Low-delay Coder Research Articles

Improved audio coding using a psychoacoustic model based on a cochlear filter bank

Perceptual audio coders use an estimated masked threshold for the determination of the maximum permissible just-inaudible noise level introduced by quantization. This estimate is derived from a psychoacoustic model mimicking the properties of. masking. Most psychoacoustic models for coding applications use a uniform (equal bandwidth) spectral decomposition as a first step to approximate the frequency selectivity of the human auditory system. However, the equal filter properties of the uniform subbands do not match the nonuniform characteristics of cochlear filters and reduce the precision of psychoacoustic modeling. Even so, uniform filter banks are applied because they are computationally efficient. This paper presents a psychoacoustic model based on an efficient nonuniform cochlear filter bank and a simple masked threshold estimation. The novel filter-bank structure employs cascaded low-order IIR filters and appropriate down-sampling to increase efficiency. The filter responses are optimized for the modeling of auditory masking effects. Results of the new psychoacoustic model applied to audio coding show better performance in terms of bit rate and/or quality of the new model in comparison with other state-of-the-art models using a uniform spectral decomposition. The low delay of the new model is particularly suitable for low-delay coders.

Low delay coder (< 25 ms) of wideband audio (20 Hz-15 kHz) scalable from 64 to 32 kbit/s

A low delay coder for speech and music signals sampled at 32kHz is described. Its algorithmic delay does not exceed 25 ms which enables audioconferencing applications without echo cancellation. Its bit rate is scalable between 64 and 32 kbit/s by steps of 8 kbit/s. The transmitter issues the binary code at 64 kbit/s with lower bit rate codes embedded in it. The receiver may operate at lower bit rates with gradual loss of quality. The proposed coder is based on a mixed scheme : the adopted solution contains elements from the CELP speech coder and frequency domain music coders. The perceptual signal is obtained in the time domain, then transformed to the frequency domain where bit allocation is calculated and transform coefficients are quantized. A first solution based on the dft is discussed, then a second solution based on a mdct with small overlap is applied. The quantization of these coefficients is done in the following way. First, a prediction of the whole spectrum is applied. Then, a mean- removed gain- shape split vq is used for amplitude spectrum quantization and a hierarchical 2- dimensional vq is used for phase spectrum quantization with amplitude correction. At the phase quantization stage, each codeword describing the selected vector index is split into parts corresponding to different bit rates. Due to the hierarchical codebook structure, truncated indices may be used, without much affecting the signal quality. Simulation results are presented and the robustness of the proposed coder is examined.

Wavelet packet filterbanks for low time delay audio coding

We study the application of wavelet packet filterbanks to low bit-rate transparent audio coding, taking the audio coders' delay requirements into account, and propose low-delay coders based on wavelet packet filterbanks. We first develop a method of comparison between filterbanks for perceptual audio coding by estimating the necessary bit-rate for a transparent compression. We use this comparison method in order to select the best filters for our audio compression scheme, from a large set of orthogonal and biorthogonal wavelets. Different wavelet filters may be used at different stages of the tree-structured decomposition with a constraint on the overall delay taken into account. The optimization is carried out with a simulated annealing procedure, proposing two wavelet packet filterbanks, exhibiting average and low delays. They are inserted in a complete audio coder that employs vector quantization and considers psychoacoustic models. The use of the proposed filterbanks leads to the design of a new bit allocation procedure, taking into account the lack of selectivity of the equivalent synthesis filters in a wavelet packet filterbank. The resulting audio scheme is validated through listening tests. The wavelet packet filterbanks are shown to be a promising tool for audio coding, especially for low-delay coding: with average delay, the quality of the wavelet packet filterbanks is as good as with MPEG-1 Layer-2, both with 80 kb/s, and when reducing the delay to 200 samples, 96 kb/s are needed to achieve the same quality.

Low bit-rate CELP speech coder with low delay

This paper presents a speech coding algorithm for a low bit-rate backward prediction coder with a bit-rate of less than 4 kbit/s, and a codebook design with a closed-loop training procedure. The coding system proposed is based on the backward prediction algorithm in LD-CELP and results in a 13.5 ms coding delay, which is 1/7 the length of delay in conventional forward prediction coders at the same transmission bit-rates. The codebook design is carried out in a closed-loop training process to generate the codebook so that we can represent harmonic portions of voiced speech appropriately. Finally, we carried out a simulation for the performance of the low-delay coder system trained in closed-loop method with the proposed initial codebook. We evaluated the effectiveness of the initial codebook with the optimal assignment and found that its reconstructed speech was clear enough to understand though it contained a slight element of noise.

Adaptive code excited predictive coding

A novel way to use the code excited linear prediction (CELP) concept that decreases the processing load while keeping the same speech quality is discussed. Rather than performing individual weighting of each candidate sequence, a global implementation of the perceptual weighting function at the codebook level is proposed. As a result, the analysis-by-synthesis procedure does not require the processing of all the candidate sequences through the synthesis and weighting filters; the complexity requirement of the algorithm is therefore much reduced. The concept is carried out with an adaptive codebook. Two fixed-point implementations of the adaptive CELP (ACELP) algorithm are reported: a 7.2 kb/s block coder (7 MIPS), and a 12 kb/s low-delay coder (11 MIPS). Both coders have been rated to provide the same quality as the 13 kb/s block coder adopted by the GSM for the European cellular telephone.< <ETX xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">&gt;</ETX>