Abstract: A sub-band adaptive differential pulse code modulation/encoding apparatus includes means (102, 103, 104, 105) having a predetermined asymmetric impulse response for receiving an audio signal and band-dividing the received audio signal into a predetermined number of sub-bands, so as to obtain a plurality of band-divided sub-band audio signals, a plurality of quantization means (110, 111, 112, 113) for quantizing the band-divided sub-band audio signals with the predetermined number of sub-bands, and encoding means (115) for performing adaptive differential pulse code modulation/encoding of the quantized sub-band audio signals.

Abstract: An audio encoder converts an input sound signal into a plurality of compressed frame data pieces in an sound signal compression coder, determines the importance of each bit in a classification unit of a transmission line coder based on the decoding quality in the presence of a transmission error, and classifies the bits into a plurality of classes. The audio encoder selects one of the three types of processing including convolution coding and addition of CRC check codes, convolution coding only, and no coding, in descending order of importance in the presence of a transmission error for each class. Then, the audio encoder adds preamble information and a synchronization signal in a multiplexer to generate a bit stream. It becomes possible to suppress degradation of a decoded sound signal without additional redundant bits.

Abstract: An input digital audio signal is divided into sub-band signals in respective sub-bands. Scale factors of the respective sub-bands are determined on the basis of the sub-band signals for every frame. Calculation is made as to differences between the determined scale factors for a first frame and the determined scale factors for a second frame preceding the first frame. Absolute values of the calculated scale-factor differences are calculated, and data representative of the calculated absolute values are generated. The data representative of the calculated absolute values are encoded into data of a Huffman code. Sign bits are generated which represent signs of the calculated scale-factor differences. The sub-band signals are quantized in response to the determined scale factors for every frame to generate quantized samples of the sub-band signals. The Huffman-code data, the generated sign bits, and the quantized samples of the sub-band signals are combined into a bit stream.

Abstract: A speech coding apparatus and speech decoding apparatus to improve audio quality. The dequantized value obtained in dequantizing section 135 is input to adaptive bit assigner 140 per a predetermined number of frames such as a pitch period basis. Adaptive bit assigner 140 calculates an energy of the dequantized value, i.e., square sum of the dequantized value as a sample, output from each of ADPCM quantizers 130a to 130d, and based on the calculated energy of the dequantized value, determines the number of bits assigned to each residual signal to be quantized in respective one of ADPCM quantizers 130a to 130d.