JP2007065497A - Signal processing apparatus - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To solve the problem, wherein when a decorrelation signal is created by a delay means and an all pass filter, when input voice is sharp in the temporal characteristics or in frequency characteristics, sharpness is lost by an unnecessary reverberation signal. <P>SOLUTION: A signal processing apparatus comprises a generating means for generating a second signal from a first signal; a mixing coefficient determination means for determining a mixing rate; and a mixing means for mixing the first signal with the second signal, on the basis of the mixing rate determined by the mixing coefficient determination means. The generating means comprises a delay means for delaying the first signal by N (N>0) unit time; a filtering means for processing the output signal of the delay means; and a processing means for processing the first signal. The generating means generates the second signal from the output signal of the filter means and the output signal of the processing means. <P>COPYRIGHT: (C)2007,JPO&INPIT

Description

  The present invention relates to a signal processing apparatus for decoding a signal obtained by down-mixing a plurality of signals and an encoded signal obtained by encoding information for separating the signal into the original signal. In particular, the present invention relates to a technique capable of decoding an encoded signal obtained by encoding multi-channel presence with a small amount of information by encoding a phase difference between signals and a level ratio.

  In recent years, a technology called Spatial Codec has been developed. This is for the purpose of compressing and encoding the presence of multi-channel with a very small amount of information. For example, the AAC system, which is a multi-channel codec that is already widely used as an audio system for digital television, is 5.1. While bit rates of 512 kbps and 384 kbps are required per channel, Spatial Codec aims to compress and encode multichannel signals at very low bit rates of 128 kbps, 64 kbps, and 48 kbps. As a technique for that purpose, for example, there is Parametric Coding for High Quality Audio (Non-patent Document 1) standardized by the MPEG audio system. According to this, a process is described in which a signal in which a sense of presence is compression-coded with a small amount of information is encoded by encoding a phase difference between channels or a level ratio. FIG. 5 shows the process. First, the input signal S is a signal obtained by downmixing what was originally a 2ch signal into a monaural signal. The input signal S is input to a processing module called decorrelation to obtain an output signal D. The decorrelation process is described in detail in Section 8.6.4.5.2 Calculate Decorrelated signal in Non-Patent Document 1, so detailed description is omitted, but decorrelation is mainly composed of two processes. The first is delay processing. This is a process of delaying the input signal by a predetermined time. Thereafter, the delayed signal is subjected to a process called All Pass Filter. This process is a process for giving a reverberation component to the signal. The signal D thus generated and the input signal S are subjected to a process called “Mixing”. This process is also described in detail in 8.6.4.6.2 Mixing of Non-Patent Document 1, so detailed description is omitted, but the two signals S and D are multiplied by coefficients h11, h12, h21, and h22. Are combined to obtain an output Lch signal and Rch signal. The equation is as shown in the figure. Here, the coefficients h11, h12, h21, and h22 are values determined by the level ratio and phase difference between the original 2ch signals based on the input monaural signal. Since this method of obtaining the coefficients h11, h12, h21, h22 from the information of the phase difference is also described in Non-Patent Document 1, it is omitted here.

By performing such processing, a spatial spread is given when a 2ch signal is generated from a monaural signal due to the effects of delay processing in decorrelation and the addition of reverberation components, which is favorable. A stereo signal is obtained.
ISO / IEC 14496-3: 2001 / FDAM 2: 2004 (E)

  However, the above method has the following problems. That is, when the input signal is matched with a signal with very severe time fluctuation (for example, at the moment of attack of a metallic percussion instrument), the delay and reverberation components in the decorrelation process are effective. The signal after decorrelation loses its sharpness. Further, since the signal after the decorrelation is added to the input signal S by the mixing process in the subsequent stage, as a result, the output signal loses the sharpness of the input signal.

  Similarly, when the frequency component of the input signal is biased to a specific frequency band (for example, when the tone of one kind of instrument continues continuously), it is inherently very solid. Although a sound image should be formed, due to the effects of delay and addition of reverberation components in the decorrelation process, the sound image with a firm localization is blurred in the signal after decorrelation. Further, the signal after the decorrelation is added to the input signal S by the subsequent mixing process, and as a result, the sound image of the output signal is blurred.

  In addition, since the generated 2ch signal is separated from the monaural signal using only the level ratio and phase difference information as a clue, the separation performance is often insufficient.

The present invention has been made in view of such conventional problems,
When a 2ch signal is generated from a monaural signal, a spatial spread is given and a good stereo signal is obtained. At the same time, the sharpness of temporal fluctuation of sound and the localization of the sound image are fixed. An object of the present invention is to provide a signal processing apparatus that can be realized.

  Another object of the present invention is to provide a signal processing device that compensates for the lack of separation.

  In order to solve the above-described problem, the signal processing device according to claim 1 is configured to mix the first signal and the second signal generated from the first signal in two degrees of mixing. A signal processing apparatus for generating two signals, wherein the generating means for generating the second signal from the first signal, the mixing coefficient determining means for determining the degree of mixing, and the mixing coefficient determining means Mixing means for mixing the first signal and the second signal based on the determined degree of mixing, and the generating means converts the first signal to N (N> 0). A delay unit that delays the unit time; a filter unit that processes an output signal of the delay unit; and a processing unit that processes the first signal. The generation unit includes the output signal of the filter unit and the processing Generating the second signal from the output signal of the means. The one in which the features.

  The signal processing apparatus according to claim 2, wherein the generation unit generates the second signal from an output signal of the filter unit and an output signal of the processing unit according to an acoustic feature amount of the first signal. And combining means for combining, wherein the acoustic feature quantity is a feature quantity that becomes large when the first signal is abruptly fluctuating.

