JPH04324900A - Voice codec with comparison attenuator - Google Patents

Abstract

PURPOSE:To improve regenerated voice quality by reducing the decrease of error ratio due to influence of a wireless transmission line when a voice codec of adaptive difference PCM type is used in a communication network device having a wireless transmission section. CONSTITUTION:A regeneration signal (e') given to an adaptive predictor 7 in an adaptive difference PCM voice encoder is made into a regeneration signal where an impulse voice signal is restrained with a prescribed threshold value by a comparison attenuator 8, and the regenerated voice signal (j) and the control variable (g) of a decoder are changed into signals (j') and (g') where an impulse signal is restrained by a comparison attenuator 10, all of which are given to an adaptive predictor 11 and an adaptive reverse quantization device 6 respectively.

Description

[Detailed description of the invention]

0001

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to measures against transmission errors in a speech codec in a speech encoding system.

0002

2. Description of the Related Art In recent years, applications have been considered in digital communications as well, such as replacing part of a wired network, typified by current analog cordless telephones, with wireless transmission. In this case, in consideration of compatibility with the wired network, it is appropriate to use a conventional audio code/decoder, such as adaptive differential PCM (ADPCM). However, in wireless transmission, the occurrence of transmission errors is unavoidable, and since digital systems are more prone to transmission errors than analog systems, some kind of countermeasure is required. This can be reduced to some extent by techniques such as retransmission if real-time characteristics are ignored, but since real-time characteristics (delay time within 10 to 20 ms) are essential for voice signal processing such as conversations, retransmission techniques cannot be applied. It is also possible to use an error correction code, but this is not appropriate since it would inevitably increase the amount of information. Therefore, if it were possible to take measures against errors in the speech codec itself, it would be of great value.

0003

One example of this is a speech codec/decoder with a comparison attenuator, which has been separately filed by the present inventor. However, since no measures are taken against transmission errors in control variables, the effects of transmission errors remain on subsequent data, and the quality of reproduced audio cannot be said to have been improved sufficiently. FIG. 4 is a waveform diagram showing the difference in error rate between the waveforms of reproduced audio signals, and shows the reproduced waveform in a 32 kbps ADPCM audio codec as an example of the influence of transmission errors. FIG. 4(A) shows the reproduced waveform when there is no error, and FIG. 4(B) shows the reproduced waveform when an error with an error rate of 10-2 occurs. In FIG. 4(B), an impulse-like signal with a very large amplitude that is not seen in FIG. 4(A) is superimposed, but it is observed. The main ones are indicated by arrows in the figure. This impulse-like signal is very unpleasant to the ear. Figure (C) shows a reproduced waveform in an ADPCM audio codec with a comparison attenuator, which was separately proposed as a countermeasure for such a problem. Impulsive signals due to transmission errors have decreased, but some still remain. An object of the present invention is to provide a speech codec with a comparison attenuator that can reduce the influence of transmission errors on the reproduced sound, which is a problem in the prior art, and reduce the deterioration in the quality of the reproduced sound of the codec. This is what we provide.

0004

[Means for Solving the Problems] A speech codec with a comparison attenuator according to the present invention quantizes the difference between the predicted value from the adaptive predictor and the input speech signal using an adaptive quantizer according to the level and encodes the result. a differential encoding encoder that sends out a encoded signal, and a residual signal obtained by dequantizing the received encoded signal using an adaptive dequantizer according to a quantization scale factor from a quantization scale factor adaptation section. A speech codec includes a predicted value from an adaptive predictor and a decoder that outputs a reproduced audio signal based on the predicted value. a comparison attenuator for attenuating the amplitude and applying it to the adaptive predictor; The present invention is characterized by comprising a comparison attenuator that attenuates the scale factor and applies the attenuated scale factor to the adaptive inverse quantizer and the adaptive predictor, respectively.

