JP7626660B2 - Sound processing device and program - Google Patents

Description

The present invention relates to an audio processing device and a program.

Conventionally, there is known a device that expands the dynamic range of an audio signal read from a CD or DVD.
For example, the sound processing device disclosed in Patent Document 1 includes an envelope detection means for detecting the envelope of an audio signal output from a sound source, a decibel conversion means for converting the envelope into decibels, a differentiation means for generating a change amount signal indicating the amount of change in the envelope, an inversion means for inverting the change amount signal, a linear conversion means for converting the inverted change amount signal into a linear signal to generate a detection signal, a complementary signal generation means for generating a complementary signal by subtracting a signal obtained by multiplying the audio signal by the detection signal from the audio signal, and an addition means for adding the complementary signal to the audio signal.

JP 2013-55384 A

However, the audio processing device disclosed in Patent Document 1 does not take into account the falling components. If the convergence of the speaker's time response is poor, the reverberation of the speaker may be added to the reverberation of the sound source, resulting in excessive reverberation.

The present invention has been made in consideration of the above circumstances, and aims to provide an audio processing device and program that can optimally control the reverberation of an audio signal.

In order to solve the above problem, the reverberation control device of the present invention includes a detection unit that detects an envelope signal of an input audio signal, a conversion unit that converts the detected envelope signal into decibels, a differentiation processing unit that performs differentiation processing on the converted envelope signal to generate a change amount signal that indicates the amount of change in the envelope signal, an inversion unit that inverts the signal level of the change amount signal, a first fluctuation detection signal generation unit that generates a first limiting signal that limits the gain of the inverted change amount signal to a signal of 0 dB or less and generates a first fluctuation detection signal in which the gain of the generated first limiting signal is indicated by an amplitude, and The audio signal processing device includes a second fluctuation detection signal generating unit that generates a second limiting signal by limiting the gain of the audio signal to 0 dB or more and generates a second fluctuation detection signal in which the gain of the generated second limiting signal is indicated by an amplitude, a limiting signal generating unit that multiplies the first fluctuation detection signal by the second fluctuation detection signal to generate a multiplied signal and generates a third limiting signal by limiting the amplitude of the generated multiplied signal to 1 or more, a first complementary signal generating unit that subtracts a signal obtained by multiplying the audio signal by the third limiting signal from the audio signal to generate a first complementary signal, and an adding unit that adds the first complementary signal to the audio signal.

According to the present invention, audio signals can be controlled to prevent excessive reverberation.

1 is a block diagram showing a schematic configuration of a sound processing device. 4 is a block diagram showing a schematic configuration of a first fluctuation detection unit; FIG. FIG. 4 is a block diagram showing a schematic configuration of a second fluctuation detection unit. FIG. 4 is a block diagram showing a schematic configuration of a first weighting unit. FIG. 4 is a block diagram showing a schematic configuration of a second weighting unit. FIG. 4 is a block diagram showing a schematic configuration of a third weighting unit. FIG. 2 is a diagram showing an example of a signal waveform of an input audio signal. FIG. 4 is a diagram showing an example of a signal waveform of a low-frequency audio signal. 2 shows an example of a signal waveform of a decibel envelope signal. FIG. 4 is a diagram showing an example of a signal waveform of a change amount signal; 13 is a diagram showing an example of a signal waveform of a change amount signal inverted by an inversion process. FIG. FIG. 4 is a diagram illustrating an example of a signal waveform of a first limiting signal. FIG. 13 is a diagram illustrating an example of a signal waveform of a second limiting signal. FIG. 4 is a diagram illustrating an example of a signal waveform of a first fluctuation detection signal. FIG. 13 is a diagram illustrating an example of a signal waveform of a second fluctuation detection signal. 13 is a diagram showing an example of a signal waveform of a first fluctuation detection signal that has been smoothed; FIG. 13 is a diagram showing an example of a signal waveform of a first fluctuation detection signal output by a first multiplier. FIG. 13 is a diagram showing an example of a signal waveform of a low-frequency attack weighting signal; FIG. 13 is a diagram showing an example of a signal waveform of a second fluctuation detection signal that has been smoothed. FIG. 13 is a diagram showing a signal waveform of a second fluctuation detection signal after multiplication processing output from a multiplier to a level limiting section. FIG. FIG. 13 is a diagram illustrating an example of a signal waveform of a third limiting signal. 13 shows an example of a signal waveform after multiplication by a third multiplier. 13 is a diagram showing an example of a signal waveform of a low-frequency reverberation weighting signal; FIG. 11 is a diagram showing a signal waveform of a signal obtained by adding a low-frequency reverberation complementary signal to a low-frequency audio signal. FIG. 13 is a diagram showing a signal waveform of a signal obtained when a low-frequency reverberation complementary signal generated based on a second fluctuation detection signal that has been subjected to repeat processing and smoothing processing is added to a low-frequency audio signal. FIG. 13 is a diagram showing a signal waveform of a signal obtained by adding a low-frequency reverberation complementary signal and a low-frequency attack complementary signal to a low-frequency audio signal. FIG. 4 is a flowchart showing the operation of the sound processing device. 13 is a flowchart showing a detailed flow of step S3. 13 is a flowchart showing a detailed flow of step S4. 13 is a flowchart showing details of step S7. FIG. 11 is a diagram illustrating a system configuration of a second embodiment.

[Overall configuration of the sound processing device 100]
FIG. 1 is a block diagram showing the configuration of a sound processing device 100. As shown in FIG.
The sound processing device 100 has a function of expanding the gain of an audio signal that has been smoothed by compressing the dynamic range, and a function of suppressing excessive reverberation in the low frequency range of the audio signal.

The sound processing device 100 includes an LPF (Low-pass filter) 110, a first fluctuation detection unit 120, a first weighting unit 140, a second weighting unit 150, a HPF (High-pass filter) 160, a second fluctuation detection unit 170, a third weighting unit 180, and a synthesis unit 190.
The synthesis unit 190 corresponds to an addition unit.

An audio signal with a compressed dynamic range is input to the sound processing device 100. This audio signal is, for example, an audio signal whose dynamic range has been compressed in conjunction with recording onto a CD (Compact Disc) or the like. The audio signal input to the sound processing device 100 is input to the LPF 110, the HPF 160, and the synthesis unit 190 as shown in FIG. 1.

The LPF 110 performs a filtering process on the input audio signal to extract low-frequency components of the audio signal. The low-frequency components of the audio signal extracted by the LPF 110 are hereinafter referred to as low-frequency audio signals.
The HPF 160 performs a filtering process on the input audio signal to extract high-frequency components of the audio signal. The high-frequency components of the audio signal extracted by the HPF 160 are hereinafter referred to as a high-frequency audio signal.

7 shows an example of a signal waveform of an audio signal input to the sound processing device 100. This audio signal is a reproduced audio signal recorded on a CD, and displays the first one second of a song.
FIG. 8 shows an example of a signal waveform of a low-frequency audio signal filtered by the LPF 110. In FIG.