  The signal processing apparatus according to claim 3, wherein the generation unit generates the second signal from an output signal of the filter unit and an output signal of the processing unit in accordance with an acoustic feature amount of the first signal. The acoustic feature amount is a feature amount that becomes large when strong energy is concentrated in a specific frequency band in the first signal.

  The signal processing apparatus according to claim 4, wherein the combining unit outputs an output signal of the filter unit when the feature amount is small, and an output of the processing unit when the feature amount is large. A signal is output.

  The signal processing apparatus according to claim 5, wherein the combining unit includes a second mixing unit that mixes the output signal of the filter unit and the output signal of the processing unit, and the feature amount is small. The output signal of the filter means is mixed in a large amount, and when the feature amount is large, the output signal of the processing means is mixed in a large amount.

  The signal processing apparatus according to claim 6, wherein the processing means includes second filter means, and the second filter means has a filter order less than the first filter means in the previous period. is there.

  The signal processing apparatus according to claim 7, wherein the processing unit includes a second delay unit, and the second delay unit has a delay amount smaller than that of the first delay unit. .

  The signal processing device according to claim 8 is provided with a third filter means, and the third filter means is a process for rotating the phase of the input signal by 90 degrees or -90 degrees. .

  The signal processing device according to claim 9 is configured such that the generation unit can process a signal independently for each of a plurality of frequency components. A signal is output, and a signal of the processing means is output for a signal in a high frequency band.

  The signal processing apparatus according to claim 10, wherein the first signal is a signal obtained by downmixing two signals, and the mixing coefficient determination unit is configured to output a level ratio L between the two original signals. The degree of mixing is determined from a value determined according to the phase difference θ.

  12. The signal processing apparatus according to claim 11, wherein the mixing coefficient determining means is configured such that the angle of the parallelogram where the angle formed by two adjacent sides is the θ and the length ratio is the L is the parallelogram. Assuming that the angles obtained by dividing the diagonal of the shape are A and B, and values d1 and d2 determined according to the level ratio L, d1 * cos (A), d1 * sin (A), d2 * cos ( -B) and d2 * sin (-B), the mixing means uses the complex part to express the real part as r1, the imaginary part as i1, and the second signal as a complex number. When the real part when expressed is r2 and the imaginary part is i2, d1 * cos (A) * r1 + d1 * sin (A) * r2 is the real part of the first output signal and d1 * cos (A ) * i1 + d1 * sin (A) * i2 is the imaginary part of the first output signal, and d2 * cos (-B) * r1 + d2 * sin (-B) * r2 is the real part of the second output signal , D2 * cos (-B) * i1 + d2 * sin (-B) * i2 is the imaginary part of the second output signal It is characterized in.

  The signal processing apparatus according to claim 12, wherein the mixing coefficient determination unit sets the values of d1 and d2 to d1 = L / ((1 + 2 * L * cos (θ) + L * L) ^ 0.5), d2 = 1 / ((1 + 2 * L * cos (θ) + L * L) ^ 0.5).

  The signal processing apparatus according to claim 13, wherein the mixing coefficient determining means sets the values of d1 and d2 to d1 = L / ((1 + L * L) ^ 0.5), d2 = 1 / ((1 + L * L) ^ 0.5).

  The signal processing device according to claim 14 further includes a feature amount receiving unit that receives data in which the acoustic feature amount is encoded, and the generation unit is configured to respond to the data in which the acoustic feature amount is encoded. A signal is generated.

  The signal processing device according to claim 15, wherein the data obtained by encoding the acoustic feature amount is 1-bit data, and the generation unit outputs an output signal of the processing unit when the data is true, In the case of false, the output signal of the filter means is output.

  17. The signal processing device according to claim 16, wherein the first signal and the second signal generated from the first signal are mixed in two degrees of mixing to generate two signals. An apparatus for generating the second signal from the first signal; a mixing coefficient determining means for determining the degree of mixing; and a mixing degree determined by the mixing coefficient determining means. Mixing means for mixing the first signal and the second signal, and sound image moving means for performing signal processing for moving the reproduced sound images of the two signals generated from the mixing means to positions separated from each other. It is characterized by having.

  The signal processing device according to claim 17, wherein the sound image moving means receives two signals generated from the mixing means and processes the signals to move the reproduced sound images of the two signals to positions separated from each other. It is characterized by this.

  19. The signal processing apparatus according to claim 18, wherein the sound image moving means processes the degree of mixing determined by the mixing coefficient determining means so that the reproduced sound images of the two signals generated by the mixing means are separated from each other. It is characterized by moving to a position.

  20. The signal processing apparatus according to claim 19, wherein the sound image moving means moves the reproduced sound image of the signal in the first direction with respect to one of the two signals generated from the mixing means. And second processing means for performing signal processing for moving the reproduced sound image of the signal in a second direction that is opposite to the first direction with respect to the other signal. It is characterized by providing.

  The signal processing apparatus according to claim 20, wherein the first processing unit performs processing of changing an amplitude of a signal in a predetermined frequency band, and the second processing unit has a frequency band different from the frequency band. A process for changing the amplitude of the signal is performed.

  The signal processing apparatus according to claim 21, wherein the two signals generated from the mixing unit are approximated to output a signal generated from the sound image moving unit, and otherwise, generated from the mixing unit. The output signal is output.

  According to the first aspect of the present invention, when a stereo signal is generated from a monaural signal, an appropriate stereo signal can be generated regardless of whether the reverberation component is useful or unnecessary.

  According to the second aspect of the present invention, it is possible to generate an appropriate stereo signal for an input signal having a sharp time variation.

  According to the third aspect of the present invention, an appropriate stereo signal can be generated for an input signal having a sharp frequency characteristic.

  According to the invention of claim 4, it is possible to switch appropriately according to the state of the input signal.

  According to the invention of claim 5, even if the state of the input signal is an intermediate state, an appropriate signal can be generated.

