JPS61234123A - System for encoding and demodulating acoustic signal - Google Patents

Abstract

PURPOSE:To reduce the occurring chance and scale of aural noises and distortion caused when acoustic signals are encoded and demodulated, by controlling the integration constant of an encoding system by detecting an encoding error or its cause and controlling the integration constant of a demodulating system also by using the signal used for controlling the encoding system. CONSTITUTION:A loop system composed of an analog comparator circuit 104, 1st temporary storage circuit 106, and a circuit which uses '110' as its fixed integration constant, is an encoding circuit which obtains an 1-bit encoded output on a line 109 synchronously to a synchronizing clock from a line 108, and an encoding demodulation feedback signal 103 advances following an input signal 101 and becomes an encoded signal 109 corresponding to the input signal 101. In order to maintain the operations of the whole system at an expected value through the integration constant of the 1st variable constant integration or variable constant integration similar circuit 110, an integration constant controlling circuit 213 controls the integration control by detecting the component of an acoustic signal, from which an encoding error is expected, or generates an electric signal corresponding to the error between an input signal and encoded signal. This signal is transmitted to the 2nd variable constant integration or variable constant integration similar circuit 115 by means of an integration constant controlling signal transmitting circuit 215 and demodulating conditions related to the integration constant is matched to the encoding side.

Description

DETAILED DESCRIPTION OF THE INVENTION There are many ways to encode audio signals. This proposal concerns a method for digitally encoding an acoustic voice signal and a method for demodulating the encoded signal, and is related to a method for reducing noise and distortion associated with encoding, which are audible problems.

Generally, acoustic signals have a wide operating range. High-speed analog-to-digital converters with 12 to 16 bits are usually used for digital encoding with good sound quality, but the drawback is that they are expensive.

- A 1-bit encoding method is available as an encoding method that is both capable and low cost. In this case, a high sampling frequency is required to compensate for the small number of bits, but there is a limit to this frequency due to restrictions on the operating speed of components for economic reasons. Even if the available performance is distributed without waste, a compromise must be found in any or all of the following items: coding noise, codeable frequency range, and coding amplitude range.

However, no matter how much compromise is made, at least one problem remains.

Specifically, when the timbre is clear, monotonous, and weak, the encoding noise is most audible, whereas when the tone is strong and contains high frequency components, high-frequency distortion is audible as dullness. . This proposal focuses on the fact that when it comes to acoustic signals, noise, distortion, and operating range are sometimes considered important, and coefficient accuracy is not so important. The present invention relates to an encoding method and demodulation method thereof, characterized in that the above-mentioned weaknesses are reinforced by controlling constants.

A detailed explanation will be given below using the drawings.

FIG. 1 is a block diagram showing an embodiment of the present invention. 10
Reference numeral 1 denotes an input signal terminal, into which an audio signal to be encoded is input. 102 is the input signal from 101;
One of the two input lines of the analog signal comparison circuit 4, 103 is the other, and 105 is its output. 106
is a first temporary storage circuit in which the digital output 105 of the analog comparison circuit 104 is temporarily stored in response to the synchronization signal 108. A synchronization signal 108 is supplied from a reference clock generation circuit 107. 109 is the output of the first temporary storage circuit, and this signal is also a 1-bit encoded signal of the input signal.

110 is a first variable constant integral or variable constant integral similar circuit. The integral constant for integrating the output 109 of the first temporary storage circuit can be varied. The integral in this case does not have to be a function with strict integral time characteristics; it can be a simple integral-like circuit such as a time constant circuit.The ways to change the integral constant are: 1) Continuously 1) Stepwise change; There is a method of controlling the integral amount by high-speed switching, but the explanation of the specific method will be omitted as it is not the essence of this proposal. 110 output 1
03 is tentatively referred to as an encoded demodulation feed balance signal in the following explanation, and is input to one of the analog signal comparison circuits 104. 211 is an input signal, and 212 is an encoded demodulation feedback signal, which is input to an integral constant control circuit 213, respectively. In the embodiment shown in FIG. 1, the integral constant control circuit 213 has two input lines, 212 and 211, but generally it has either or both of the input lines 211 and 212. The output 214, hereinafter tentatively referred to as an integral constant control signal, controls the integral constant of the first variable constant integral or variable constant integral similar circuit. 215 is an integral constant control signal transmission circuit, 216 is its output, and this signal is 1
The integral constant of the 14 second variable constant integral or variable constant integral similar circuit is controlled. 111 is an encoded audio signal read out from memory. 112 is a second temporary memory circuit
The readout signal from the memory is temporarily stored by a synchronization signal 113. 6 114 is the output and becomes the input of the second variable constant integral or variable constant integral similar circuit. 116 is its output, which is a demodulated acoustic signal. 117 is its output terminal. 311 is a digital memory circuit for encoded acoustic signals. 312 is a control signal for the memory circuit; 313
It is supplied from a memory control circuit. 314 is a synchronization signal supplied from the reference clock generation circuit 107 to the memory control circuit.

104 analog comparison circuits with 101 as input signals;
A loop system consisting of a first temporary storage circuit 106 and a circuit with a fixed integral constant 110 generates a 1-bit code in which a 1-bit encoded output is obtained on 109 in synchronization with the synchronization clock from 108. The operation of this loop proceeds in such a way that the encoded demodulation feedback signal 103 follows the input signal 101. Since the integration result of the encoded signal is 103 and that signal follows the input signal, 109 is the encoded signal corresponding to the input signal.

This encoding characteristic has a differential or differential analogous characteristic corresponding to the integral characteristic possessed by the first integral or integral analogous circuit.

