JPH03201757A - Processing method for voice guidance - Google Patents

Abstract

PURPOSE:To prevent occurrence of malfunction by attenuating at least part of a frequency band of a pushbutton signal from a voice guidance. CONSTITUTION:A voice guidance 11 of a processed original sound is digitized and inputted to a Fourier transformation section 12. An output applying Fourier transformation to the original voice guidance, that is, a power spectrum 13 passes through a comb-line filter 14 cutting off only a frequency used for a PB tone. A signal processed by the comb-line filter 14, that is, a power spectrum 15 passing through the comb-line filter 14 is subjected to inverse Fourier transformation by an inverse Fourier transformation section 16 and the signal is restored to a time region from a frequency region. Thus, the processed voice guidance 17 is a voice signal not including the frequency used for the PB tone. Then occurrence of malfunction is prevented.

Description

DETAILED DESCRIPTION OF THE INVENTION "Field of Industrial Application" The present invention relates to a processing method for creating voice guidance used for network service guidance, etc.

"Conventional technology" In order to use network services such as message dialing that are currently in use, users usually dial the telephone number of the center providing the service, and then call the service provider (center). The user listens to voice guidance and presses a push button (hereinafter abbreviated as PB) to receive the desired service. For example, when using a message dial, the sequence is shown in Figure 2.

In such services, as users become accustomed to the service, they begin to press push buttons without listening to the voice guidance all the way through. In order to cope with this, it is necessary to operate a push button signal detector (hereinafter abbreviated as PBR) on the center side before the reproduction of the audio guidance ends.

However, current telephone networks always have echo back, so if the PBR is operated before the end of voice guidance playback, this voice guidance will be input to the PBH by echo back. This situation is shown in FIG. 3. The voice has a wide frequency band, including the frequency used for the PB double signal. Therefore, during reproduction of audio guidance, the PB signal frequency PBR may malfunction and be detected as a PB signal.

In order to avoid this, a conventional method has been to provide PBH with high accuracy. In other words, by incorporating a circuit that improves the distinguishability from voice, some voice signals can be
This is a method in which B is not detected.

However, since such PBR is complicated, it tends to be expensive. In addition, there is a possibility that the opposite malfunction may occur, that is, the PB multiplied signal may not be detected because it is regarded as an audio signal. Moreover, a large number of conventional inexpensive PBRs are already connected to the telephone network. It is practically impossible to replace these.

SUMMARY OF THE INVENTION An object of the present invention is to provide a processing method for creating audio guidance that does not cause a malfunction in detecting a PB multiplied signal from an audio signal even when input to a conventional PBR.

"Means for Solving the Problem" The present invention provides a method for reducing echo back on a telephone network by attenuating at least part of the frequency used by the FB from the frequency band of voice guidance.
The most important feature is to prevent the PBR from operating even if voice guidance is input to the BR.

"Example 1" FIG. 1 shows an embodiment of the invention.

First, the original audio guidance 11 to be processed is digitized and input to the Fourier transform section 12 . The Fourier transform unit usually performs fast Fourier transform (FFT), and the input original voice guidance is Fourier transformed and converted from the time domain to the frequency domain. The output obtained by Fourier transforming this original voice guidance, that is, the power spectrum 13, is passed through a comb filter 14 that cuts only the frequencies used by the PB. PB double signal One frequency from low frequency 697.770.852.941Hz and high frequency 1209.1336.1447.
1633) 4 by combining with one frequency of 1z
Specify X 4 =16 types. And each frequency band is 1.5%.

A filter that cuts this band is used as the comb filter 14. However, although the frequency band of PB is ±1.5z, the center frequency of actual PBR is shifted because it includes analog system errors. Therefore, it is necessary to increase the bandwidth to be cut (attenuated) to some extent. ± to the center frequency
3. When cutting OX, the frequency to be attenuated by the comb filter 14 is 676 to 718.747 to 793.82.
6-877, 913-969.1172-1245, 1
296-1376, 1433-1521.1584-1
It becomes 682Hz. The power spectrum 15 processed by the comb filter 14, that is, passed through the comb filter 14, is subjected to an inverse Fourier transform in an inverse Fourier transform unit 16 and returned from the frequency domain to the time domain.

The audio guidance 17 processed through the above operations becomes audio that does not include the frequency used by the PB. Therefore, even if this audio is input to the PBR, the PBR will not detect the PB multiplied signal.

``Example 2'' The voice guidance processed in Example 1 suffers from deterioration in voice quality. An improved method to reduce this will be shown.

PB double signal One of four high frequency frequencies and four low frequency frequencies
It consists of a combination of one of the following: Therefore, if processing is performed to attenuate the four high frequencies or the four low frequencies in the original audio guidance, the PBR will not malfunction even with the audio guidance processed in this way. By doing so, the power of the attenuated frequency in the original voice guidance is reduced, so that the deterioration of the voice quality of the processed voice guidance is reduced.

"Embodiment 3" An improvement method will be described when the deterioration of the voice quality of voice guidance processed by the method shown in Embodiment 2 is noticeable.

As a practical matter, even if audio guidance is input to P B H, all PB frequencies will not be detected by PBR. In other words, the voice guidance may include only part of the PB frequency components, or the level of the included PB frequency components may be low.

In this case, for example, if the voice guidance is input to the PBH and, for example, *'' is detected in the PBR, either or both of 941Hz and 13368z in the voice guidance may be attenuated.

As in Example 1, this attenuation process is performed by Fourier transform, and 941Hz or/and 1336Hz in the power spectrum is
After attenuating only Hz, the power spectrum may be subjected to inverse Fourier transform. The voice guidance obtained in this way has even less deterioration in voice quality.

"Effects of the Invention" As explained above, according to the present invention, voice guidance that does not include the PB signal frequency that would cause the PBR to malfunction can be created. Therefore, even if the voice guidance created by this invention is manually input to the PBR, the PBR will not be able to reproduce the voice.
Does not cause malfunction when detected as a signal. Therefore PBR
Even if the user uses the PB while the voice guidance created by this invention is being played, even if the PB does not include a circuit that improves the distinguishability from the voice, the conventional PB will not work.
There is also an advantage that R will not malfunction.

The present invention can be used for voice processing when the PB and voice need to coexist in network services that use a PB and voice guidance, such as message dialing.

[Brief explanation of drawings]

Fig. 1 is a block diagram showing Embodiment 1 of the present invention, Fig. 2 is a diagram showing the sequence of message dialing currently used, and Fig. 3 is a simple explanatory diagram of the reason why voice guidance is input to the PBR. It is.

Claims (1)

[Claims] (1) In a service system that uses a telephone line and plays voice guidance and the user presses a push button in accordance with the voice guidance, the voice is characterized by attenuating at least part of the frequency band of the push button signal from the voice guidance. Guidance processing method