  According to the sixth aspect of the invention, an appropriate reverberant signal can be generated by a filter having a small filter order.

  According to the seventh aspect of the invention, an appropriate reverberation signal can be generated by the delay means having a small delay amount.

  According to the eighth aspect of the present invention, an appropriate reverberant signal can be generated by a filter without a filter delay.

According to the ninth aspect of the invention, the reverberation component can be controlled independently for each frequency component.
According to the tenth aspect of the present invention, the degree of mixing can be determined by the level ratio and the phase difference information.

  According to the eleventh aspect of the invention, when the degree of mixing is obtained from the level ratio and the phase difference information, the phase distribution can be determined mathematically and correctly.

  According to the twelfth aspect of the present invention, when the degree of mixing is obtained from the level ratio and the phase difference information, the gain distribution can be determined mathematically and correctly.

  According to the thirteenth aspect of the invention, when the degree of mixing is obtained from the level ratio and the phase difference information, the gain distribution can be easily determined.

  According to the fourteenth aspect of the invention, since a signal for determining whether the reverberation component is useful or unnecessary is encoded and given in advance, an appropriate stereo signal can be easily generated.

  According to the fifteenth aspect of the invention, since the signal for determining whether the reverberation component is useful or unnecessary is 1 bit, an appropriate stereo signal can be generated by simple switching.

  According to the sixteenth and seventeenth aspects of the invention, the separation performance is insufficient because the generated 2ch signal is separated from the monaural signal based on only the level ratio and phase difference information. You can make up for it.

  According to the eighteenth aspect of the invention, since only the coefficient of the mixing means is changed, the separation performance can be improved with almost no increase in the calculation amount.

  According to the nineteenth aspect of the invention, the insufficient separation performance can be compensated by moving the sound images of both signals in opposite directions.

  According to the twentieth aspect of the invention, insufficient separation performance can be compensated with a very small amount of calculation.

  According to the twenty-first aspect of the present invention, insufficient separation performance can be compensated only when separation performance is insufficient.

(Embodiment 1)
Hereinafter, a signal processing apparatus according to Embodiment 1 of the present invention will be described with reference to the drawings.
FIG. 1 is a diagram showing the configuration of the signal processing apparatus according to the first embodiment. The signal processing apparatus includes a first encoded signal obtained by encoding a signal obtained by down-mixing two audio signals, and a second code obtained by encoding a value determined according to the level ratio L between the two audio signals. And a third encoded signal obtained by encoding a value determined according to a phase difference θ between the two audio signals.

  In FIG. 1, reference numeral 100 denotes decoding means for decoding the first encoded signal to generate a first signal, 101 denotes generation means for generating the second signal from the first signal, and 102 denotes Mixing coefficient determining means for determining a mixing coefficient from the second encoded signal and the third encoded signal, 103 is based on the degree of mixing determined by the mixing coefficient determining means 102, Mixing means for mixing the first signal and the second signal; 104, delay means for delaying the first signal by N (N> 0) unit time; and 105, processing the output signal of the delay means 104 First filter means 106, second filter means 106 for processing the output signal of the delay means 104, 107 feature quantity detection means for detecting an acoustic feature quantity of the first signal, 108, According to the acoustic feature amount, Serial first combining means for combining the second signal from the output signal and the output signal of said second filter means 106 of the filter means 105, it is.

The operation of the signal processing apparatus configured as described above will be described below.
First, the decoding unit 100 decodes the first encoded signal to generate a first signal. Here, the first encoded signal is obtained by encoding a monaural signal obtained by downmixing two audio signals, and is encoded by, for example, an MPEG AAC standard encoder. Here, it is assumed that the decoding means 100 performs the process until the PCM signal obtained by decoding such an AAC standard encoded signal is converted to a frequency signal composed of a plurality of frequency bands. In the following description, processing for a signal in one specific band among the signals in such a plurality of frequency bands will be described.

  The generation unit 101 generates a second signal from the first signal, and this is performed as follows. That is, the delay means 104 first delays the first signal by N (N> 0) unit time. Next, the first filter means 105 performs a filtering process on the output signal of the delay means 104. For example, as this process, an All Pass Filter whose degree is P order is performed. The All Pass Filter process may be performed by any conventionally known method. For example, the All Pass Filter described in Section 8.6.4.5.2 of Non-Patent Document 1 may be used. On the other hand, the second filter means 106 performs an All Pass Filter process on the output signal of the delay means 104, the order of which is less than the P order.

  The output signal from the first filter unit 105 and the output signal from the second filter unit 106 thus generated are processed by the synthesis unit 108 to generate the second signal. . This process is as follows. That is, the feature quantity detection means 107 detects the acoustic feature quantity of the first signal, and the output signal from the first filter means 105 and the second filter means 106 according to the feature quantity. The ratio of mixing with the output signal from is determined.

  For example, the acoustic feature amount is a feature amount that becomes large when the first signal fluctuates sharply, and the synthesizing unit determines that the first feature is small when the acoustic feature amount is small. The output signal of the filter means 105 is output, or the output signal of the first filter means 105 is increased and the output signal of the second filter means 106 is mixed slightly and output. On the other hand, when the acoustic feature quantity is large, the output signal of the second filter means 106 is output, or the output signal of the first filter means 105 is reduced, and the second filter means A large amount of output signals from the means 106 are mixed and output.

  Here, the acoustic feature amount may be a feature amount that becomes large when strong energy is concentrated in a specific frequency band of the first signal. Alternatively, a combination of such feature amounts may be used.

  What is important here is that the acoustic feature amount is a feature amount that represents the sharpness of temporal variation in sound and the sense of localization of a sound image. This is because the filter means 105 is a P-th order All Pass Filter and gives a reverberation to the sound. Therefore, when such a reverberation is not necessary, that is, sharpness of the temporal variation of the sound. This is because, when a firm sense of localization of the sound image is required, it is necessary to reduce the reverberation by reducing the order of the All Pass Filter.