At this point, the explanation of FIG. 1 will be temporarily interrupted, and the problems of the above-mentioned encoding loop according to the conventional method in which 110 is a fixed integral constant will be explained. Figures 2, 3, 4, and 5 are diagrams for explaining the waveforms of the encoded demodulated feedback signal of 103 following the input signal, and in each case, the dotted line is the input signal waveform, and the solid line is the encoded signal waveform. This is a demodulated feedback signal.

Figure 2 shows an example where the integral constant of 110 is large, the input signal waveform is monotonous, there are few high frequency components, or the input amplitude is small; in this case, tracking is achieved with a small error. Figure 3 shows an example where the input signal has a large high frequency component or a large amplitude with the same integral constant as in Figure 2, and the tracking speed is limited by the voltage increase speed limit determined by the integral constant, making it impossible to follow the input. A large error is occurring. This error appears as high-frequency distortion, resulting in a muddy sound to the auditory sense.

FIG. 5 shows an example in which the integral constant is small, the high frequency component is large, and the input amplitude is large. In this case, although an error occurs due to large vertical fluctuations, the partial average value follows the input waveform, so no distortion is felt in the sound quality obtained. Although the noise component increases slightly, since the input signal is large, the increase in noise is not perceptible to the auditory sense. Figure 4 is the 5th
An example is shown in which the input signal is monotonous and has a small amplitude with the same integration constant as in the figure.

Since the input signal is small, an increase in noise is felt, and residual noise especially when there is no signal becomes a problem. Purely theoretically, these tracking error problems can be brought as close to zero as possible by making the synchronizing signal frequency of 108 in FIG. 1 as high as possible and the integral constant of 110 as small as possible.
There is a limit to the performance of parts considering economic efficiency. In particular, due to the asymmetry between the operating speed of the analog comparison circuit 104 and the switching speed of the first temporary storage circuit 106, as well as the time coordination with the memory circuit, the cost will increase significantly if the requirements exceed a certain level. The explanation will now return to Figure 1. In order to solve these contradictory problems, this proposal detects the occurrence of encoding errors and the factors that cause them, controls the integral constant, and makes improvements to suppress auditory distortion and noise. be. The integral constant control circuit 212 has this role, and the first
The operation of the entire system is controlled to maintain the expected value through the integral constant of a variable constant integral or variable constant integral similar circuit. There are two types of integral constant control circuits. The first is a feedforward control type method in which components of the acoustic signal that are predicted to cause coding errors are detected and the integral constant is controlled. A specific example is a method of generating an electrical signal corresponding to the proportional value, differential value, second-order differential value, or composite value of an input signal, that is, the amplitude or change that causes noise or distortion. This is an attempt to automatically determine the integral constant according to the speed. The second type is a feedback control type, which generates an electrical signal corresponding to the error between the input signal and the encoded signal. That is, if the tracking error becomes large, the integral constant is made small to increase the tracking speed, and conversely, if the error becomes small, the integral constant is made large to reduce residual noise.

This integral constant control signal is passed through the second integral constant control signal transmission circuit 215, which demodulates the encoded signal.
It is transmitted to a variable constant integral or variable constant integral similar circuit, and the demodulation conditions regarding the integral constant are matched with those on the encoding side. Three embodiments of this integral constant control signal transmission circuit will be described. The first is a method of direct transmission, which is applied when the transmission delay time of the encoded signal by the digital memory system is small and when the time difference between the integral constant control signal and the demodulated signal does not cause any audible problem. The second method is to transmit data using a time constant circuit.

Although the delay time described above is rather large, it is applied when the time difference or magnitude difference between the integral constant control amount and the encoded audio signal does not cause a problem in terms of auditory sensation. The third method is to transmit data using a delay circuit that has the same delay time as the digital memory, and is applied when accurately demodulating encoded signals. Since the specific means of these integral constant control signal transmission circuits are not the essence of the present invention, their explanation will be omitted.

As explained above, the integral constant of the encoding system is controlled by detecting the encoding error or the factors that cause the encoding error, and the integral constant control signal is transmitted to the demodulating side, and the integral constant of the demodulating system is By controlling the constants, it is possible to reduce the chance of occurrence of auditory noise and distortion that accompanies the encoding and demodulation of acoustic signals, and to reduce their size.

[Brief explanation of drawings]

FIG. 1 is a block diagram for explaining an embodiment of the present invention. FIG. 2, FIG. 3, FIG. 4, and FIG. 5 are signal waveforms showing the problems of the conventional method.

Claims (1)

[Claims] 1. A circuit 1 having a function of generating an electrical signal affected by an error between an acoustic signal to be digitally encoded and a signal resulting from digitally encoding it; a circuit 2 having an integral or integral-like function in which the integral constant is controlled by an electrical signal output from the circuit 1;
A circuit 3 having a function of storing a digital signal in synchronization with a synchronization pulse and supplying its output to the circuit 1, and the circuit 2 having one input as the acoustic signal to be encoded.
a circuit 4 having a function of comparing the magnitudes of analog signals whose output is inputted to the other input and whose output is supplied to the input of the circuit 3; An acoustic signal encoding method characterized by obtaining a 1-bit time-series digital code of a signal. 2. An acoustic signal code in which the circuit 5 is a circuit having a function of generating an electrical signal affected by factors in the acoustic signal that cause errors in digital encoding, and the circuit 5 is used in place of the circuit 1 of claim 1 above. method. 3. A circuit 6 having an integral or integral-like function in which the integral constant for demodulating a digitally encoded acoustic signal is controlled by an electrical signal, and the circuit 2 according to claim 1 or claim 2 above. The above-mentioned circuit 6 is controlled by an electric signal controlling the integral constant of
A digitally encoded acoustic signal demodulation system comprising: a circuit 7 for controlling an integral constant of .