The first variation detection unit 120 detects the envelope of the low-frequency audio signal input from the LPF 110. The first variation detection unit 120 detects the amount of change in the detected envelope, outputs a first variation detection signal indicating the amount of change in the detected envelope to the first weighting unit 140, and outputs a second variation detection signal indicating the amount of change in the envelope to the second weighting unit 150. A detailed explanation of the first variation detection signal and the second variation detection signal will be given later.

The second variation detection unit 170 detects the envelope of the high-frequency audio signal input from the HPF 160. The second variation detection unit 170 detects the amount of change in the detected envelope, and outputs a third variation detection signal indicating the amount of change in the detected amplitude to the third weighting unit 180.

The first weighting unit 140 generates a complementary signal for low-frequency attack based on the first fluctuation detection signal. The complementary signal for low-frequency attack is a signal that amplifies the amplitude of the low-frequency audio signal that has been smoothed by dynamic range compression. The complementary signal for low-frequency attack corresponds to the second complementary signal. The first weighting unit 140 outputs the generated complementary signal for low-frequency attack to the synthesis unit 190.

The second weighting unit 150 generates a low-frequency reverberation complementary signal based on the second variation detection signal. The low-frequency reverberation complementary signal is a signal that attenuates the reverberation of the low-frequency audio signal. The low-frequency reverberation complementary signal corresponds to the first complementary signal. The second weighting unit 150 outputs the generated low-frequency reverberation complementary signal to the synthesis unit 190.

The third weighting unit 180 generates a high-frequency attack complementary signal based on the third variation detection signal. The high-frequency attack complementary signal is a signal that amplifies the amplitude of the high-frequency audio signal that has been smoothed by dynamic range compression. The third weighting unit 180 outputs the generated high-frequency attack complementary signal to the synthesis unit 190.

The synthesis unit 190 adds a low-frequency attack complementary signal, a low-frequency reverberation complementary signal, and a high-frequency attack complementary signal to the input audio signal. This emphasizes the rising edge of the audio signal, improving the audible sense of sharpness, realism, and sound quality. It also speeds up the convergence of the reverberation of the falling edge of the low-frequency audio signal, suppressing excessive reverberation caused by the sound source, speakers, etc.

[Configuration of the first fluctuation detection unit 120 and the second fluctuation detection unit 170]
FIG. 2 is a block diagram showing a schematic configuration of the first fluctuation detection unit 120. As shown in FIG.
The configuration of the first fluctuation detection unit 120 will be described.
The first fluctuation detection unit 120 includes a first maximum value detection unit 121, a first maximum value hold unit 122, a first smoothing processing unit 123, a first decibel conversion unit 124, a first differential processing unit 125, a first inversion processing unit 126, a first level limiting unit 127, a first linear conversion unit 128, a second level limiting unit 129, and a second linear conversion unit 130.
The first maximum value detection section 121, the first maximum value hold section 122 and the first smoothing processing section 123 correspond to a detection section.
The first decibel converter 124 corresponds to a converter.
The first differential processing unit 125 corresponds to the differential processing unit.
The first inversion processing unit 126 corresponds to an inversion unit.
The first level limiting section 127 and the first linear conversion section 128 correspond to a first fluctuation detection signal generating section.
The second level limiting section 129 and the second linear conversion section 130 correspond to a second fluctuation detection signal generating section.

The low-frequency audio signal filtered by the LPF 110 is input to the first maximum value detection section 121 .
The first maximum value detection unit 121 converts the amplitude of the input low-frequency audio signal into an absolute value, and detects the maximum value of the absolute value of the amplitude of the low-frequency audio signal within a predetermined time. The first maximum value detection unit 121 outputs the maximum value for every m samples, which is the number of samples within the predetermined time, to the first maximum value hold unit 122. m is an arbitrary natural number.

The first maximum value hold section 122 receives a signal indicating the maximum value from the first maximum value detection section 121 .
The first maximum value hold unit 122 obtains the maximum value that is the largest among the past n maximum values including the maximum value indicated by the input signal. n is any natural number. In other words, the first maximum value hold unit 122 obtains the highest maximum value among n×m samples. The first maximum value hold unit 122 outputs a signal indicating the obtained maximum value to the first smoothing processing unit 123 at predetermined time intervals.
It should be noted that the first maximum value detection section 121 and the first maximum value hold section 122 have a preset detection range for detecting maximum values, and do not detect signal levels below the set detection range.

The first smoothing processing unit 123 performs integration processing on the signal indicating the maximum value input from the first maximum value hold unit 122, and performs smoothing processing to smooth fluctuations in the maximum value. By performing smoothing processing on the signal indicating the maximum value, an envelope signal indicating the envelope of the low-frequency audio signal is detected.

The first decibel converter 124 converts the envelope signal of the low-frequency audio signal obtained by the smoothing process into an envelope signal in decibels. The envelope signal input to the first decibel converter 124 has linear signal characteristics in which the gain is indicated by the amplitude. The decibel envelope signal converted by the first decibel converter 124 has signal characteristics in which the gain is indicated in decibel units. The first decibel converter 124 outputs the converted decibel envelope signal to the first differential processor 125.
FIG. 9 shows an example of a signal waveform of the decibel envelope signal output from the first decibel converter 124 to the first differential processor 125. In FIG.

The first differential processing unit 125 performs differential processing on the input decibel envelope signal to generate a change amount signal indicating the amount of change in the envelope signal. The change amount signal changes according to the change in the envelope signal. For example, when the amount of change in the envelope signal is large, the change amount signal indicates a large value, and when the amount of change in the envelope signal is small, the change amount signal indicates a small value. The first differential processing unit 125 outputs the generated change amount signal to the first inversion processing unit 126.
FIG. 10 shows an example of a signal waveform of the change amount signal.

The differentiation process in the first differentiation processing unit 125 is realized by performing a filtering process using a high-pass filter on the audio signal from which the low-frequency components have been extracted by the LPF 110. In addition, by changing the setting value of the normalized cutoff frequency in this high-pass filter, it is possible to adjust the change time of the amount of change obtained by the filtering process.

For example, the state in which the dynamic range is compressed and the amplitude is smoothed varies depending on the compressor and limiter settings, and the dynamic range compression state differs between CDs and DVDs. Therefore, by adjusting the normalized cutoff frequency according to the dynamic range compression state of the audio signal, it is possible to adjust the dynamic range expansion time to suit the dynamic range compression state.

The first inversion processing unit 126 performs an inversion process to invert the signal level of the input change amount signal. The first inversion processing unit 126 inverts the signal level, i.e., the positive and negative sides of the signal, by multiplying the change amount signal by -1. The first inversion processing unit 126 outputs the inverted change amount signal to the first level limiting unit 127 and the second level limiting unit 129.
11 shows an example of a signal waveform of a change amount signal whose signal level has been inverted by inversion processing. The change amount signal shown in FIG 11 and the change amount signal shown in FIG 10 have the positive and negative signals inverted.

