Low-frequency-band component and high-frequency-band audio encoding/decoding apparatus, and communication apparatus thereof
View Patent ↗An audio encoding apparatus capable of improving a frame cancellation error tolerance without increasing a number of bits of a fixed codebook in a CELP type audio encoding. A linear prediction analyzer analyzes an input digital speech signal and outputs linear predictive coefficients. A linear predictive coefficients quantizer quantizes the linear predictive coefficients. A low-frequency-band component encoder encodes a down-sampled linear-predictive residual signal by a pulse-code-modulation encoder and generates low-frequency-band component encoded information, while a high-frequency-band component encoder encodes an error signal between a linear-predictive residual signal and an up-sampled signal of a decoded down-sampled linear-predictive residual signal by a code-excited-linear-prediction encoder and generates high-frequency-band component encoded information.
1. A speech encoding apparatus, comprising:
a linear prediction analyzer that analyzes an input digital speech signal and outputs linear predictive coefficients;
a linear predictive coefficients quantizer that quantizes the linear predictive coefficients;
a low-frequency-band component encoder that encodes a down-sampled linear-predictive residual signal by a pulse-code-modulation encoder and generates low-frequency-band component encoded information; and
a high-frequency-band component encoder that encodes an error signal between a linear-predictive residual signal and an up-sampled signal of a decoded down-sampled linear-predictive residual signal by a code-excited-linear-prediction encoder and generates high-frequency-band component encoded information.
2. The speech encoding apparatus according to claim 1 , wherein the high-frequency-band component encoder encodes the error signal using an adaptive codebook and a fixed codebook.
3. A communication apparatus comprising the speech encoding apparatus according to claim 1 .
4. The speech encoding apparatus according to claim 2 , wherein the high-frequency-band component encoder quantizes pitch lag information in the adaptive codebook based on the up-sampled signal of the decoded down-sampled linear-predictive residual signal.
5. A speech decoding apparatus, comprising:
a linear predictive coefficients de-quantizer that decodes quantized linear predictive coefficients;
a low-frequency-band component decoder that decodes a down-sampled linear-predictive residual signal by a pulse-code-modulation decoder and up-samples the down-sampled linear-predictive residual signal and outputs an up-sampled signal of the decoded down-sampled linear-predictive residual signal;
a high-frequency-band component decoder that decodes an error signal between a linear-predictive residual signal and the up-sampled signal of the decoded down-sampled linear-predictive residual signal by a code-excited-linear-prediction decoder, and outputs the decoded error signal; and
a synthesis filter that synthesizes a decoded speech signal excited by a summation of the up-sampled signal of the decoded down-sampled linear-predictive residual signal and the decoded error signal.
6. A communication apparatus comprising the speech decoding apparatus according to claim 5 .
7. A speech encoding method, comprising:
analyzing an input digital speech signal using a linear predictive analyzer and outputting linear predictive coefficients using linear prediction;
quantizing the linear predictive coefficients;
encoding a down-sampled linear-predictive residual signal by a pulse-code-modulation encoder and generating low-frequency-band component encoded information; and
encoding an error signal between a linear-predictive residual signal and an up-sampled signal of a decoded down-sampled linear-predictive residual signal by a code-excited-linear-prediction encoder and generating high-frequency-band component encoded information.