JP3596580B2 - Audio signal processing circuit - Google Patents

Description

[0001]
TECHNICAL FIELD OF THE INVENTION
The present invention relates to an audio signal processing device used in fields such as hearing aids, telephones, loudspeakers, and voice communications.
[0002]
[Prior art]
When transmitting or reproducing sound, if there is much reverberation or echo in the transmission system or reproduction system, the intelligibility of the sound is reduced. Therefore, in such a case, processing such as lowering the utterance speed, finely decomposing a continuously uttered word sound, and reproducing it at a later time is performed.
[0003]
When it is difficult to hear a high frequency sound such as a consonant, the high frequency may be emphasized by a frequency equalizer process. Further, a process of applying a weighting function in consideration of so-called successive masking (a phenomenon in which a consonant is masked by a vowel when the consonant follows a vowel having a large energy) has been attempted.
[0004]
Further, the above processing may be performed for a hearing-impaired person or an elderly person.
[0005]
[Problems to be solved by the invention]
However, as described above, if the utterance speed is reduced or a speech sound that is continuously uttered is decomposed, the following problems occur.
[0006]
1. A time lag occurs between the original sound and the original sound. Therefore, it cannot be used for conversations. In addition, even when listening to a broadcast or the like, the time required to finish listening is long.
2. Since the rate of change of the voice component is also an important clue for the perceptual judgment of the speech sound, if the speech speed is reduced, this clue may change and be perceived as another speech sound.
3. When a speech sound is decomposed and played back slowly, information as a unit of the speech sound or information of a transient change portion may be lost, and the intelligibility may deteriorate.
4. A sound whose high frequency is always amplified by the frequency equalizer processing may be uncomfortable due to an imbalance in timbre or may be difficult to hear.
5. In the process of applying the weighting function in consideration of the successive masking, at least the time length of the weighting function is delayed, and the immediacy is lost. As a result, a time lag may occur between the movement of the mouth and the processed sound, which may adversely affect the clarity. Also, when there is acoustic feedback from the earphone to the microphone, a time delay causes a phenomenon such as reverberation.
[0007]
In order to address the above-described problems, the present applicant has disclosed in Japanese Patent Application No. 9-99144 a low-frequency component as a method of compensating for the masking of a low-frequency component of a sound into a high-frequency component. A "sound signal processing circuit" has been proposed in which the amplification factor in the high frequency band is increased only in the time zone in which.
[0008]
However, in this "audio signal processing circuit", since the amplification factor in the high frequency band is increased, a margin is required for the gain and the maximum output of the amplifier, and a margin for the howling margin is required to prevent howling. It is not suitable for severe hearing loss.
[0009]
The present invention seeks to eliminate the above problems.
[0010]
[Means for Solving the Problems]
Therefore, in the present invention,
A level reduction circuit that reduces the level of a low-frequency component of the input audio signal,
A first detection circuit for detecting a start point of a high-frequency component of the audio signal;
A second detection circuit that detects an end point of a high-frequency component of the audio signal;
A level control circuit that controls a level reduction degree of the level reduction circuit according to detection outputs of the first detection circuit and the second detection circuit;
When the first detection circuit detects the start point of the high frequency component, the level control circuit supplies a control signal to the low frequency level reduction circuit so as to increase the degree of level reduction,
An audio signal processing circuit configured to supply a control signal to the low frequency level reduction circuit so as to return the level reduction degree when the second detection circuit detects the end point of the high frequency component; Is what you do.
Therefore, the level of the low-frequency component is reduced by an amount corresponding to the masking of the high-frequency component by the low-frequency component and the successive masking, and the clarity of the processed voice is improved.
[0011]
BEST MODE FOR CARRYING OUT THE INVENTION
By the way, the speech of a normal conversation is composed of a combination of unvoiced sound and voiced sound, that is, a combination of a low frequency component and a high frequency component. In addition, sounds that exist in general living environments are often combinations of low-frequency components and high-frequency components.
[0012]
In the sense of hearing, it is known that a low-frequency component masks a high-frequency component, and this masking also works when a voice is perceived. In a normal hearing person, the masking disturbance is small and the voice can be perceived correctly even with the masking. It has become one of.
[0013]
In addition, even for a normal hearing person, if there is noise having a large low-frequency component, the hearing of words may be reduced. Further, the so-called successive masking also reduces the clarity of words.
[0014]
Therefore, the present invention is intended to suppress a decrease in clarity due to such masking or successive masking.
[0015]
Therefore, in one embodiment of the present invention, a period of several milliseconds to several tens of milliseconds from the start point of the high frequency component of the audio is set as the start period, and ten and several milliseconds from the end point of the high frequency component of the audio. When a period of up to several tens of milliseconds is set as the end period, during the processing period from the start period to the end period, the level of the low frequency component of the audio signal is reduced.