The first level limiting unit 127 limits the signal level of the input change amount signal to a first range of -6 dB or more and 0 dB or less to generate a first limiting signal.
When the signal level of the input change amount signal is smaller than -6 dB, the first level limiting unit 127 converts the signal level of the change amount signal to -6 dB to generate a first limiting signal. When the signal level of the input change amount signal is larger than 0 dB, the first level limiting unit 127 converts the signal level of the change amount signal to 0 dB to generate a first limiting signal.
The first level limiting section 127 outputs the generated first limiting signal to the first linear conversion section 128. Fig. 12 shows an example of a signal waveform of the first limiting signal.

The second level limiting unit 129 limits the signal level of the input change amount signal to a second range of 0 dB or more and 6 dB or less to generate a second limiting signal.
If the signal level of the input change amount signal is greater than 6 dB, the second level limiting unit 129 converts the signal level of the change amount signal to 6 dB to generate a second limiting signal. If the signal level of the input change amount signal is less than 0 dB, the second level limiting unit 129 converts the signal level of the change amount signal to 0 dB to generate a second limiting signal.
The second level limiting section 129 outputs the generated second limiting signal to the second linear conversion section 130 .
FIG. 13 shows an example of a signal waveform of the second limiting signal.

The first linear conversion unit 128 converts the input first limiting signal into a signal having linear signal characteristics in which the gain of the first limiting signal is represented by an amplitude, and outputs the converted signal to the first weighting unit 140 as a first fluctuation detection signal.
FIG. 14 shows an example of a signal waveform of the first fluctuation detection signal.

The second linear conversion unit 130 converts the input second limiting signal into a signal having linear signal characteristics in which the gain of the second limiting signal is represented by an amplitude. The second linear conversion unit 130 outputs the converted signal to the second weighting unit 150 as a second fluctuation detection signal.
FIG. 15 shows an example of a signal waveform of the second fluctuation detection signal.

FIG. 3 is a block diagram showing a schematic configuration of the second fluctuation detection section 170. As shown in FIG.
The second fluctuation detection unit 170 includes a second maximum value detection unit 171, a second maximum value hold unit 172, a second smoothing processing unit 173, a second decibel conversion unit 174, a second differential processing unit 175, a second inversion processing unit 176, a third level limiting unit 177 and a third linear conversion unit 178.

The operations of the second maximum value detection unit 171, the second maximum value hold unit 172, the second smoothing processing unit 173, the second decibel conversion unit 174, the second differential processing unit 175, and the second inversion processing unit 176 are the same as those of the first fluctuation detection unit 120, so detailed explanations are omitted.

The third level limiting unit 177 limits the signal level of the change amount signal input from the second inversion processing unit 176 to a fourth range of −30 dB or more and 0 dB or less to generate a fourth limited signal.
The third level limiting unit 177 converts the signal level of the input change amount signal to −30 dB to generate a fourth limiting signal when the signal level of the input change amount signal is smaller than −30 dB, and converts the signal level of the input change amount signal to 0 dB to generate a fourth limiting signal when the signal level of the input change amount signal is larger than 0 dB.
The third level limiting section 177 outputs the generated fourth limiting signal to the third linear conversion section 178 .

The third linear conversion unit 178 converts the input fourth limiting signal into a signal having linear signal characteristics that indicate the gain of the fourth limiting signal by amplitude. The third linear conversion unit 178 outputs the converted signal to the third weighting unit 180 as a third fluctuation detection signal.

[First weighting section 140, second weighting section 150, and third weighting section 180]
4 is a block diagram showing a schematic configuration of the first weighting unit 140. The first weighting unit 140 includes a first repeat processing unit 141, a third smoothing processing unit 142, a first multiplier 143, a first subtractor 144, and a first gain adjustment unit 145.
The first multiplier 143 and the first subtractor 144 correspond to a second complementary signal generating unit.
The first gain adjustment section 145 corresponds to the gain adjustment section.

The first repeat processing unit 141 receives the first fluctuation detection signal from the first fluctuation detection unit 120. The first repeat processing unit 141 executes a repeat process for matching the frame length of the input first fluctuation detection signal with the frame length of the low-frequency audio signal output from the LPF 110.
For example, assume that the frame length of the low-frequency audio signal output from the LPF 110 is 256 samples, and the frame length of the first fluctuation detection signal is 32 samples. In this case, the first repeat processing unit 141 repeatedly outputs the same first fluctuation detection signal to the third smoothing processing unit 142, with 8 samples being 256/32.

The third smoothing unit 142 performs a smoothing process to smooth the signal waveform of the input first fluctuation detection signal by low-pass filtering. The third smoothing unit 142 outputs the first fluctuation detection signal after the smoothing process to the first multiplier 143 and the second weighting unit 150.
FIG. 16 shows an example of a signal waveform of the first fluctuation detection signal after the smoothing process.

The first multiplier 143 receives the low-frequency audio signal output by the LPF 110 and the first fluctuation detection signal that has been subjected to the smoothing process.
The first multiplier 143 performs a multiplication process to multiply the input low-frequency audio signal by the smoothed first fluctuation detection signal. The first multiplier 143 outputs the multiplied first fluctuation detection signal to the first subtractor 144. By the multiplication process of the first multiplier 143, a portion of the low-frequency audio signal where the signal level changes from a low level to a high level is amplified in accordance with the detection amount of the first fluctuation detection signal.
FIG. 17 shows an example of a signal waveform of the first fluctuation detection signal output by the first multiplier 143.

The first subtractor 144 receives the low-frequency audio signal output by the LPF 110 and the multiplied first fluctuation detection signal output by the first multiplier 143 .
The first subtractor 144 subtracts the multiplied first fluctuation detection signal from the input low-frequency audio signal.
By performing subtraction processing in the first subtractor 144, the signal level difference between the low-frequency audio signal before amplification and the audio signal after amplification is calculated as a weighting amount at the point in the low-frequency audio signal where the signal level changes from a low level to a high level.
The first subtractor 144 outputs a signal indicating the calculated weighting amount to the first gain adjustment section 145 as a low-frequency attack weighting signal.
FIG. 18 shows an example of a signal waveform of a low-frequency attack weighting signal.

The first gain adjustment unit 145 amplifies and attenuates the amplitude of the input low-frequency attack weighted signal to adjust the weighting amount, and generates a low-frequency attack complement signal. The first gain adjustment unit 145 outputs the generated first low-frequency complement signal to the synthesis unit 190.

The low-frequency attack weighting signal is generated based on the change in the maximum value in the envelope signal, so it is possible to expand the gain in the areas where the dynamic range has been compressed and smoothed. However, the gain of the dynamic range to be expanded differs for each audio signal whose dynamic range has been compressed and smoothed, and adding the low-frequency attack weighting signal to the audio signal as is may result in insufficient expansion of the dynamic range, or the amount of expansion of the dynamic range may be too great, which may actually result in a deterioration of sound quality.