  From this point of view, the generation unit 101 may have a configuration as shown in FIG. In FIG. 2, the delay means 104, the first filter means 105, and the synthesis means 108 are the same as those shown in FIG. In FIG. 2, reference numeral 200 denotes second delay means for delaying the first signal by n (N> n ≧ 0) unit time. Reference numeral 201 denotes third filter means for rotating the phase of the input signal by 90 degrees or -90 degrees.

  The delay means 104 and the filter means 105 have an effect of giving a sense of spatial spread and reverberation of the sound, but when they are unnecessary, that is, the sharpness of the temporal variation of the sound and the localization of the sound image. When feeling is needed, it is necessary to reduce the amount of delay or the amount of reverberation. In such a case, the second delay unit 200 having a delay amount smaller than that of the delay unit 104 is used, and further, the third filter having a low reverberation feeling is used. The delay amount of the second delay means 200 may be zero. That is, the second delay means 200 may be omitted. The third filter means 201 rotates the phase of the input signal by 90 degrees or -90 degrees. However, this is a very small amount of computation, and a signal that is uncorrelated with the input signal and has no delay is generated. Therefore, it is convenient as a means for generating a sharp signal that is uncorrelated with the input signal. Here, it is very important that the generated signal is uncorrelated with the input signal (the first signal). This is because if the signal has a high correlation, it is simply a monaural sound (a sound without a sense of stereo) when it is mixed with the first signal by the processing by the subsequent mixing means.

The output signal from the filter means 105 obtained in this way and the third filter means 201 are synthesized by the synthesis means 108 according to the acoustic feature quantity, but the method is the same as described above. It's okay. By doing so, a sharp and well-positioned sound can be generated when a feeling of reverberation or a feeling of sound spread is unnecessary.
Now, in this way, the second signal generated by the generating unit 101 and the first signal are mixed by the mixing unit 103. The operation will be described below.

  First, the mixing coefficient determining means 102 determines a mixing coefficient from the second encoded signal and the third encoded signal. The second encoded signal is obtained by encoding a value determined according to the level ratio L between the two original audio signals, and the third encoded signal is a level between the two original audio signals. A value determined according to the phase difference θ is encoded. A method for obtaining the mixing coefficients h11, h12, h21, h22 from such level ratio information and phase difference information is described in detail in, for example, Section 8.6.4.6.2 Mixing of Non-Patent Document 1 described above. The following method may be used.

  That is, the angles obtained by dividing the parallelogram θ where the angle between two adjacent sides is θ and the length ratio is the L by the diagonal of the parallelogram are A and B. When the values determined according to the level ratio L are d1 and d2, h11 = d1 * cos (A), h21 = d1 * sin (A), h12 = d2 * cos (−B), h22 = d2 * Let sin (-B). In the above, the values of d1 and d2 are as follows: d1 = L / ((1 + 2 * L * cos (θ) + L * L) ^ 0.5), d2 = 1 / ((1 + 2 * L * cos (θ ) + L * L) ^ 0.5). In this way, the downmixed and monaural signal can be mathematically and accurately separated into two signals according to the phase difference and level ratio of the original two signals. The reason is shown in FIG. In the parallelogram XYZW where the angle formed by two adjacent sides is θ and the length ratio is L, the angle YXZ obtained by dividing the parallelogram XYZW is A, and the angle WXZ is B. The diagonal length XZ is mathematically obtained as ((1 + 2 * L * cos (θ) + L * L) ^ 0.5. Therefore, the above d1 and d2 are d1 = L / ((1+ 2 * L * cos (θ) + L * L) ^ 0.5), d2 = 1 / ((1 + 2 * L * cos (θ) + L * L) ^ 0.5).

  In the above, the values of d1 and d2 may be simply obtained as d1 = L / ((1 + L * L) ^ 0.5) and d2 = 1 / ((1 + L * L) ^ 0.5).

The first signal and the second signal are mixed by the mixing unit 103 using the generated mixing coefficients h11, h21, h12, and h22. The method is as follows. That is, when the real part when the first signal is expressed by a complex number is r1, the imaginary part is i1, the real part when the second signal is expressed by a complex number is r2, and the imaginary part is i2, h11 * r1 + h21 * r2 is the real part of the first output signal, h11 * i1 + h21 * i2 is the imaginary part of the first output signal, and h12 * r1 + h22 * r2 is the real part of the second output signal , H12 * i1 + h22 * i2 is the imaginary part of the second output signal.
As described above, according to the present embodiment, the first signal and the second signal generated from the first signal are mixed in two degrees of mixing (in the combination of h11 and h21). In the signal processing device that generates two signals by mixing in a combination of h12 and h22), the generating means for generating the second signal from the first signal, and the mixing Mixing coefficient determining means for determining the degree of the above, and mixing means for mixing the first signal and the second signal based on the degree of mixing determined by the mixing coefficient determining means, The generating means includes delay means for delaying the first signal by N (N> 0) unit time, an All Pass Filter for processing an output signal of the delay means, and processing means for processing the first signal. The processing means includes the delay means and an All Pass Filter. Therefore, a signal with less sound spread or reverberation than the generated signal is generated, and the first signal is a signal that fluctuates sharply, or strong energy is concentrated in a specific frequency band. When the 2ch signal is generated from the monaural signal by mixing a large amount of the output signal of the processing means to the second signal, Thus, a good stereo signal can be obtained, and at the same time, the sharpness of temporal fluctuation of sound and the localization of sound image can be realized.

