JPH0562754B2 - - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION [Field of Industrial Application] The present invention relates to the discrimination of audio signals, and particularly to a discrimination device capable of discriminating between audio signals corresponding to music and audio signals corresponding to conversation.

[Prior Art] Distinguishing between a signal corresponding to music and a signal corresponding to conversation such as news in an audio device can be used for automatic adjustment of sound quality for music and conversation, automatic tuning of music programs on radio, etc. , which is very useful.

Generally, when an audio signal is converted into a DC voltage according to its signal level, it is known that the DC voltage of the audio signal corresponding to music is higher than that of the audio signal corresponding to conversation [1981] Published December 20th, "Electroacoustic Engineering" (Corona Publishing), see pages 28-29].

Based on such literature, for example, a discriminating device as shown in FIG. 5 can be considered. In the figure, 2 is a voltage averaging circuit that equalizes the audio signal from signal source 1 and converts it into a DC voltage, 3 is a reference voltage source, and the reference voltage value set for this reference voltage source 3 is generally This is a threshold value that distinguishes between a DC voltage value obtained by leveling an audio signal corresponding to ordinary music and a DC voltage value obtained by leveling an audio signal corresponding to ordinary conversation. 4
is a comparison circuit that compares the voltage value from the voltage averaging circuit 2 and the reference voltage value, and outputs a discrimination signal when the voltage value from the voltage averaging circuit 2 becomes equal to or higher than the reference voltage from the reference voltage source 3. It is.

Therefore, the signal discriminating device assumes that the audio signal from the signal source 1 corresponds to music when the discriminating signal is output from the comparator circuit 4, and that the audio signal from the signal source 1 corresponds to music when the discriminating signal is not output from the comparator circuit 4. The audio signals from source 1 are respectively distinguished as corresponding to speech.

[Problems to be Solved by the Invention] However, in the signal discrimination described in the above-mentioned documents, audio signals corresponding to music and audio signals corresponding to conversations cannot be uniformly distinguished based on their signal levels. Because they are separated, accurate discrimination is difficult. For example, in the case of quiet classical music, the audio signal level may be lower than that of a conversation, making it impossible to distinguish between them.

[Means for Solving the Problems] The present invention has been made in view of the above problems, and an object of the present invention is to provide an audio signal discrimination device that can distinguish audio signals corresponding to music and conversation as accurately as possible. To achieve this objective, the present invention distinguishes audio signals corresponding to music and conversation by the presence or absence of periodicity of their signal waveforms, and the presence or absence of so-called rhythm in music and conversation. Focusing on this, we have developed a low-pass filter 11 that passes components below a predetermined frequency of a signal input from an audio signal source containing a mixture of music and conversation, and outputs a trigger signal in synchronization with the signal that has passed through this low-pass filter 11. and determining means 13, 14, 15, 16 for measuring the period between the trigger signals outputted from the signal generating means 12 and determining the periodicity between the measured consecutive trigger signals. The device is configured to determine whether an input signal corresponds to music or conversation.

In this case, the discrimination means includes measuring means 13, 14, 15 for measuring the period between the trigger signals outputted from the signal generating means 12, and based on the period measured by the measuring means 13, 14, 15. After that, the reference period is compared with the period newly measured by the measuring means 13, 14, and 15, and it is determined whether the signal corresponds to voice or conversation. Arithmetic circuit 1 that determines
It is desirable that the configuration is made up of 6.

Further, the measuring means includes a pulse generating circuit 13 that outputs a pulse signal of a predetermined period, and a count value is cleared every time a trigger signal from the signal generating means 12 is input.
a counter 14 for counting pulse signals from 3;
a sample hold circuit 15 for storing and holding the count value in the counter 14 in synchronization with the trigger signal from the signal generating means 12;
The sample and hold circuit 15 compares each count value sampled within a predetermined time to detect the minimum value, and when each count value becomes approximately an integer multiple of the minimum value, periodicity is detected. It is preferable to use a circuit that outputs a determination signal indicating that there is a presence.

[Embodiments of the Invention] Hereinafter, embodiments of the present invention will be described based on the drawings.