For this reason, the first gain adjustment unit 145 amplifies and attenuates the gain of the low-frequency attack weighting signal in advance depending on the compression state of the audio signal, etc., to generate a first low-frequency complement signal with an adjusted weighting amount, making it possible to appropriately expand and adjust the dynamic range of the signal obtained by adding the first low-frequency complement signal to the audio signal in the synthesis unit 190.

The first limiting signal is a signal that limits the gain of the change amount signal to 0 dB or less. A first fluctuation detection signal that indicates the gain of this first limiting signal by amplitude is generated in the first linear conversion unit 128. Therefore, the first fluctuation detection signal is a signal with an amplitude of 1 or less.
The second limiting signal is a signal in which the gain of the change amount signal is limited to 0 dB or more. A second fluctuation detection signal indicating the gain of the second limiting signal by its amplitude is generated in the second linear conversion unit 130. Therefore, the second fluctuation detection signal is a signal with an amplitude of 1 or more.
If the complementary signal for low-frequency attack output from the first subtractor 144 is "Y2," the audio signal input to the first multiplier 143 and the first subtractor 144 is "S1," and the first fluctuation detection signal input to the first multiplier 143 is "W2," then the complementary signal for low-frequency attack Y2 is calculated by the following formula.
Y2=S1-(W2×S1)=(1-W2)×S1
Since the first fluctuation detection signal W2 is a signal with an amplitude of 1 or less, 1-W2 is 0 or a positive value, and the complementary signal for low frequency attack Y2 is a signal in phase with the audio signal S1. Therefore, by adding the in-phase complementary signal for low frequency attack Y2 to the audio signal S1, the amplitude of the audio signal S1 can be amplified by the complementary signal for low frequency attack Y2.

[Second weighting unit 150]
5 is a block diagram showing a schematic configuration of the second weighting unit 150. The second weighting unit 150 includes a second repeat processing unit 151, a fourth smoothing processing unit 152, a second multiplier 153, a fourth level limiting unit 154, a third multiplier 155, a second subtractor 156, and a second gain adjustment unit 157.
The second multiplier 153 and the fourth level limiting section 154 correspond to a limiting signal generating section.
The third multiplier 155 and the second subtractor 156 correspond to a first complementary signal generating unit.
The second gain adjustment section 157 corresponds to the gain adjustment section.

The second repeat processing unit 151 receives the second fluctuation detection signal from the first fluctuation detection unit 120. The second repeat processing unit 151 performs repeat processing to match the frame length of the input second fluctuation detection signal to the frame length of the low-frequency audio signal output from the LPF 110. The operation of the second repeat processing unit 151 is the same as that of the first repeat processing unit 141, so a detailed description will be omitted.

The fourth smoothing processor 152 performs a smoothing process by low-pass filtering to smooth the signal waveform of the second fluctuation detection signal input from the second repeat processor 151. The fourth smoothing processor 152 outputs the smoothed second fluctuation detection signal to the second multiplier 153.
FIG. 19 shows an example of a signal waveform of the second fluctuation detection signal after the smoothing process.

The second multiplier 153 receives the first fluctuation detection signal output by the third smoothing processor 142 of the first weighting unit 140 and the second fluctuation detection signal output by the fourth smoothing processor 152 .
The second multiplier 153 multiplies the input first fluctuation detection signal and the input second fluctuation detection signal to generate a multiplied signal. The second multiplier 153 outputs the generated multiplied signal to the fourth level limiting section 154. Fig. 20 shows an example of a signal waveform of the multiplied signal output by the second multiplier 153.

The fourth level limiting unit 154 generates a third limited signal by limiting the amplitude of the input multiplied signal to a third range of 1 or more.
The fourth level limiting unit 154 converts the amplitude of the second fluctuation detection signal, which has an amplitude of 1 or less, to 1 to generate a third limiting signal. The fourth level limiting unit 154 outputs the third limiting signal, whose amplitude is limited to 1 or more, to the third multiplier 155.

FIG. 21 shows an example of a signal waveform of the third limiting signal.
In the rising edge section (0.2 seconds to 0.4 seconds) of the audio signal shown in range a of the smoothed second fluctuation detection signal shown in Fig. 19, the amplitude gradually converges to 1, but in the rising edge section (0.2 seconds to 0.4 seconds) of the audio signal shown in range b of the third limiting signal shown in Fig. 21, the amplitude is sharply limited to 1. This is achieved by multiplying the second fluctuation detection signal by the first fluctuation detection signal.
By sharply limiting the amplitude to 1 at the rising edge of the audio signal, it is possible to suppress a reduction in the sense of attack at the rising edge of the audio signal due to the low-frequency reverberation complement signal.

The third multiplier 155 receives the low-frequency audio signal output by the LPF 110 and the third limited signal.
The third multiplier 155 performs a multiplication process of multiplying the low-frequency audio signal by the third limiting signal. The third multiplier 155 outputs the multiplied signal to the second subtractor 156. Fig. 22 shows an example of a signal waveform of the signal multiplied by the third multiplier 155.

The second subtractor 156 receives the low-frequency audio signal output by the LPF 110 and the output signal of the third multiplier 155 .
The second subtractor 156 performs subtraction processing to subtract the output signal of the third multiplier 155 from the low-frequency audio signal. By performing subtraction processing in the second subtractor 156, a signal level difference between the low-frequency audio signal before amplification and the audio signal after amplification is calculated as a weighting amount at a portion of the low-frequency audio signal where the signal level changes from a high level to a low level.
The second subtractor 156 outputs a signal indicating the calculated weighting amount to the second gain adjustment section 157 as a low-frequency reverberation weighting signal.
FIG. 23 shows an example of a signal waveform of a low-frequency reverberation weighting signal.

The second gain adjustment unit 157 amplifies and attenuates the amplitude of the input low-frequency reverberation weighting signal to adjust the weighting amount, and generates a low-frequency reverberation complement signal. The second gain adjustment unit 157 outputs the generated low-frequency reverberation complement signal to the synthesis unit 190.

The first limiting signal is a signal in which the gain of the change amount signal is limited to 0 dB or less. A first fluctuation detection signal indicating the gain of this first limiting signal by amplitude is generated by the first linear conversion unit 128. Therefore, the first fluctuation detection signal is a signal with an amplitude of 1 or less.
The second limiting signal is a signal in which the gain of the change amount signal is limited to 0 dB or more. A second fluctuation detection signal indicating the gain of the second limiting signal by its amplitude is generated by the second linear conversion unit 130. Therefore, the second fluctuation detection signal is a signal with an amplitude of 1 or more.
Furthermore, a third limited signal, which is obtained by limiting the amplitude of a multiplied signal obtained by multiplying the first fluctuation detection signal by the second fluctuation detection signal to 1 or more, becomes a signal with an amplitude of 1 or more.
If the low-frequency reverberation complementary signal output from the second subtractor 156 is "Y1," the audio signal input to the third multiplier 155 and the second subtractor 156 is "S1," and the third limiting signal input to the third multiplier 155 is "W1," the low-frequency reverberation complementary signal Y1 is calculated by the following formula.
Y1=S1-(W1×S1)=(1-W1)×S1
Since the third limiting signal W1 is a signal with an amplitude of 1 or more, 1-W1 is 0 or a negative value, and the low-frequency reverberation complementary signal Y1 is a signal with an opposite phase to the audio signal S1. Therefore, by adding the opposite-phase low-frequency reverberation complementary signal Y1 to the audio signal S1, the amplitude of the audio signal S1 can be attenuated by the low-frequency reverberation complementary signal Y1.