  In the present embodiment, the acoustic feature quantity is detected by the feature quantity detection unit 107, but it is not always necessary to receive data obtained by encoding the acoustic feature quantity in advance. Also good. The configuration diagram in that case is as shown in FIG. The only difference between FIG. 1 and FIG. 6 is that a feature amount receiving means 109 is provided instead of the feature amount detecting means 107. The feature amount receiving unit 109 receives data obtained by encoding the acoustic feature amount of the input signal as the fourth encoded signal. For example, the fourth encoded signal is an encoded signal that is true when strong energy is concentrated in a specific frequency band, and is false otherwise. When the fourth encoded signal is true, the generating means 101 is a signal with a small reverberation component (that is, a signal processed with a filter with a short filter tap length for a signal with a small delay amount or no delay, If not, a signal having a large reverberation component (that is, a signal obtained by processing a signal having a large delay amount with a filter having a long filter tap length) is generated. By doing so, the processing as intended on the encoding apparatus side can be performed, so that a high-quality sound signal can be generated. In this case, it is needless to say that the synthesizing means 108 has only a function of a selector.

(Embodiment 2)
A signal processing apparatus according to Embodiment 2 of the present invention will be described below with reference to the drawings. The second embodiment is greatly different from the first embodiment in that the first embodiment sequentially adapts the method of generating the second signal according to the sequentially input signals. On the other hand, in the second embodiment, it is considered that the signal in the low frequency band greatly contributes to the reverberation and spread feeling of the sound, and the signal in the high frequency band greatly contributes to the sharpness of the sound. However, the generation means is changed between the low range and the high range.

  FIG. 4 is a diagram showing the configuration of the signal processing apparatus according to the second embodiment. The signal processing apparatus includes a first encoded signal obtained by encoding a signal obtained by down-mixing two audio signals, and a second code obtained by encoding a value determined according to the level ratio L between the two audio signals. And a third encoded signal obtained by encoding a value determined according to a phase difference θ between the two audio signals.

  In FIG. 4, reference numeral 400 denotes decoding means for decoding the first encoded signal to generate a first signal, 401 denotes generation means for generating the second signal from the first signal, and 402 denotes Mixing coefficient determining means for determining a mixing coefficient from the second encoded signal and the third encoded signal, 403, based on the degree of mixing determined by the mixing coefficient determining means 402, Mixing means for mixing the first signal and the second signal.

  Here, the first signal is a frequency signal composed of a plurality of frequency bands, and the generating means 401 independently processes the signals of the respective frequency bands as shown in FIG. Generate a signal. For example, a low frequency band signal is processed by a delay unit and a filter unit, but a high frequency band signal is processed only by a filter unit. May be. Further, the delay amount for the signal in the low frequency band may be the same or larger than that in the higher frequency signal. Further, the filter order of the filter means for the signal in the low frequency band may be the same or larger than that in the higher frequency band. Further, the filter means having a band higher than the predetermined band may be a process of rotating the input signal by 90 degrees or -90 degrees.

The operation of the signal processing apparatus configured as described above will be described below.
First, the decoding unit 400 decodes the first encoded signal to generate a first signal. Here, the first encoded signal is obtained by encoding a monaural signal obtained by downmixing two audio signals, and is encoded by, for example, an MPEG AAC standard encoder. Here, it is assumed that the decoding means 400 performs the process until the PCM signal obtained by decoding such an AAC standard encoded signal is converted into a frequency signal composed of a plurality of frequency bands. The generation unit 401 generates a second signal from the first signal, which is performed as follows. That is, among the plurality of frequency bands constituting the first signal, for the low frequency band, the signal is delayed by a preset value N unit time, and for the signal thus delayed, All pass filter processing of the order P is performed. Here, the All Pass Filter process may be any method known in the art, but may be, for example, the All Pass Filter described in Section 8.6.4.5.2 of Non-Patent Document 1 described above. .

  Further, for a signal in a frequency band higher than the frequency band described above, the signal is delayed by a time unit of a value n (N ≧ n ≧ 0) that is equal to or smaller than N, and so The delayed signal is subjected to an All Pass Filter process of the order p (P ≧ p ≧ 0) whose order is equal to or smaller than P. Alternatively, instead of the All Pass Filter process, a process of rotating the input signal by 90 degrees or -90 degrees may be used.

  In short, the lower frequency band signal gives more delay and longer filter taps, and the higher frequency band signal gives less delay and shorter filter taps. Reduce the sense of sound spread and reverberation. The reason for doing this is that, in general, the signal in the low frequency band greatly contributes to the reverberation and spread of the sound, and the signal in the high frequency band greatly contributes to the sharpness of the sound. is there. Of course, when the perceptual characteristics of hearing are analyzed precisely for each fine frequency band and based on the result, it should be limited to a method in which the value decreases in a minor manner as it goes from low to high as described above. is not. What is important here is that the value is controlled independently for each frequency band.

  Now, the second signal generated in this way and the first signal are mixed by the mixing unit 403 using the mixing coefficient determined by the mixing coefficient determining unit 402. The operation may be the same as that shown in the first embodiment.

  As described above, according to the present embodiment, the first signal and the second signal generated from the first signal are mixed in two degrees of mixing (in the combination of h11 and h21). In the signal processing device that generates two signals by mixing in a combination of h12 and h22), the generating means for generating the second signal from the first signal, and the mixing Mixing coefficient determining means for determining the degree of the above, and mixing means for mixing the first signal and the second signal based on the degree of mixing determined by the mixing coefficient determining means, The generating means includes a delay means for delaying a relatively large value N (N> 0) unit time for a signal in a low frequency band of the first signal, and an All Pass having an order of a relatively large value P. A signal is generated with the filter, and the first signal is high. For a signal in the frequency band, a delay means for delaying a relatively small value n unit time (or not delaying it at all), an All Pass Filter having an order of a relatively small value p (or an input signal of 90 degrees or -90 degrees) When a 2ch signal is generated from a monaural signal, a spatial spread is given and a good stereo signal can be obtained at the same time. The sharpness of temporal fluctuations and the sound localization can be realized.