FIG. 1 is a block diagram showing an example of an audio signal discrimination device according to the present invention. In the same figure,
11 is a low-pass filter that passes a frequency signal below a predetermined frequency among the audio signal components from the signal source 1; 12 is a differentiating circuit as a signal generating means for differentiating the signal from the low-pass filter 11; The output stage of is clamped, for example by a diode, so that there is no response to negative changes in the input signal. 13 is a pulse generation circuit that outputs a pulse signal of a predetermined period; 14 is a counter that clears the current count value and counts the pulse signal from the pulse generation circuit 13 every time a trigger signal from the differentiating circuit 12 is input; 15 is a sample hold circuit that stores and holds the count value in the counter 14 in synchronization with the trigger signal from the differentiating circuit 12. 16 is an arithmetic circuit, and this arithmetic circuit 16 compares each count value sampled by the sample hold circuit 15 within a predetermined time, detects the minimum value, and detects the minimum value of each count value. A discrimination signal is output when the value is approximately an integer multiple of the minimum value. The above pulse generation circuit 13, counter 14, sample hold circuit 15, and arithmetic circuit 16 constitute a discriminating means.

Next, the operation will be explained according to the timing chart shown in FIG.

When listening to a music broadcast program from a signal source 1, for example, a radio, when an audio signal corresponding to the music is input to the low-pass filter 11, a signal of a predetermined low frequency component of the audio signal is output from the low-pass filter 11. Then, the output signal is subjected to differentiation processing by a differentiation circuit 12.
At this time, since the rhythm of music is generally carved by a bass instrument, the signal passed through the low-pass filter 11 has a waveform that is synchronized with the rhythm of the music, and the output signal from the differentiator circuit 12 is also synchronized with the rhythm. Become something. Here, for example, as shown in FIG. 2, times t1, t2, t3, t4, . . .
When the trigger signal is output from the differentiating circuit 12,
The counter 14 clears the count value up to now every time the trigger signal is input, and the pulse generation circuit 13
Count the pulse signals from C1, C
2, C3, . . . are sampled by the sample-hold circuit 15, and the arithmetic circuit 16 detects the minimum value from the counted values C1, C2, C3, . is approximately an integer multiple of the minimum value (determination of periodicity). At this time, since the output of the trigger signal from the differentiating circuit 12 is synchronized with the rhythm of the music, the arithmetic circuit 16 outputs a discrimination signal.

On the other hand, when the audio signal from the signal source 1 corresponds to conversation such as news, the signal of the predetermined low frequency component of the audio signal that has passed through the low-pass filter 11 is transmitted to the differentiating circuit 12 in the same way as above.
The signal is differentiated and output as a trigger signal. Then, in synchronization with this trigger signal, the count value of the pulse signal from the pulse generation circuit 13 by the counter 14 is sequentially sampled by the sample hold circuit 15. At this time, since various conversations do not have a fixed rhythm, the audio signals corresponding to these conversations have no periodicity. Therefore, as mentioned above, each count value sampled by the sample-and-hold circuit 15 has no correlation as long as the count value is approximately an integer multiple of the minimum value, and the arithmetic circuit 1
No discrimination signal is output from 6 onwards.

That is, in the embodiment described above, the audio signals are distinguished depending on the presence or absence of the discrimination signal from the discrimination circuit 16.

In addition, when the signal discrimination device according to the above embodiment is used in an automatic sound quality adjustment device, the sound quality is adjusted in advance so that the frequency characteristics are suitable (easy to hear) for conversation, for example, the frequency characteristics are attenuated in the low and treble ranges. An adjustment circuit and a sound quality adjustment circuit that adjusts the frequency characteristics to match the user's taste in music, for example, frequency characteristics that emphasize the low and treble ranges, are provided. Therefore, each of the above-mentioned sound quality adjustment circuits to which the audio signal from the signal source 1 is input may be switched.

