Computationally efficient adaptive bit allocation for coding method and apparatus

(57) Summary book invention relates to the elaborate consciousness model which calculates efficiently. By the consciousness model of the present invention, the foundation for useful adaptation bit assignment of encoded information and quantization can be offered with a quality low bit rate coding system. In the one case of the operation, an input signal is divided into a frequency subband signal by the audio division zone encoder, a subband signal is quantized according to the value decided with a quota function, and a quantization subband signal is assembled by the coded signal. By a quota function, a quota value is decided according to the sound psychology principle based on a masking threshold. A masking threshold is (). The power spectral density (PSD) of an input signal is presumed, () A stimulus pattern is generated by applying a diffusion function to PSD, () A stimulus pattern is adjusted in a quantity equal to the signal-to-noise-ratio (SNR) offset which is sufficient for attaining sound psychological masking, and it is (). The level of an adjusted pattern is compared with the threshold of hearing, It is decided by generating a masking threshold equal to any of an adjusted pattern level and a hearing threshold, or the larger one. A diffusion function can be performed by applying one or the un-going type filter beyond it to the input signal PSD. In a forward direction adaptive coding system, a quota value is explicitly transmitted to a decoder using the information assembled by the coded signal. In an opposite direction adaptive coding system, a quota value is explicitly transmitted to a decoder using the information assembled by the coded signal which assigns with a decoder quota function and can guide a value. In a hybrid adaptive coding system, a basic quota value is explicitly transmitted with the change quota value transmitted explicitly.