[0016]
FIG. 1 shows a configuration of an embodiment of the present invention. An unprocessed audio signal S11 is supplied to two band-pass filters 12, 13 in common through an input terminal 11, and these band-pass filters 12, 13 are provided. Output signals S12 and S13 are supplied to an addition circuit 16 through fixed gain and variable gain amplifiers 14 and 15.
[0017]
In this case, the first formant component of the voice is treated as the low-frequency component of the voice, and the second formant component and the consonant component are treated as the high-frequency component of the voice. Therefore, the pass band of the band-pass filter 12 is set to a band of the second formant component and the consonant component of the voice, for example, 1.2 kHz to 8 kHz, and the pass band of the band-pass filter 13 is set to the pitch component of the voice and the first band. The band of the formant component, for example, 200 Hz to 1.2 kHz.
[0018]
Therefore, when the gain of the fixed gain amplifier 14 is equal to the reference gain of the variable gain amplifier 15, the high frequency signal component S 12 and the low frequency signal component S 13 included in the input audio signal S 11 are output from the adder circuit 16. An audio signal S16 having the same ratio is obtained. Then, the signal S16 is taken out to the output terminal 17.
[0019]
Further, the audio signal S11 from the input terminal 11 is sequentially supplied to the bandpass filter 21 and the level calculation circuit 22 for preprocessing. In this case, the band-pass filter 21 makes it easy to detect the start point and the end point of the high-frequency signal component, and reduces the influence of noise on the second formant component and the consonant component of the input audio signal S11. Are extracted as a signal S21. Therefore, the pass band of the band pass filter 21 is set, for example, equal to the pass band of the band pass filter 12.
[0020]
The level calculation circuit 22 rectifies the signal S21 in two-waves and extracts the low-frequency component (for example, a component of 60 Hz or less) of the signal S21, thereby obtaining the level of the high-frequency signal component S12 output from the band-pass filter 12. Is formed.
[0021]
Then, the calculation signal S22 of the level calculation circuit 22 is supplied to the detection circuit 23, and the start point and the end point of the high frequency signal component S12 are detected. The detection signal S23 is supplied to the level control circuit 24 to perform level control. A signal S24 is formed, and the signal S24 is supplied to the variable gain amplifier 15 as a control signal for the gain G15.
[0022]
In this case, the detection of the start point and the end point of the high-frequency signal component S12 and the magnitude of the gain G15 of the variable gain amplifier 15 have, for example, a relationship as shown in FIG. That is, when the level of the high-frequency signal component S12 indicated by the calculation signal S22 is smaller than the threshold value for the start point determination, the gain G15 of the amplifier 15 is set to the reference value equal to the gain of the amplifier 14, but at time t1. When the level of the high-frequency signal component S12 becomes larger than the threshold value for determining the start point, the gain G15 gradually decreases to a predetermined value in a period Ts from the time point t1 to a time point t2 several milliseconds to several tens of milliseconds later. Is done.
[0023]
Further, when the level of the high-frequency signal component S12 indicated by the calculation signal S22 is larger than the threshold value for the end point determination, the gain G15 is kept small, but at the time t3, the level of the high-frequency signal component S12 becomes When the gain G15 becomes smaller than the threshold value for the end point determination, the gain G15 is gradually increased to the reference value in a period Te from a time t3 to a time t4 several tens of milliseconds to 200 milliseconds later.
[0024]
According to such a configuration, when the high-frequency signal component S12 is included in the audio signal S11 before processing, the level control signal S24 controls the amplifier 15 during the period t1 to t4 from the start point to the end point. Since the gain G15 is reduced, the level of the low-frequency signal component S13 passing through the amplifier 15 is reduced.
[0025]
Therefore, during the period from the start point to the end point of the high-frequency signal component S12, the level of the high-frequency signal component S12 of the audio signal S16 output to the terminal 17 becomes relatively high, so that the reproduced sound of the signal S16 is masked. Even if it occurs, the high-frequency component of the reproduced sound becomes relatively large corresponding to the masking, so that the listening of words is improved. In the case of successive masking, the degree of masking is large near the start point of the high frequency component, but as shown in FIG. 2, the level of the low frequency signal component is close to the start point of the high frequency signal component S12. Since the level of the high-frequency signal component S12 is relatively increased relatively quickly, the clarity can be effectively increased even for successive masking.
[0026]
FIG. 3 shows an embodiment of a method in which the detection circuit 23 and the level control circuit 24 form the level control signal S24 from the calculation signal S22. That is, in this case, all of the circuits shown in FIG. 1 are digitized and configured by, for example, a DSP. The audio signal S11 is a digital audio signal obtained by A / D-converting the original analog audio signal before processing.