[0005]

DESCRIPTION OF THE PREFERRED EMBODIMENTS FIG. 2 is a block diagram showing an embodiment of a comparison attenuator which forms the essential part of the present invention. FIG. 2A is a block diagram of a comparison attenuator provided on the encoder side. The input reproduction signal Sr(k) is compared with the threshold value by comparator 1, and Sr
When (k) does not exceed the threshold, the input playback signal Sr(k) is output as is, and when it exceeds the threshold, the playback signal Sr'(k) attenuated through attenuator 2 is output. Switch the output using the switch 3 so that Figure 2 (B
) is a block diagram of a comparison attenuator for the decoder side. A comparator 1 compares the threshold value and the reproduced signal Sr(k), and based on the result, a switch 3 selects between the unchanged reproduced signal Sr(k) and the reproduced signal Sr'(k) passed through the attenuator 2. Switch and output. At this time, the switch 21 is controlled so that the control variable is outputted through the attenuator 20 at the same time. As described above, the amplitude of the reproduced signal output that is finally output is limited to within a predetermined level, and the control variables are also modified in the decoder. A speech codec with a comparison attenuator is constructed by incorporating this comparison attenuator into a speech encoder and a codec, respectively. The encoder attenuates the impulsive speech signal to a certain level to make it easier to detect impulse noise caused by transmission errors in the decoder. If you do this, the encoder will suppress the impulsive voice signal, but such cases are rare in general voice calls, and the difference caused by adjusting the attenuation amount of the comparison attenuator will be There is almost no deterioration in the quality of reproduced audio due to the provision of the comparative attenuator of the present invention.

Next, as an embodiment of the present invention, ADPCM
A case will be described in which each of the above-mentioned comparison attenuators is incorporated into a speech codec according to the method. FIG. 1 is a block diagram of an ADPCM speech codec with a comparison attenuator according to the present invention. The flow of the process is shown below. First, the encoder takes the difference (residual signal a) between the input audio signal and the reproduced signal one step before, that is, the predicted signal f. (2) The input residual signal a is adaptively quantized according to the level by the adaptive quantizer 4, and transmitted as the encoded output c. (2) The quantization width used at that time, that is, the quantization scale factor b, is derived by the quantization scale factor adaptation section 5. (2) Based on the value quantized by the adaptive quantizer 4 (encoded output c), the adaptive inverse quantizer 6 reproduces and outputs a reproduction residual d. (2) Regenerate the reproduced signal e using the reproduction residual d and the prediction signal f from one step before. (2) The reproduced signal e is input to the above-mentioned comparison attenuator 8, and when the value exceeds the threshold value, the value of the reproduced sound e is attenuated and a reproduced signal e' is output. (2) This reproduced speech e' is also input to the adaptive predictor 7 to generate a prediction signal f used for prediction in the next step. (2) Quantize the difference between the prediction signal f output from the adaptive predictor 7 and the input voice of the next step using the above procedure. This is the flow of processing on the encoder side.

On the other hand, on the decoder side, the encoded output c is transmitted from the encoder, and the quantization scale factor adaptor 9 calculates the quantization width, that is, the quantum The scale factor g is calculated and output. ■
The adaptive inverse quantizer 10 decodes the residual h by inputting the quantization scale factor g' and code c' processed by the comparison attenuator 12 in the previous step. (2) The adaptive predictor 11 outputs a predicted signal i using the decoded residual h and the reproduced speech j' one step before. (2) Add this predicted signal i and the reproduction residual h to synthesize the reproduced audio j. ■Playback audio
is input to the above-mentioned comparison attenuator 12, and when it exceeds the threshold value, it is assumed that a transmission error is included, and the reproduced sound j is attenuated and output. This is the reproduced audio signal j'. Also,
The quantization scale factor g used as a control variable in the next step is also attenuated to give the quantization scale factor g′
Output as . ■Similarly to the encoder, this reproduced audio j′
is input to the adaptive predictor 11 to generate a prediction signal i used for prediction in the next step. (2) The quantization scale factor g' is input to the adaptive dequantizer 10 and used for dequantization in the next step. This process is performed. FIG. 3 is a block diagram of a comparative attenuator in which numerical values of a specific example are written in the main parts of the present invention. In the case of an adaptive encoding method such as the ADPCM method, better results can be obtained by determining the threshold value from a variable that has a large relationship with the reproduced audio, such as the quantization scale factor, rather than leaving the threshold unchanged.