FIG. 3 is a block diagram showing another example of the audio signal discrimination device according to the present invention. This is configured such that a pulse count is performed every other trigger signal obtained by differentiating an audio signal passed through a low-pass filter 11 in a differentiating circuit 12. That is, a changeover switch 17 is provided that switches contacts 1 and 2 in synchronization with the trigger signal from the differentiating circuit 12, and a pulse signal from the pulse generation circuit 13 inputted via contact 1 of the changeover switch 17 is countered. 14 are configured to count. Then, the sample and hold circuit 15 sequentially samples the counted value of the counter 14, and the arithmetic circuit 16 detects the minimum value of each counted value sampled within a predetermined time, and each counted value is approximately an integer multiple of the minimum value. I am trying to determine whether it is familiar or not.

In the above-mentioned signal discrimination device, for example, as shown in FIG. 4, times t1, t2, t3, t4, . . .
When a trigger signal is output from the differentiating circuit 12, the changeover switch 17 switches the contacts 1 and 2 in synchronization with the trigger signal at each time. As a result, the counter 14 counts the pulse signals from the pulse generation circuit 13 between times t1 and t2 when the changeover switch 17 is on the contact 1 side, and between times t3 and t4, and the counted value C1 ,C
2, . . . are sampled by the sample hold circuit 15. Then, the arithmetic circuit 16 performs the discrimination operation based on the sampled value.

[Effects of the Invention] As described above, according to the present invention, the periodicity of a predetermined low frequency component in particular of an input audio signal mixed with music and conversation is determined, and the periodicity of a predetermined low frequency component is determined based on the determination result. Since the audio signals are differentiated, it becomes possible to reliably distinguish between the audio signals corresponding to music whose rhythm is generally carved by a low-pitched instrument and the audio signals corresponding to conversations without rhythm. As a result, in the audio device, it becomes possible to accurately adjust the sound quality of music and conversation, or automatically tune a music program on the radio, etc.

[Brief explanation of the drawing]

FIG. 1 is a block diagram showing an example of an audio signal discriminating device according to the present invention, FIG. 2 is a timing chart showing the operation of the discriminating device shown in FIG. 1, and FIG.
The figure is a block diagram showing another example of the audio signal discrimination device according to the present invention, FIG. 4 is a timing chart showing the operation of the discrimination device shown in FIG. 3, and FIG. 5 is an example of the conventional audio signal discrimination device. FIG. 1...Signal source, 11...Low pass filter, 1
2...Differential circuit, 13...Pulse generation circuit, 14
... Counter, 15 ... Sample hold circuit,
16... Arithmetic circuit, 17... Changeover switch.

Claims (1)

[Claims] 1. A low-pass filter 11 that passes components below a predetermined frequency of a signal input from an audio signal source containing a mixture of music and conversation, and a trigger signal that is synchronized with the signal that has passed through the low-pass filter 11. signal generating means 12 that outputs
and determining means 13, 14, 1 for measuring the period between the trigger signals output from the signal generating means 12 and determining the periodicity between the measured consecutive trigger signals.
5 and 16, and is characterized in that it determines whether an input signal is a signal corresponding to music or a signal corresponding to conversation. 2. The determining means includes measuring means 13, 14, 15 for measuring the period between the trigger signals outputted from the signal generating means 12, and a standard based on the period measured by the measuring means 13, 14, 15. Set the cycle, and then
The reference period and the measuring means 13, 14, 1
The audio signal according to claim 1, further comprising an arithmetic circuit 16 for comparing the newly measured period in step 5 and determining whether the signal corresponds to voice or conversation. Discrimination device. 3. The measurement means includes a pulse generation circuit 13 that outputs a pulse signal of a predetermined period, and a count value is cleared every time a trigger signal from the signal generation means 12 is input, and the pulse generation circuit 13 outputs a pulse signal from the pulse generation circuit 13. The arithmetic circuit 16 includes a counter 14 that counts signals, and a sample hold circuit 15 that stores and holds the count value in the counter 14 in synchronization with a trigger signal from the signal generating means 12. The minimum value is detected by comparing each count value sampled within a predetermined time, and when each count value is approximately an integer multiple of the minimum value, a determination signal is output indicating that there is periodicity. The audio signal discrimination device according to claim 2, characterized in that it is a circuit.