[0027]
Then, in the detection circuit 23 and the level control circuit 24, the level control routine 100 of FIG. 3 is executed for each sample of the digital audio signal S11, and the gain G15 of the variable gain amplifier 15 is changed as shown in FIG. Controlled. In the routine 100 and the following description, the meaning of each variable is as follows.
[0028]
e (i): the level indicated by the i-th sample of the audio signal S11.
threshold1: threshold value for determining the end point of the high frequency component.
When the signal S11 becomes smaller than this value, it is determined that the high-frequency component ends.
threshold2: threshold value for determining the start point of the high frequency component.
When the signal S11 becomes larger than this value, it is determined to be the start point of the high frequency component.
threshold1 ≦ threshold2 is set.
w: weight coefficient for controlling the gain G15. 0 ≦ w ≦ 1
When w = 0, G15 = reference gain w = 1, G15 = minimum gain d1: step width when decreasing coefficient w 1.
d2: Step width when increasing the coefficient w 2.
[0029]
That is, in the level control routine 100, first, at step 101, it is determined whether or not the signal level e (i) of the i-th sample is smaller than a threshold value threshold2 for starting point determination. The process proceeds from step 101 to step 102.
[0030]
Then, in this step 102, it is determined whether or not the signal level e (i) of the i-th sample is smaller than a threshold value threshold1 for determining the end point. When it is smaller, the process proceeds from step 102 to step 103, In this step 103, the coefficient w 1 is reduced by the step width d1, and the routine 100 ends. Therefore, as shown in FIG. 2, when the end point of the high frequency component is detected, thereafter, the gain G15 gradually increases, and the level of the low frequency component gradually increases.
[0031]
If the signal level e (i) of the i-th sample is equal to or higher than the threshold value threshold1 for determining the end point in step 102, the process ends the routine 100 from step 102. Therefore, as shown in FIG. 2, during a period until the end point of the high frequency component is detected (a period in which the gain G15 is small), the gain G15 is maintained and the level of the low frequency component is maintained low.
[0032]
Further, in step 101, when the signal level e (i) of the i-th sample is equal to or higher than the threshold value threshold2 for starting point determination, the process proceeds from step 101 to step 104, where the coefficient w 2 The width is increased by d2, and the routine 100 ends. Therefore, as shown in FIG. 2, when the start point of the high frequency component is detected, thereafter, the gain G15 gradually decreases, and the level of the low frequency component gradually decreases.
[0033]
Thus, according to the level control routine 100, the level of the high frequency component is relatively increased by controlling the gain G15 of the low frequency component amplifier 15 in the decreasing direction. The auditory decay of the high-frequency component can be compensated for, and the clarity of voice such as a consonant part of a conversation can be improved.
[0034]
4A and 4B show the results of observation of the audio waveform. FIG. 4A shows the waveform of the input audio signal S11 that has not been processed by the routine 100, and FIG. 4B shows the waveform of the output audio signal S16 that has been processed by the routine 100. This is an example of observation. The utterance content at this time is "Please write on the first line".
[0035]
During the period from the start point to the end point of the high-frequency component (arrows A, B, C, D, and E), the level of the low-frequency component is low and the level of the high-frequency component is relatively high. are doing. In particular, in the portions of arrows A, B and C, the syllables are separated and emphasized, and it sounds as if they were pronounced clearly.
[0036]
Therefore, according to the above-described processing circuit, the following effects can be obtained when transmitting or reproducing sound in a system such as reverberation or echo, or when a hearing-impaired person or elderly person hears sound.
1. Among the uttered voices, the level of the low-frequency component is reduced so that the masking to the high-frequency component is reduced, and consonants are audibly emphasized, and the syllables are emphasized, so that the voice is clear, Clarity can be improved.
2. Since it is not necessary to increase the high-frequency component in order to reduce the masking, there is no need to have a margin in the gain, the maximum output, or the howling margin of the high-frequency component amplifier. Therefore, the amplifier for the high frequency component can always be operated in the maximum gain state.
3. Only when the masking is occurring, the high-frequency component is audibly enhanced, so that there is no discomfort as if the balance of the timbre was broken as in the case where the high-frequency component is always emphasized.
4. Since the processing can be performed immediately in principle, there is no time difference between the movement of the mouth of the speaker and the processed sound. Also, even if there is acoustic feedback from the earphone to the microphone, it does not sound like reverberation, so it is easy to hear.
5. The rate of change of the speech component, which is important for the perceptual judgment of the speech sound, the information as a unit of the speech sound, and the information of the transient change portion are not lost.