  In the second embodiment, the processing method (delay amount and filter order) of each frequency band signal is fixed regardless of the nature of the input signal. However, of course, it is not necessary to limit to this method. You may switch suitably according to it. For example, the frequency band below the frequency band T performs delay and All Pass Filter processing, the frequency band above T has zero delay, and the filter processing only rotates the input signal by 90 degrees or -90 degrees. The value of T may be appropriately switched according to the input signal.

(Embodiment 3)
Hereinafter, a signal processing apparatus according to Embodiment 3 of the present invention will be described with reference to the drawings.

  FIG. 7 is a diagram showing the configuration of the signal processing apparatus according to the third embodiment. In FIG. 7, reference numeral 700 denotes decoding means, 701 denotes generation means, 702 denotes mixing coefficient determination means, and 703 denotes mixing means. In the second embodiment, decoding means 400, generation means 401, and mixing coefficient determination means 402 according to the second embodiment. , The same as the mixing means 403. The difference from the second embodiment is that a sound image moving unit 704 is arranged at the subsequent stage of the mixing unit 703.

The operation of the signal processing apparatus configured as described above will be described below.
In FIG. 7, the operations of the decoding unit 700, the generation unit 701, the mixing coefficient determination unit 702, and the mixing unit 703 are the same as those described in the second embodiment, and thus will be omitted.

  The two signals generated from the mixing unit 703 are processed by the sound image moving unit 704. This process is a process that applies a so-called head-related transfer function, and is a process that makes it feel as if sound is being produced in a wider space than the space surrounded by the speakers that are actually arranged.

First, the purpose of performing such processing will be described with reference to FIGS.
In the first place, the signal processing handled by the present application is to convert a signal that was originally a multi-channel signal downmixed into a few channels into a multi-channel signal from only the phase difference information and level ratio information between the original multiple channels. To separate. However, the original phase difference information and level ratio information alone do not completely return to the original state, but the original stereo signal is completely different from the original stereo signal. The signal does not return and becomes a mono signal. That is, a sound image generated by two speakers becomes like a sound image generated from speakers arranged at a narrow interval. FIG. 8 shows this, and the sound generated from the speaker arranged in a narrow space as drawn by the dotted line, even though the speaker drawn by the solid line is actually arranged. It sounds like

  Therefore, the purpose of the present application is to bring the sound image closer to the original stereo signal image by introducing a technique for expanding the sound image.

  FIG. 9 shows the case of 4 channels. When the front left channel signal and the rear left channel signal are downmixed, and the front right channel signal and the rear right channel signal are downmixed, the left and right separation is not impaired, but the front and rear separation Therefore, it is desired to recover the sense of separation that has been lost by generating a sound image as if the speaker of the front channel is arranged further forward and the speaker of the rear channel is arranged further rearward. Aim.

Now, the description returns to the operation of the third embodiment.
FIG. 10 is a diagram showing the concept of the head-related transfer function. When it is desired to localize the sound image so that the speaker is present at a position different from the position of the speaker actually arranged (when the sound image is localized at S and S ′ in FIG. 10), the sound image is received from the position where the sound image is to be localized. It is known that a sound image can be localized at a desired position by faithfully reproducing an acoustic transfer function up to a listener's ear, convoluted with a sound source signal, and presenting it to a listener. The transfer function (Hl (f, φ), etc.) in the path of the curved arrow shown in FIG. 10 is the head-related transfer function.

  FIG. 11 is a diagram showing structural features in the amplitude frequency characteristics of such a head-related transfer function, and FIG. 12 is a diagram showing interaural time differences and interaural level differences. It is already known that the clues related to the localization before and after the sound image and the top and bottom are in the peak and dip included in the amplitude frequency characteristic of the head-related transfer function shown in FIG. Further, it is already known that a clue related to the localization in the left-right direction is the time difference (ITD) and level difference (ILD) on the left and right of the head-related transfer function shown in FIG. (See Japanese Patent Application 2005-161602).

It is the operation of the sound image moving means 704 to perform processing of the head-related transfer function having such a characteristic amount. As an example, processing relating to localization of the sound image in the front-rear method will be described below.
As described above, it is known that the clues related to the localization in the front-rear direction of the sound image are the peak and dip included in the amplitude frequency characteristic of the head-related transfer function shown in FIG. On the other hand, the mixing unit 703 shown in FIG. 7 sends a plurality of frequency band signals to the sound image moving unit 704. Therefore, as shown in FIG. 13, the sound image moving unit 704 adjusts the amplitude levels of the plurality of frequency band signals input from the mixing unit 703 so as to match the amplitude frequency characteristics of the head-related transfer function.

  The curve shown in FIG. 13 is the same as the amplitude frequency characteristic of the head-related transfer function shown in FIG. 11, and the square indicated by the oblique line pattern increases the gain of the frequency band signal by the height. A square indicated by a lattice pattern indicates that the gain of the band signal is reduced by the height.

Here, it is not necessary to adjust the amplitude level so as to completely simulate the amplitude-frequency characteristic of the head-related transfer function, and only a few peaks and dips that are important for hearing can be simulated. Alternatively, only a few dips that are particularly important for hearing need be simulated. By doing so, the process of moving the sound image can be realized efficiently with a small amount of calculation.
FIG. 14 shows an example in which the bandwidth of each frequency band signal is wider than the width of the dip. In such a case, if the amplitude of the band signal is collectively reduced, the shape of the dip cannot be appropriately expressed, so by applying a filter having a predetermined frequency characteristic to the signal in the frequency band, A dip as shown in FIG. 14 can be formed. For example, by appropriately setting the value A of the filter of F (z) = 1 + A * z−1 + z−2 between −2 and 2, and applying the filter processing to the frequency band signal, A dip can be formed at a predetermined position. The curve shown by the red dotted line in FIG.

