5C. Therefore, depending on the speaker arrangement and the environment setting in the real user's home, the reproduction of the expected hall sound field may not be sufficiently obtained.

  The present invention has been made in view of the above circumstances, and the user has selected the sound field processing coefficient data of the early reflection sound for the actual multi-channel speaker arrangement in the actual audio reproduction space on the listener (user) side. Sound field processing is calculated for each mode (predetermined hole) each time, and sound field processing is performed based on this, realizing sound field processing suitable for the actual user's multi-channel audio playback space and storing it in the memory device An object of the present invention is to provide a multi-channel sound field processing apparatus that can reduce the amount of data necessary for sound field processing to be performed and reduce the required memory capacity.

  The present invention relates to multi-channel sound sources PA Lch, PA Rch arranged with respect to a multi-channel audio input signal L, R, Sl, Sr, C, Bl, Br with a hole space H having a predetermined size and shape as targets. , ... to the predetermined listening position, the output signal of each channel based on the acoustic characteristic data energy, angle, del obtained by simulating the geometric response of the amplitude, angle and arrival time of multiple early reflections The digital signal processing device obtains sound field processing coefficient data Dist1, Dist2, Gain1, Gain2, and Delay according to the arrival direction, amplitude, and delay time of the initial reflected sound to be added to a predetermined algorithm. A multi-channel sound field processing device 20 that performs multi-channel sound field processing according to the sound field processing coefficient data in a DAP, wherein the initial reflected sound The field processing coefficient data Dist1, Dist2, Gainl, Gain2, and Delay are input to the acoustic characteristic data stored in advance in the memory device 7 when the multi-channel sound field processing is started by the digital signal processing device DAP. To provide a multichannel sound field processing device 20 comprising sound field processing coefficient data calculating means for calculating with the predetermined algorithm based on the position information of the speaker of each channel in the multichannel audio reproduction space. Thus, the above problem is solved.

  The hall space in the present invention means a space where a large number of people can view videos and music, such as so-called concert halls, large and small movie theaters, stadiums, arenas, and churches.

Since the multi-channel sound field processing apparatus according to the present invention is configured as described above,
(1) Since the sound field processing coefficient data calculating means for calculating the sound field processing coefficient data for the mode (predetermined hall space) selected by the user each time when the multi-channel sound field processing is activated, Since it is not necessary to store the sound field processing coefficient data of hundreds of early reflection sounds for a large number of hall spaces in the memory device in advance, the necessary memory capacity of the memory device can be reduced, and the cost can be reduced. To contribute.
(2) Appropriate sound field processing of the initial reflected sound with respect to the multi-channel speaker arrangement in the multi-channel audio reproduction space on the actual user side can be realized.

  Embodiments of a multichannel sound field processing apparatus according to the present invention will be described with reference to the drawings.

  The description of the process similar to the above-described related art regarding DAP is omitted, and only the sound field processing coefficient data calculating means will be described in detail.

  In the embodiment of the present invention, the calculation step of the sound field processing coefficient data is changed for the initial reflected sound in the multi-channel sound field processing by DAP, and the reverberant sound and the direct sound are the same as before.

  FIG. 1 is a block diagram for explaining a signal generation algorithm of a sound field processing device 20 corresponding to a multi-channel audio input signal (in the case of 7.1 channels) according to the present invention. 2 and 3 are diagrams for explaining an algorithm for obtaining sound field processing coefficient data in the multi-channel sound field processing apparatus according to the present invention.