Figure 6 is a block diagram showing a schematic configuration of the third weighting unit 180. The third weighting unit 180 includes a third repeat processing unit 181, a fifth smoothing processing unit 182, a fourth multiplier 183, a third subtractor 184, and a third gain adjustment unit 185.

The third weighting unit 180 has functional units similar to those of the first weighting unit 140. The third weighting unit 180 subtracts the high-frequency audio signal that has been subjected to an amplification process corresponding to the detection amount indicated by the third fluctuation detection signal from the high-frequency audio signal input from the third subtractor 184 and the HPF 160, thereby determining the weighting amount for the high-frequency audio signal.
The third weighting unit 180 generates a high-frequency complement signal by adjusting the determined weighting amount in a third gain adjustment unit 185. The configuration and roles of the functional units in the third weighting unit 180 are the same as those of the first weighting unit 140 shown in FIG. 4, and therefore will not be described here.

[Composition section]
The synthesis unit 190 adds the low-frequency attack complement signal obtained by the first weighting unit 140, the low-frequency reverberation complement signal obtained by the second weighting unit 150, and the high-frequency complement signal obtained by the third weighting unit 180 to the audio signal input to the sound processing device 100.
By performing the addition process, audio signals of low and high frequency components that have been gain-complemented according to the amount of change in the signal envelope are added to the audio signal of the sound source that has been smoothed by compressing the dynamic range, thereby expanding the dynamic range and improving the sense of clarity, realism, and sound quality perceived by the ear.
Furthermore, by adding the low-frequency reverberation complement signal to the audio signal, the reverberation of the low-frequency audio signal can be attenuated.

The sound processing device 100 according to this embodiment does not perform dynamic range expansion processing on mid-frequency components. This is because mid-frequency components are generally highly sensitive to hearing and contain many audio components such as vocals, and expanding the dynamic range of mid-frequency components may cause the listener to feel a sense of fluctuation in hearing. However, the sound processing device 100 according to this embodiment only shows an example of a case in which dynamic range expansion processing is performed on low-frequency components and high-frequency components, and it is also possible to configure the sound processing device 100 according to this embodiment to perform dynamic range expansion processing on mid-frequency components as well. Furthermore, it is also possible to perform dynamic range expansion processing using various combinations, such as only low-frequency components, only mid-frequency components, only high-frequency components, only low-frequency components and mid-frequency components, and only high-frequency components and mid-frequency components.

Fig. 24 shows a signal waveform of a signal obtained by adding a low-frequency reverberation complementary signal to a low-frequency audio signal. The signal shown in Fig. 24 does not include a low-frequency attack complementary signal and a high-frequency complementary signal. The dotted envelope in Fig. 24 shows a signal when the low-frequency reverberation complementary signal is not added to the low-frequency audio signal, that is, the envelope of the low-frequency audio signal.
24, the amplitude of the signal waveform is smaller than the envelope after 0.5 seconds, which indicates that the signal is attenuating.

25 shows a signal waveform of a signal when the second weighting section 150 does not include the second multiplier 153 and the fourth level limiting section 154, and the second fluctuation detection signal that has been subjected to the repeat processing and smoothing processing is directly input to the third multiplier 155, and a low-frequency reverberation complementary signal generated by this is added to the low-frequency audio signal. That is, the signal waveform of the low-frequency audio signal is shown when the first fluctuation detection signal is not multiplied by the second fluctuation detection signal. In this case, too, the low-frequency attack complementary signal and the high-frequency complementary signal are not added to the low-frequency audio signal. The envelope shown by the dashed line in FIG. 25 indicates the envelope of the low-frequency audio signal.
In the signal waveform shown in Figure 25, the amplitude is smaller than the envelope after 0.5 seconds, but the amplitude attenuates at the rising edge of the signal, for example, in the range c surrounded by a solid line in Figure 25, and the sense of attack of the rising edge of the signal is reduced.

26 shows a signal waveform of a signal obtained by adding a low-frequency reverberation complementary signal and a low-frequency attack complementary signal to a low-frequency audio signal. The envelope curve shown by the dashed line in FIG. 26 indicates the envelope curve of the low-frequency audio signal.
As shown in FIG. 26, by adding a low-frequency reverberation complementary signal and a low-frequency attack complementary signal to a low-frequency audio signal, the audible sense of sharpness, realism, and sound quality can be improved, and by speeding up the convergence time of the fall of the low-frequency audio signal, excessive reverberation caused by the sound source, speakers, etc. can be suppressed.

27 to 30 are flowcharts showing the operation of the sound processing device 100.
The operation of the sound processing device 100 will be described with reference to the flowcharts shown in FIGS.
First, when an audio signal is input to the sound processing device 100 (step S<b>1 ), the sound processing device 100 processes the input audio signal using the LPF 110 and the HPF 160 .
The LPF 110 filters the input audio signal to generate a low-frequency audio signal (step S2), while the HPF 160 filters the input audio signal to generate a high-frequency audio signal (step S3).
The LPF 110 outputs the generated low-frequency audio signal to the first weighting section 140 , the first fluctuation detection section 120 , and the second weighting section 150 .
The HPF 160 outputs the generated high-frequency audio signal to the second fluctuation detection section 170 and the third weighting section 180 .

The first fluctuation detection unit 120 generates a first fluctuation detection signal and a second fluctuation detection signal based on the input low-frequency audio signal (step S4). The first fluctuation detection unit 120 outputs the generated first fluctuation detection signal to the first weighting unit 140, and outputs the generated second fluctuation detection signal to the second weighting unit 150. Details of step S4 will be described with reference to the flowchart in FIG. 28.

The second variation detection unit 170 also generates a third variation detection signal based on the input high-frequency audio signal (step S5). The second variation detection unit 170 outputs the generated third variation detection signal to the third weighting unit 180.

Next, the first weighting unit 140 generates a low-frequency attack complementary signal based on the input low-frequency audio signal and the first fluctuation detection signal (step S6). The first weighting unit 140 outputs the generated low-frequency attack complementary signal to the synthesis unit 190. Details of step S6 will be described with reference to the flowchart in FIG. 29.

The second weighting unit 150 also generates a low-frequency reverberation complementary signal based on the input low-frequency audio signal and the second fluctuation detection signal (step S7). The second weighting unit 150 outputs the generated low-frequency reverberation complementary signal to the synthesis unit 190. Details of step S7 will be described with reference to the flowchart in FIG. 30.