  In the explanation of the operation of the sound image moving means using FIGS. 13 and 14, the method for efficiently realizing the amplitude frequency characteristic of the head-related transfer function for a specific channel has been described. The processing of the partial transfer function is performed, and the processing of actually moving the sound image is performed by adding the outputs of the predetermined transfer functions. For this, a widely known method may be used. (See Japanese Patent Application No. 2005-161602).

  Of course, if the amplitude frequency characteristic of the head-related transfer function shown in FIGS. 13 and 14 is to move the sound image forward, the amplitude frequency characteristic of the head-related transfer function for the other channel is the sound image. Needless to say, if the sound is moved rearwardly, the feeling of expansion of the acoustic space is increased.

  As described above, according to the present embodiment, the first signal and the second signal generated from the first signal are mixed in two degrees of mixing (in the combination of h11 and h21). In the signal processing device that generates two signals by mixing in a combination of h12 and h22), the generating means for generating the second signal from the first signal, and the mixing A mixing coefficient determining means for determining the degree of the mixing means, a mixing means for mixing the first signal and the second signal based on the degree of mixing determined by the mixing coefficient determining means, and the mixing means And a sound image moving means for performing signal processing for moving the reproduced sound images of the two signals generated to separate positions, thereby generating a spatial signal when generating a 2ch signal from the monaural signal. Good spread and good At the same time stereo signal is obtained, sharp sheath temporal variation of the sound, becomes the localization can be realized with a solid sound image, moreover, so that the separation sense channel is further improved.

  In the present embodiment, it has been disclosed that the amplitude of the frequency band signal output from the mixing unit is changed as the sound image moving unit. In this case, the configuration shown in FIG. It may be taken. Each unit shown in FIG. 15 is the same as that shown in FIG. 7, but a mixing coefficient changing unit 1504 for changing the degree of mixing determined by the mixing coefficient determining unit 1502 is provided. As shown in FIG. 13, the mixing means 1503 has the same operation as the mixing means 703 shown in FIG. 7 by increasing or decreasing the mixing coefficient in advance for a predetermined frequency band. Regardless, the two generated signals have increased separation performance.

  In the third embodiment, the generation unit 701 is the same as that described in the second embodiment. However, from the gist of the invention described in the third embodiment, the generation unit 701 The operation 701 may be any operation. For example, it may be configured by a delay device and an AllPassFilter according to the conventional technique described with reference to FIG. 5, or the generation means 101 shown in the first embodiment (FIG. 1: details of the inside are shown in FIG. ). Further, it may be configured by only the third filter means 201 in FIG. The gist of the invention shown in the third embodiment is that the generated multi-channel signal is separated from a monaural signal by using information on the level ratio and phase difference as clues. This is because the sound image moving means 704 compensates for the insufficiency.

  Further, the sound image moving means 704 may be controlled so as to operate only when its function is necessary. When it is necessary is when the two signals to be separated are approximate. In particular, when the level ratio information indicates 0 dB or a value close thereto, the separation performance deteriorates. Therefore, in such a case, the sound image moving means 704 is used, and if not, the control is performed so that it is not used. May be.

  The signal processing apparatus according to the present invention can decode an encoded signal that expresses a phase difference or level ratio between a plurality of channels with a bit number that is not very low, while maintaining acoustic characteristics, so that music broadcasting at a low bit rate can be performed. It can be applied to services, music distribution services, and receiving devices.

It is a figure which shows the structure of the signal processing apparatus in this Embodiment 1. FIG. It is a figure which shows an example of a structure of a production | generation means. It is a figure explaining level ratio information and phase difference information using a parallelogram. It is a figure which shows the structure of the signal processing apparatus in this Embodiment 1. FIG. It is a figure which shows the basic composition of the prior art. It is a figure which shows the structure of the signal processing apparatus in embodiment of the structure which receives the encoding data which show an acoustic feature-value. It is a figure which shows the structure of the signal processing apparatus in this Embodiment 3. It is a figure which shows an example of the subject which this application solves in this Embodiment 3. FIG. It is a figure which shows another example of the subject which this application solves in this Embodiment 3. FIG. It is a figure which shows the basic concept of the signal processing for moving a sound image. It is a figure which shows the structural characteristic in the amplitude frequency characteristic of a head-related transfer function. It is a figure which shows the time difference between both ears of a head-related transfer function, and the level difference between both ears. It is a figure which shows an example of operation | movement of a sound image moving means. It is a figure which shows another example of operation | movement of a sound image moving means. It is a figure which shows another structure of the signal processing apparatus in this Embodiment 3.

Explanation of symbols

DESCRIPTION OF SYMBOLS 100 Decoding means 101 Generation means 102 Mixing coefficient determination means 103 Mixing means 104 Delay means 105 First filter means 106 Second filter means 107 Feature quantity detection means 108 Synthesis means 200 Second delay means 201 Third filter means 400 decoding means 401 generating means 402 mixing coefficient determining means 403 mixing means 109 feature quantity receiving means 700 decoding means 701 generating means 702 mixing coefficient determining means 703 mixing means 704 sound image moving means 1500 decoding means 1501 generating means 1502 mixing coefficient determination Means 1503 Mixing means 1504 Mixing coefficient changing means (sound image moving means)

Claims (21)