  In FIG. 1, a plurality of multi-channel sound field processing devices 20 are arranged for a hole space H having a predetermined size and shape with respect to multi-channel audio input signals L, R, S1, Sr, C, Bl, Br. Obtained by simulating the geometric response of the amplitude energy, angle angle, and arrival time del of a plurality of early reflections from a multi-channel sound source PA Lch, PA Rch,... To a predetermined listening position (listener M) Based on the acoustic characteristic data, the sound field processing coefficient data Dist1, Dist2, Gainl, Gain2, distributed according to the direction of arrival of the initial reflected sound, the amplitude, and the delay time, which are added to the output signal of each channel, Multi-channel which obtains Delay by a predetermined algorithm and performs multi-channel sound field processing in accordance with the sound field processing coefficient data by digital signal processing device DAP The sound field processing device 20 is configured to use the sound field processing coefficient data Dist1, Dist2, Gainl, Gain2, and Delay of the initial reflection sound when the multi-channel sound field processing is started by the digital signal processing device DAP. The multi-channel audio reproduction space (the living room of the user's home) input by the user from the outside by the numeric keypad input device 6 and the like, and the acoustic characteristic data (initial reflection sound amplitude energy, angle angle and arrival time del) stored in advance And a sound field processing coefficient data calculation means for calculating with the predetermined algorithm based on the position information (arrangement angle information, etc.) of the speakers of each channel in a space in which a multi-channel AV system is installed. It has become.

  Here, in the conventional multi-channel sound field processing apparatus 10, DAP modes (multi-channel sound field processing modes) are prepared for the 10 types of predetermined holes shown in [Table 1], and 5ch and 7ch are provided for each sound source. If two sound field processes are prepared for two listening positions, a large number (about 450) of sound field processing coefficient data of the initial reflected sound is stored in advance in the external memory device 9 for a total of 40 patterns. Although it is necessary, in the multi-channel sound field processing device 20 of the present invention, the sound field processing coefficient data is not stored in the memory device 7 in advance, and the data stored in the memory device 7 in the DAP is Acoustic characteristic data for two listening positions for each sound source obtained by simulation for the ten predetermined holes. Is only a memory capacity in need thereof is remarkably reduced in comparison with the conventional.

  The sound field processing coefficient data of 450 early reflected sounds necessary for multi-channel sound field processing is stored in the memory device 7 by the CPU 8 in the DAP when the DAP is activated only for one DAP mode selected by the user. Are appropriately calculated on the basis of the acoustic characteristic data at the corresponding listening position of the corresponding number of channels read out from the above and the angle data of each speaker arrangement of the actual multi-channel speaker input by the user, and used for sound field processing.

  As described above, by automatically calculating and setting the coefficient data for multichannel sound field processing inside the DAP, the external memory capacity around the DAP can be reduced, and the actual multichannel audio reproduction space on the user side can be reduced. The generation of the optimal multi-channel sound field processing in consideration of the speaker arrangement is realized.

  Here, the algorithm for automatic calculation of the sound field processing coefficient data Dist1, Dist2, Gainl, Gain2, and Delay is as follows.

  First, in step 1, the speaker position (speaker arrangement angle) in the actual multi-channel audio reproduction space is set.

  First, the user inputs and sets position information (speaker arrangement angle) of each multi-channel speaker in an actual listening environment or viewing environment by a numeric keypad input operation using the input device 6 from the outside. The information on the speaker arrangement angle is a counterclockwise angle with the front of the user as the reference 0 °. For example, in FIG. 2, when the front left speaker FL is shifted by α ° as shown in (b) with respect to the ideal 7ch speaker arrangement of (a), the position (angle) of this speaker FL is 30 °. + Α °. In addition to the angle information, distance information can also be used as the position information of each speaker.

  Next, the amplitude (energy), angle (angle), and arrival time (del) of the number of initial reflected sounds of the acoustic characteristic data at the predetermined listening location in the DAP mode (predetermined hall) selected by the user in step 2 are stored in memory. Read from device 7.

  Next, in step 3, the arrival direction (Dist1, Dist2), the amplitude of the initial reflected sound (Gain1, Gain2), and the delay time (Delay), which are sound field processing coefficient data of each initial reflected sound, are provided in the DAP. The calculation is performed as follows based on the speaker position information (angle information) obtained in step 1 in the calculation routine according to the above algorithm.

  First, the amplitude (Gain) of the initial reflected sound shown in FIG. 3 is expressed by the following equation.

なお、ｍ＝Ｙ／Ｘ＋Ｙ，ｎ＝Ｘ／Ｘ＋Ｙである。

Note that m = Y / X + Y and n = X / X + Y.

  Next, the delay time sample of the initial reflected sound is calculated by the following equation [Equation 5].