Furthermore, the third weighting unit 180 generates a high-frequency attack complementary signal based on the input high-frequency audio signal and the third variation detection signal (step S8). The third weighting unit 180 outputs the generated high-frequency attack complementary signal to the synthesis unit 190.

Next, the synthesis unit 190 synthesizes the input audio signal with the low-frequency attack complementary signal, the low-frequency reverberation complementary signal, and the high-frequency attack complementary signal (step S9), and outputs the synthesized signal to a downstream processing circuit (step S10). The downstream processing circuit corresponds to, for example, the D/A converter 12 and the amplifier 13 shown in FIG. 31. Step S9 corresponds to the "step of adding the first complementary signal to the audio signal."

FIG. 28 is a flowchart showing the details of step S4.
The operation of the first fluctuation detection unit 120 will be described with reference to the flowchart shown in FIG.
When the low-frequency audio signal is input from the LPF 110, the first variation detection unit 120 detects a maximum value of the low-frequency audio signal by the first maximum value detection unit 121, and detects an envelope signal by holding the detected maximum value of the low-frequency audio signal by the first maximum value hold unit 122 (step S401). Step S401 corresponds to a "step of detecting an envelope signal of an input audio signal".

Next, the first fluctuation detection unit 120 smoothes the detected envelope signal using the first smoothing processing unit 123, and converts the smoothed envelope signal into an envelope signal in decibels (step S402). Step S402 corresponds to the "step of converting the envelope signal into decibels."

Next, the first fluctuation detection unit 120 performs differentiation processing on the decibel envelope signal to generate a change amount signal (step S403). Step S403 corresponds to the "step of generating a change amount signal indicating the amount of change in the envelope signal".
Next, the first fluctuation detection unit 120 performs inversion processing by multiplying the generated change amount signal by −1 using the first inversion processing unit 126 (step S404). Step S404 corresponds to a "step of inverting the signal level of the change amount signal."

Next, the first fluctuation detection unit 120 generates a first limited signal by limiting the signal range of the inverted change amount signal to a first range by the first level limiting unit 127, and generates a second limited signal by limiting the signal range to a second range by the second level limiting unit 129 (step S405). Step S405 corresponds to "a procedure for generating a first limited signal in which the gain of the change amount signal is limited to a signal of 0 dB or less" and "a procedure for generating a second limited signal in which the gain of the change amount signal is limited to 0 dB or more". The first range is a range of -6 dB or more and 0 dB or less. The second range is a range of 0 dB or more and 6 dB or less.

Next, the first fluctuation detection unit 120 linearly converts the first limiting signal using the first linear conversion unit 128 to generate a first fluctuation detection signal, and linearly converts the second limiting signal using the second linear conversion unit 130 to generate a second fluctuation detection signal (step S406). Step S406 corresponds to "a procedure for generating a first fluctuation detection signal in which the gain of the first limiting signal is indicated by an amplitude" and "a procedure for generating a second fluctuation detection signal in which the gain of the second limiting signal is indicated by an amplitude".

FIG. 29 is a flowchart showing the details of step S6.
The operation of the first weighting section 140 will be described with reference to the flowchart shown in FIG.
First, when the first weighting unit 140 receives the first fluctuation detection signal from the first fluctuation detection unit 120, the first weighting unit 140 repeats the input first fluctuation detection signal in the first repeat processing unit 141, and smoothes the first fluctuation detection signal after the repeat processing in the third smoothing processing unit 142 (step S601). The first weighting unit 140 outputs the first fluctuation detection signal after the repeat processing and smoothing processing to the first multiplier 143 and the second weighting unit 150.

Next, the first weighting unit 140 multiplies the first fluctuation detection signal that has been subjected to the repeat process and the smoothing process by the low-frequency audio signal by the first multiplier 143 (step S602).
Next, the first weighting unit 140 subtracts the signal multiplied by the first multiplier 143 from the low-frequency audio signal to generate a low-frequency attack weighted signal (step S603).

Next, the first weighting unit 140 adjusts the gain of the weighted signal using the first gain adjustment unit 145 to generate a complementary signal for low-frequency attack (step S604). Next, the first weighting unit 140 outputs the generated complementary signal for low-frequency attack to the synthesis unit 190 (step S605).

FIG. 30 is a flowchart showing the details of step S7.
The operation of the second weighting section 150 will be described with reference to the flowchart shown in FIG.
First, when the second weighting unit 150 receives a second fluctuation detection signal from the first fluctuation detection unit 120, it repeats the input second fluctuation detection signal in the second repeat processing unit 151, and smoothes the repeat-processed second fluctuation detection signal in the fourth smoothing processing unit 152 (step S701).

Next, the second weighting unit 150 multiplies the second fluctuation detection signal that has been subjected to the repeating process and the smoothing process by the first fluctuation detection signal input from the first weighting unit 140 (step S702). Step S702 corresponds to the "step of multiplying the first fluctuation detection signal by the second fluctuation detection signal to generate a multiplied signal."
Next, the second weighting unit 150 generates a third limited signal by limiting the signal range of the product signal multiplied by the second multiplier 153 to a third range by the fourth level limiting unit 154 (step S703). The third range is a range in which the amplitude is 1 or more. Step S703 corresponds to a "step of generating a third limited signal in which the amplitude of the product signal is limited to 1 or more."

Next, the second weighting unit 150 multiplies the low-frequency audio signal by the third limiting signal using the third multiplier 155 (step S704), and subtracts the signal multiplied by the third multiplier 155 from the low-frequency audio signal using the second subtractor 156 to generate a weighting signal for reverberation (step S705). Step S705 corresponds to "the step of subtracting the signal obtained by multiplying the audio signal by the third limiting signal from the audio signal to generate a first complementary signal."

Next, the second weighting unit 150 adjusts the gain of the weighted signal for reverberation using the second gain adjustment unit 157 to generate a low-frequency reverberation complement signal (step S706). Next, the second weighting unit 150 outputs the generated low-frequency reverberation complement signal to the synthesis unit 190 (step S707).

As described above, the sound processing device 100 of this embodiment includes the first variation detection section 120 , the second weighting section 150 , and the synthesis section 190 .
The first fluctuation detection unit 120 includes a first maximum value detection unit 121, a first maximum value hold unit 122, a first decibel conversion unit 124, a first differential processing unit 125, a first inversion processing unit 126, a first level limiting unit 127, a first linear conversion unit 128, a second level limiting unit 129 and a second linear conversion unit 130.

The first maximum value detection section 121 and the first maximum value hold section 122 detect the envelope signal of the input audio signal.
The first decibel converter 124 converts the detected envelope signal into a decibel signal.
The first differential processing unit 125 performs differential processing on the decibel-converted envelope signal to generate a change amount signal indicating the amount of change in the envelope signal.
The first inversion processing unit 126 inverts the signal level of the generated change amount signal.
The first level limiting section 127 generates a first limiting signal by limiting the gain of the inverted change amount signal to a signal equal to or lower than 0 dB.
The first linear conversion section 128 generates a first fluctuation detection signal in which the gain of the generated first limit signal is indicated by an amplitude.
The second level limiting unit 129 and the second linear conversion unit 130 generate a second limiting signal that limits the gain of the inverted change amount signal to 0 dB or more, and generate a second fluctuation detection signal in which the gain of the generated second limiting signal is indicated by its amplitude.