[0037]
6. The routine 100 has a small number of processing steps, and therefore can sufficiently cope with a DSP whose processing is somewhat slow.
[0038]
The level control routine 200 shown in FIG.
threshold = threshold1 = threshold2
This is a case where the level control routine 100 is simplified. That is, in the routine 200,
threshold: threshold value for determining the start point and the end point of the high frequency component.
If the signal S11 is smaller than this value, it is determined to be the end point, and if it is larger than this value, it is determined to be the start point.
The rest is the same as the routine 100.
[0039]
Then, in step 201, the signal level e (i) of the i-th sample is compared with the threshold value threshold, and if the level e (i) is smaller than the threshold value threshold, in step 202, the coefficient w 2 Otherwise, in step 203, the coefficient w 2 is increased by the step width d2.
[0040]
Therefore, according to the level control processing routine 200, the processing is further simplified, and the load on the DSP is further reduced.
[0041]
In the above description, the variable gain amplifier 15 is used to reduce the level of the low frequency component of the input audio signal S11. However, instead of this, a fixed gain amplifier and a variable attenuator circuit (so-called electronic volume) are used. It can also be a combination. In the above description, two band-pass filters 12 and 21 having the same characteristics are used for the so-called main line and the pre-processing, but the band-pass filter 21 for the pre-processing is omitted and the band-pass filter 12 is used. By distributing the high frequency component S12 to the level calculation circuit 22, a simpler configuration can be achieved.
[0042]
Further, by supplying the calculation signal S22 of the level calculation circuit 22 to the level control circuit 24, the level reduction degree of the low frequency component in the variable gain amplifier 15 is adjusted according to the level of the high frequency component S12 of the input audio signal S11. It can also be controlled.
[0043]
【The invention's effect】
ADVANTAGE OF THE INVENTION According to this invention, a sound is clear and intelligibility can be improved. Further, the requirement for the amplifier for the high frequency component can be reduced. Furthermore, there is no unpleasant sensation as when always emphasizing the high frequency range of sound. Further, there is no time difference between the movement of the mouth of the speaker and the processed sound.
[0044]
Furthermore, even if there is acoustic feedback from the earphone to the microphone, the sound does not look like reverberation and is easy to hear. In addition, effective information for perceptual judgment of speech sounds and the like is not lost. In addition, the burden of digital processing can be reduced.
[Brief description of the drawings]
FIG. 1 is a system diagram illustrating one embodiment of the present invention.
FIG. 2 is a diagram for explaining the present invention.
FIG. 3 is a flowchart illustrating one embodiment of the present invention.
FIG. 4 is a waveform chart for explaining the present invention.
FIG. 5 is a flowchart illustrating one embodiment of the present invention.
[Explanation of symbols]
11 = input terminal, 12 = bandpass filter, 13 = bandpass filter, 14 = fixed gain amplifier, 15 = variable gain amplifier, 16 = addition circuit, 17 = output terminal, 21 = bandpass filter, 22 = level calculation circuit , 23 = detection circuit, 24 = level control circuit, 100 and 200 = level control routine

Claims (5)

A band division circuit that divides the input audio signal into a low-frequency component and a high-frequency component,
A level variable circuit that varies the level of the low-frequency component divided by the band dividing circuit;
A first detection circuit for detecting a start point of the high frequency component extracted by the band division circuit;
A second detection circuit that detects an end point of the high-frequency component extracted by the band division circuit;
A level control circuit that controls the level variable circuit according to the detection outputs of the first detection circuit and the second detection circuit;
When the first detection circuit detects the start point of the high frequency component, it supplies a control signal from the level control circuit so as to reduce the level to the level variable circuit,
When the second detection circuit detects an end point of the high frequency component, a control signal is supplied from the level control circuit to the level variable circuit so as to cancel the level reduction. Signal processing circuit. The audio signal processing circuit according to claim 1,
The first detection circuit is a sound signal processing circuit configured to detect, when the signal level of the high frequency component exceeds a predetermined threshold value, to output the detection signal of the start point. The audio signal processing circuit according to claim 1,
The second detection circuit is a sound signal processing circuit configured to detect, when the signal level of the high frequency component falls below a predetermined threshold value, to output the end point detection signal. The audio signal processing circuit according to claim 1,
A sound in which a control signal is supplied from the level control circuit to the level variable circuit so as to increase a level according to an elapsed time from a point in time when the first detection circuit detects the start point of the high frequency component. Signal processing circuit. The audio signal processing circuit according to claim 1,
A sound in which a control signal is supplied from the level control circuit to the variable level circuit so as to reduce a level according to an elapsed time from a point in time when the second detection circuit detects the end point of the high frequency component. Signal processing circuit.

Images