[0008] As a concrete example, the results of implementation when the threshold value is determined based on the quantization scale factor will be shown below. FIG. 5 is a waveform diagram showing the effect of the present invention, and
2 shows the implementation results of an ADPCM speech codec with a comparison attenuator. FIG. 5(A) shows the results when a comparison attenuator that does not process the control variables is used, and FIG. 5(B) shows the results when the comparison attenuator of the present invention is inserted. The threshold value at this time is 100× using the quantization scale factor Y(k).
2Y(k), and the attenuation coefficient was 0.9 for both the reproduced audio and the quantization scale factor. As shown in the figure, the effects of transmission errors are reduced by providing the comparative attenuator of the present invention, which also processes control variables, even when looking at the reproduced audio waveform. Quantitatively, the present invention improves the segmental SNR
We were able to achieve an improvement of approximately 0.5 dB. The unpleasant impulse-like crackling sound was also reduced in terms of auditory sense. Furthermore, the processing amount of the present invention is as small as 0.14 MIPS,
This could be achieved with approximately 3% of the throughput of the ADPCM codec (approximately 4.54 MIPS). The above examples are AD
Although the PCM audio codec has been described, the present invention can also be applied to audio codecs of other systems.

[0009]

[Effects of the Invention] As explained in detail above, by incorporating the comparison attenuator of the present invention into a speech codec, the impulsive noise that occurs when a transmission error occurs can be reduced, and the quality of reproduced speech can be improved by segmental SNR. This can be improved by about 6%. This enables higher quality communication in systems that have wireless sections where transmission errors are likely to occur. Also,
Since the amount of processing is extremely small, there is little increase in power consumption even when a comparison attenuator is added, making it extremely effective for wireless devices that require low power consumption.

[Brief explanation of drawings]

[Fig. 1] Block diagram showing an embodiment of the present invention

[Fig. 2] Block diagram showing main parts of the present invention

[Fig. 3] Block diagram showing main parts of the present invention

[Figure 4] Reproduction waveform diagram of 32kbps ADPCM audio codeccoder

[Figure 5] Reproduction waveform diagram showing the effects of the present invention

[Explanation of symbols]

1 Comparator 2, 20 Attenuator 3, 21 Switcher 4 Adaptive quantizer 5, 9 Quantization scale factor adaptor 6, 10
Adaptive inverse quantizer 7, 11 Adaptive predictor 8, 12 Comparison attenuator

Claims (2)

[Claims] Claim 1: An encoder using a differential encoding method that quantizes the difference between a predicted value from an adaptive predictor and an input audio signal according to the level by an adaptive quantizer and sends out a coded signal; A decoder that outputs a reproduced audio signal from a residual signal obtained by inversely quantizing the encoded signal using an adaptive inverse quantizer according to the quantization scale factor from the quantization scale factor adaptation unit and a predicted value from the adaptive predictor. The encoder includes a comparison attenuator that attenuates the amplitude of the reproduced audio signal to be equal to or less than a predetermined threshold value and supplies it to the adaptive predictor, and the decoder When the amplitude of the reproduced audio signal becomes equal to or less than a predetermined threshold value, the amplitude and the quantization scale factor from the quantization scale factor adaptation section are attenuated, and the adaptive inverse quantizer and the adaptive prediction 1. A speech codec with a comparison attenuator, characterized in that it is provided with a comparison attenuator for providing signals to respective signals. 2. An encoder using a differential encoding method that quantizes the difference between a predicted value from an adaptive predictor and an input audio signal according to the level using an adaptive quantizer and sends out a coded signal; A decoder that outputs a reproduced audio signal from a residual signal obtained by inversely quantizing the encoded signal using an adaptive inverse quantizer according to the quantization scale factor from the quantization scale factor adaptation unit and a predicted value from the adaptive predictor. The encoder includes a comparison attenuator that attenuates the amplitude of the reproduced audio signal to be equal to or less than a predetermined threshold value and supplies it to the adaptive predictor, and the decoder When the amplitude of the reproduced audio signal becomes equal to or less than a predetermined threshold value, the amplitude and the quantization scale factor from the quantization scale factor adaptation section are attenuated by independently determined attenuation amounts to perform the adaptation. 1. A speech codec with a comparison attenuator, comprising a comparison attenuator provided to each of the inverse quantizer and the adaptive predictor.