ここに、上記first.delは直接音Ｓの到達時間サンプルであり、fsはサンプリング周波数、delは反射音Ａ＋Ｂの到達時間である。

Here, the first.del is the arrival time sample of the direct sound S, fs is the sampling frequency, and del is the arrival time of the reflected sound A + B.

  Also, the arrival direction (Dist1, Dist2) of the reflected sound is determined by which position (which speaker is between) the reflected sound angle angle with respect to the actual speaker arrangement angle in step 1. For example, when speaker FL = 35 °, speaker C = 10 °, and angle = 20 °, Dist1 is given to speaker FL and Dist2 is given to speaker C.

  As described above, it is understood that appropriate level / delay distribution of sound field processing can be performed according to the speaker arrangement angle in the actual multi-channel audio reproduction space on the user side obtained in step 1.

It is a block diagram for demonstrating the signal generation algorithm of the sound field processing apparatus corresponding to the multichannel audio | voice input signal which concerns on this invention (in the case of 7.1 channel). It is a figure for demonstrating the algorithm which calculates | requires the sound field process coefficient data in the multichannel sound field processing apparatus which concerns on this invention. It is a figure for demonstrating the algorithm which calculates | requires the sound field process coefficient data in the multichannel sound field processing apparatus which concerns on this invention. It is a block diagram for demonstrating the signal generation algorithm of the sound field processing apparatus corresponding to a multi-channel audio | voice input signal (in the case of 7 channels). Schematic diagram in the case of (a) 2 channels, (b) 5.1 channels, (c) 7.1 channels representing the speaker arrangement for an ideal listener M in the user's audio playback space (such as a living room at home) It is. (A) is a schematic diagram showing the direction and distance (delay) of the sound line of the direct sound from each 5.1-channel sound source in the hall, and (b) is an ideal 7ch speaker in the audio reproduction space on the user side. It is a schematic diagram which shows the relationship between the distance with arrangement | positioning, and the direction of the said sound ray. (A) is a schematic diagram of primary reflected sound using PARch as a sound source in a hall, and (b) is a schematic diagram showing a relationship between a distance and a direction with an ideal 7-channel speaker arrangement in an audio reproduction space on the user side. is there. (A) is an example of the sound ray of the primary reflection sound of the sound source of each channel reflected by the left and right wall surfaces, and (b) is an ideal speaker arrangement in the audio reproduction space on the user side and the sound ray of the primary reflection sound. It is a schematic diagram which shows the relationship of these directions. It is a three-dimensional schematic diagram showing an example of a geometric response diagram of a sound ray in a predetermined hole by conventional DAP processing.

Explanation of symbols

6 Input device
7 Memory device (ROM)
8 CPU
9 External memory device 10, 20 Multi-channel sound field processing device
H Hall space L, R, Sl, Sr, C, Bl, Br Multi-channel audio input signal PA Lch, PA Rch, ... Sound source
Energy Amplitude of acoustic characteristics data of early reflections
Angle Angle of acoustic characteristic data of early reflection sound
del Time of arrival of acoustic characteristics data of early reflections Dist1, Dist2 Sound field processing coefficient data of arrival direction of early reflections Gain1, Gain2 Sound field processing coefficient data of amplitude of early reflections Delay Sound field of delay time of early reflections Processing coefficient data

Claims (1)

A geometric response of the amplitude, angle, and arrival time of multiple early reflections from a multi-channel sound source arranged in a hole space of a predetermined size and shape to a predetermined listening position in response to a multi-channel audio input signal Sound field processing coefficient data distributed according to the speaker arrangement angle of the arrival direction, amplitude and delay time of the initial reflected sound added to the output signal of each channel based on the acoustic characteristic data obtained by simulating A multi-channel sound field processing device that performs multi-channel sound field processing according to the sound field processing coefficient data in a digital signal processing device,
When the multi-channel sound field processing by the digital signal processing device is activated, the acoustic characteristic data stored in advance in the memory device and the multi-channel audio reproduction space input from the outside are used for the sound field processing coefficient data of the initial reflected sound. And a sound field processing coefficient data calculating means for calculating with the predetermined algorithm based on the position information of the speaker of each channel in the multi-channel sound field processing device.

Images