The second weighting section 150 includes a second multiplier 153 , a fourth level limiting section 154 , a third multiplier 155 and a second subtractor 156 .
The second multiplier 153 multiplies the first fluctuation detection signal by the second fluctuation detection signal to generate a multiplied signal.
The fourth level limiting section 154 generates a third limited signal by limiting the amplitude of the generated multiplied signal to 1 or more.
A third multiplier 155 multiplies the audio signal by the third limited signal.
The second subtractor 156 subtracts the signal obtained by multiplying the audio signal by the third limiting signal from the audio signal to generate a low-frequency reverberation complementary signal.
The synthesis unit 190 adds the low-frequency reverberation complementary signal to the audio signal.

The second limiting signal is a signal in which the gain of the change amount signal is limited to 0 dB or more. The second linear conversion unit 130 generates a second fluctuation detection signal that indicates the gain of the second limiting signal by its amplitude. Therefore, the second fluctuation detection signal has an amplitude of 1 or more.
The first limiting signal is a signal in which the gain of the change amount signal is limited to 0 dB or less. The first linear conversion unit 128 generates a first fluctuation detection signal that indicates the gain of the first limiting signal by its amplitude. Therefore, the first fluctuation detection signal has an amplitude of 1 or less.
Furthermore, a third limited signal, which is obtained by limiting the amplitude of a multiplied signal obtained by multiplying the first fluctuation detection signal by the second fluctuation detection signal to 1 or more, becomes a signal with an amplitude of 1 or more.
If the low-frequency reverberation complementary signal output from the second subtractor 156 is "Y1," the audio signal input to the third multiplier 155 and the second subtractor 156 is "S1," and the third limiting signal input to the third multiplier 155 is "W1," the low-frequency reverberation complementary signal Y1 is calculated by the following formula.
Y1=S1-(W1×S1)=(1-W1)×S1
Since the third limiting signal W1 is a signal with an amplitude of 1 or more, 1-W1 is 0 or a negative value, and the low-frequency reverberation complementary signal Y1 is a signal with an opposite phase to the audio signal S1. Therefore, by adding the opposite-phase low-frequency reverberation complementary signal Y1 to the audio signal S1, the amplitude of the audio signal S1 can be attenuated by the low-frequency reverberation complementary signal Y1.

The sound processing device 100 also includes a first weighting unit 140 .
The first weighting section 140 includes a first multiplier 143 and a first subtractor 144 .
The first multiplier 143 multiplies the audio signal by the first fluctuation detection signal.
The first subtractor 144 subtracts a signal obtained by multiplying the audio signal by the first fluctuation detection signal from the audio signal to generate a complementary signal for low-frequency attack.
The synthesis section 190 adds the low-frequency attack complementary signal and the low-frequency reverberation complementary signal to the audio signal.

The first limited signal is a signal whose amplitude is 1 or less, as described above.
If the complementary signal for low-frequency attack output from the first subtractor 144 is "Y2," the audio signal input to the first multiplier 143 and the first subtractor 144 is "S1," and the first fluctuation detection signal input to the first multiplier 143 is "W2," then the complementary signal for low-frequency attack Y2 is calculated by the following formula.
Y2=S1-(W2×S1)=(1-W2)×S1
Since the first fluctuation detection signal W2 is a signal with an amplitude of 1 or less, 1-W2 is 0 or a positive value, and the complementary signal for low frequency attack Y2 is a signal in phase with the audio signal S1. Therefore, by adding the in-phase complementary signal for low frequency attack Y2 to the audio signal S1, the amplitude of the audio signal S1 can be amplified by the complementary signal for low frequency attack Y2.

The sound processing device 100 also includes a second gain adjustment section 157 that adjusts the gain of the low-frequency reverberation complementary signal, and a first gain adjustment section 145 that adjusts the gain of the low-frequency attack complementary signal.
The synthesis unit 190 adds the low-frequency reverberation complementary signal and the low-frequency attack complementary signal, the gains of which have been adjusted by the second gain adjustment unit 157 and the first gain adjustment unit 145, to the audio signal.
Therefore, by generating a low-frequency reverberation complementary signal and a low-frequency attack complementary signal with adjusted weighting amounts, it is possible to appropriately adjust the amplification rate and attenuation rate of the amplitude of the audio signal.

As the differentiation process, the first differential processing unit 125 performs filtering process on the audio signal using a high-pass filter.
Furthermore, the change time of the change amount of the envelope signal obtained by the filtering process is adjusted by changing the setting value of the normalized cutoff frequency in the high-pass filter.
Therefore, it is possible to adjust the dynamic range expansion time in accordance with the state of dynamic range compression.

[Second embodiment]
FIG. 31 is a diagram showing a system configuration of the second embodiment.
The second embodiment includes a sound source device 11, a sound processing device 200, a D/A converter 12, an amplifier 13, and a speaker 14.

The sound source device 11 plays back the content and outputs an audio signal. The content includes, for example, audio sources recorded on a CD, DVD, BD (Blu-ray (registered trademark) Disc), hard disk, or semiconductor memory. The content may also be a streaming audio signal received by the sound source device 11 from an external device such as a server, when the sound source device 11 is equipped with a wireless communication unit.

The D/A converter 12 receives the audio signal processed by the sound processing device 200. The D/A converter 12 converts the input digital audio signal into an analog signal and outputs the converted analog audio signal to the amplifier 13.
The amplifier 13 amplifies the input audio signal and outputs the amplified audio signal to the speaker 14 .

The sound processing device 200 is a computer device equipped with a control unit 230 having a memory 210 and a processor 220. The control unit 230 controls the overall operation of the sound processing device 200 by the processor 220 operating in accordance with a control program 211 stored in the memory 210.

The memory 210 includes semiconductor memories such as a RAM (Random Access Memory) and a ROM (Read Only Memory). The RAM is used for temporary storage of various data, and the ROM stores a control program 211 for controlling the operation of the sound processing device 200, various setting information, etc.

The processor 220 is an arithmetic processing device that is configured with a CPU (Central Processing Unit) and an MPU (Micro Processing Unit). The processor 220 executes the control program 211 to control each part of the sound processing device 200. The processor 220 may be configured with a single processor, or may be configured with multiple processors. The processor 220 may also be configured with a SoC (System-on-a-chip) that is integrated with part or all of the memory 210 and other circuits. The processor 220 may also be configured with a combination of a CPU that executes a program and a DSP that executes a specified arithmetic processing. Furthermore, all of the functions of the processor 220 may be implemented in hardware, or may be configured using a programmable device.

When the processor 220 executes the control program 211, the control unit 230 functions as the LPF 110, the first fluctuation detection unit 120, the first weighting unit 140, the second weighting unit 150, the HPF 160, the second fluctuation detection unit 170, the third weighting unit 180, and the synthesis unit 190 shown in FIG. 1.