A signal processing device that generates two signals by mixing a first signal and a second signal generated from the first signal at two degrees of mixing,
Generating means for generating the second signal from the first signal;
Mixing coefficient determining means for determining the degree of mixing;
Mixing means for mixing the first signal and the second signal based on the degree of mixing determined by the mixing coefficient determining means;
The generating means includes
Delay means for delaying the first signal by N (N> 0) unit time;
Filter means for processing the output signal of the delay means;
Processing means for processing the first signal,
The signal processing apparatus, wherein the generation unit generates the second signal from an output signal of the filter unit and an output signal of the processing unit.   The generating unit includes a synthesizing unit that synthesizes the second signal from the output signal of the filter unit and the output signal of the processing unit in accordance with an acoustic feature amount of the first signal. The signal processing apparatus according to claim 1, wherein the target feature amount is a feature amount that becomes large when the first signal is abruptly fluctuating.   The generating unit includes a synthesizing unit that synthesizes the second signal from the output signal of the filter unit and the output signal of the processing unit in accordance with an acoustic feature amount of the first signal. The signal processing apparatus according to claim 1, wherein the target feature amount is a feature amount that becomes large when strong energy is concentrated in a specific frequency band in the first signal.   The synthesizing unit outputs an output signal of the filter unit when the feature amount is small, and outputs an output signal of the processing unit when the feature amount is large. The signal processing device according to any one of claims 1 to 3.   The synthesizing unit includes a second mixing unit that mixes the output signal of the filter unit and the output signal of the processing unit. When the feature amount is small, the output signal of the filter unit is increased. 4. The signal processing apparatus according to claim 1, wherein when the mixing is performed and the feature amount is large, the output signal of the processing unit is mixed in a large amount. 5.   The said processing means has a 2nd filter means, and the said 2nd filter means has the order of a filter less than the 1st filter means in the previous period, The any one of Claims 1-5 characterized by the above-mentioned. A signal processing device according to 1.   6. The processing device according to claim 1, wherein the processing unit includes a second delay unit, and the second delay unit has a delay amount smaller than that of the first delay unit. The signal processing apparatus as described.   8. The processing unit according to claim 1, wherein the processing unit includes a third filter unit, and the third filter unit is a process of rotating the phase of the input signal by 90 degrees or -90 degrees. The signal processing device according to claim 1.   The generating means is configured to be able to process a signal independently for a plurality of frequency components, and outputs a signal of the filter means for a signal in a low frequency band, and a signal in a high frequency band. The signal processing apparatus according to claim 1, wherein a signal of the processing means is output. The first signal is a signal obtained by downmixing two signals,
10. The mixing coefficient determining unit determines a degree of mixing from a value determined according to a level ratio L and a phase difference θ between the two original signals. The signal processing device according to claim 1. The mixing coefficient determining means is an angle obtained by dividing the parallelogram with the angle formed by the two adjacent sides being the θ and the length ratio being the L by the diagonal of the parallelogram. , A and B, and values d1 and d2 determined according to the level ratio L, d1 * cos (A), d1 * sin (A), d2 * cos (-B), d2 * sin (-B ), And the mixing means uses r1 as the real part when the first signal is expressed as a complex number, i1 as the imaginary part when the first signal is expressed as a complex number, and r2 as the real part when the second signal is expressed as a complex number. When the part is i2,
Let d1 * cos (A) * r1 + d1 * sin (A) * r2 be the real part of the first output signal,
Let d1 * cos (A) * i1 + d1 * sin (A) * i2 be the imaginary part of the first output signal,
Let d2 * cos (-B) * r1 + d2 * sin (-B) * r2 be the real part of the second output signal,
11. The signal according to claim 1, wherein d2 * cos (-B) * i1 + d2 * sin (-B) * i2 is an imaginary part of the second output signal. Processing equipment. The mixing coefficient determining means calculates the values of d1 and d2.
d1 = L / ((1 + 2 * L * cos (θ) + L * L) ^ 0.5), d2 = 1 / ((1 + 2 * L * cos (θ) + L * L) ^ 0.5) The signal processing device according to claim 1, wherein the signal processing device is obtained.   The mixing coefficient determining means calculates the values of d1 and d2 as d1 = L / ((1 + L * L) ^ 0.5), d2 = 1 / ((1 + L * L) ^ 0.5) The signal processing device according to claim 1, wherein the signal processing device is characterized in that:   The image processing apparatus further includes a feature amount receiving unit that receives data in which the acoustic feature amount is encoded, and the generation unit generates a signal according to the data in which the acoustic feature amount is encoded. The signal processing device according to any one of claims 1 to 13.   The data obtained by encoding the acoustic feature amount is 1-bit data, and the generation unit outputs an output signal of the processing unit when the data is true, and the data of the filter unit when the data is false. The signal processing apparatus according to claim 14, wherein an output signal is output. A signal processing device that generates two signals by mixing a first signal and a second signal generated from the first signal at two degrees of mixing,
Generating means for generating the second signal from the first signal;
Mixing coefficient determining means for determining the degree of mixing;
Mixing means for mixing the first signal and the second signal based on the degree of mixing determined by the mixing coefficient determining means;
Sound image moving means for performing processing for moving the reproduced sound images of the two signals generated from the mixing means to positions separated from each other;
A signal processing apparatus comprising:   17. The sound image moving means receives two signals generated from the mixing means and processes the signals to move the reproduced sound images of the two signals to positions separated from each other. Signal processing device.   The sound image moving means processes the degree of mixing determined by the mixing coefficient determining means so that the reproduced sound images of the two signals generated by the mixing means move to positions separated from each other. The signal processing apparatus according to claim 16. The sound image moving means includes a first processing means for performing a process for moving a reproduced sound image of the signal in a first direction on one of the two signals generated from the mixing means; 17. The apparatus according to claim 16, further comprising second processing means for performing a process for moving a reproduced sound image of the signal in a second direction that is opposite to the first direction. Signal processing device.   The first processing means performs processing to change the amplitude of a signal in a predetermined frequency band, and the second processing means performs processing to change the amplitude of a signal in a frequency band different from the frequency band. The signal processing apparatus according to claim 17.   When the two signals generated from the mixing unit are approximate, a signal generated from the sound image moving unit is output; otherwise, a signal generated from the mixing unit is output. The signal processing device according to any one of claims 16 to 18.

Images