The control unit 230 executes the steps of detecting an envelope signal of the input audio signal, converting the detected envelope signal into decibels, differentiating the converted envelope signal to generate a change amount signal indicating the change amount of the envelope signal, inverting the signal level of the generated change amount signal, generating a first limiting signal by limiting the gain of the inverted change amount signal to a signal of 0 dB or less, generating a first fluctuation detection signal in which the gain of the generated first limiting signal is indicated by an amplitude, generating a second limiting signal by limiting the gain of the inverted change amount signal to 0 dB or more, generating a second fluctuation detection signal in which the gain of the generated second limiting signal is indicated by an amplitude, multiplying the first fluctuation detection signal by the second fluctuation detection signal, generating a third limiting signal by limiting the amplitude of the signal obtained by multiplying the first fluctuation detection signal by the second fluctuation detection signal to 1 or more, subtracting the signal obtained by multiplying the audio signal by the third limiting signal from the audio signal to generate a first complementary signal, and adding the first complementary signal to the audio signal.

The above-described embodiment is merely an example of one aspect of the present invention, and any modification or application is possible without departing from the spirit and scope of the present invention.

For example, the configuration of the sound processing device 100 shown in Figures 1 to 6 is a schematic diagram showing the functions of these devices classified according to the main processing content. The configuration of the sound processing device 100 can also be divided into more blocks depending on the processing content. Furthermore, these functional blocks may be configured so that even more processing is executed by one block shown in Figures 1 to 6. Furthermore, the processing of each block may be executed by one piece of hardware or by multiple pieces of hardware. Furthermore, the processing of each block may be realized by one program or by multiple programs.

The processing units in the flowcharts shown in Figures 27 to 30 are divided according to the main processing content in order to make the processing of the sound processing device 100 easier to understand, and the present invention is not limited by the manner in which the processing units are divided or the names of the processing units. The processing of the sound processing device 100 may be divided into more processing units according to the processing content. Furthermore, the processing of the sound processing device 100 may be divided so that one processing unit includes even more processes.

When the program of the present invention is realized using a computer, the program executed by the computer can be configured in the form of a recording medium or a transmission medium for transmitting the program. The recording medium can be a magnetic or optical recording medium or a semiconductor memory device. Specific examples include portable or fixed recording media such as flexible disks, HDDs, CD-ROMs (Compact Disk Read Only Memory), DVDs, Blu-ray Discs, magneto-optical disks, flash memories, and card-type recording media. The recording medium can also be a non-volatile storage device such as a ROM or HDD provided in the sound processing device 100.

11 Sound source device 12 D/A converter 13 Amplifier 14 Speaker 100, 200 Sound processing device 110 LPF
120 First fluctuation detection section 121 First maximum value detection section 122 First maximum value hold section 123 First smoothing processing section 124 First decibel conversion section 125 First differentiation processing section 126 First inversion processing section 127 First level limiting section 128 First linear conversion section 129 Second level limiting section 130 Second linear conversion section 140 First weighting section 141 First repeat processing section 142 Third smoothing processing section 143 First multiplier 144 First subtractor 145 First gain adjustment section 150 Second weighting section 151 Second repeat processing section 152 Fourth smoothing processing section 153 Second multiplier 154 Fourth level limiting section 155 Third multiplier 156 Second subtractor 157 Second gain adjustment section 160 HPF
170 Second fluctuation detection section 171 Second maximum value detection section 172 Second maximum value hold section 173 Second smoothing processing section 174 Second decibel conversion section 175 Second differentiation processing section 176 Second inversion processing section 177 Third level limiting section 178 Third linear conversion section 180 Third weighting section 181 Third repeat processing section 182 Fifth smoothing processing section 183 Fourth multiplier 184 Third subtractor 185 Third gain adjustment section 190 Synthesis section 210 Memory 211 Control program 220 Processor 230 Control section

Claims (5)

a detector for detecting an envelope signal of an input audio signal;
A conversion unit that converts the detected envelope signal into decibels;
a differential processing unit that performs differential processing on the converted envelope signal to generate a change amount signal that indicates a change amount of the envelope signal;
an inversion unit that inverts a signal level of the change amount signal;
a first fluctuation detection signal generating unit that generates a first limiting signal by limiting the gain of the inverted change amount signal to a signal equal to or less than 0 dB, and generates a first fluctuation detection signal in which the gain of the generated first limiting signal is indicated by an amplitude;
a second fluctuation detection signal generating unit that generates a second limiting signal by limiting the gain of the inverted change amount signal to 0 dB or more, and generates a second fluctuation detection signal in which the gain of the generated second limiting signal is indicated by an amplitude;
a limiting signal generating unit that generates a multiplied signal by multiplying the first fluctuation detection signal by the second fluctuation detection signal, and generates a third limiting signal by limiting the amplitude of the generated multiplied signal to 1 or more;
a first complementary signal generating unit that generates a first complementary signal by subtracting a signal obtained by multiplying the audio signal by the third limited signal from the audio signal;
an adder for adding the first complementary signal to the audio signal;
An audio processing device comprising: a second complementary signal generating unit that generates a second complementary signal by subtracting a signal obtained by multiplying the audio signal by the first fluctuation detection signal from the audio signal,
The sound processing device according to claim 1 , wherein the adder adds the first complementary signal and the second complementary signal to the audio signal. a gain adjustment unit that adjusts gains of the first complementary signal and the second complementary signal generated by the first complementary signal generation unit and the second complementary signal generation unit,
3. The sound processing device according to claim 2, wherein the adder adds the first complementary signal and the second complementary signal, the gains of which have been adjusted by the gain adjuster, to the audio signal. the differential processing unit performs filtering processing on the audio signal using a high-pass filter as the differential processing;
4. The sound processing device according to claim 1, wherein the change time of the change amount of the envelope signal obtained by the filtering process is adjusted by changing a setting value of a normalized cutoff frequency in the high-pass filter. On the computer,
detecting an envelope signal of an input audio signal;
converting the detected envelope signal into decibels;
a step of differentiating the converted envelope signal to generate a change amount signal indicating a change amount of the envelope signal;
inverting a signal level of the generated change amount signal;
generating a first limiting signal by limiting the gain of the inverted change amount signal to a signal equal to or less than 0 dB;
generating a first fluctuation detection signal in which the gain of the generated first limiting signal is indicated by an amplitude;
generating a second limiting signal by limiting the gain of the inverted change amount signal to 0 dB or more;
generating a second fluctuation detection signal in which the gain of the generated second limiting signal is indicated by an amplitude;
multiplying the first fluctuation detection signal by the second fluctuation detection signal to generate a multiplied signal;
A step of generating a third limited signal by limiting the amplitude of the generated multiplication signal to 1 or more;
multiplying the audio signal by the third limited signal and subtracting the resultant signal from the audio signal to generate a first complementary signal;
adding said first complementary signal to said audio signal;
A program characterized by:

Images