Implementation of an object audio system based on MPEG-4 audio lossless coding on DSP

A Lossless Multichannel Bio-Signal Compression Based on Low-Complexity Joint Coding Scheme for Portable Medical Devices

“…To reduce the hardware complexity through the selection operation of the reference channel for joint coding, we use the Rice code and Block Gilbert Moore code (BGMC) [ 4 ] for choosing the reference channel before entropy coding. For MPEG-4 ALS entropy coding, the bit stream size of the residual signal can be estimated from the Rice parameter, which determines the required data size to encode one sample [ 8 , 9 ]. The average residual signal, AbsMean , for the Rice parameter is calculated as follows: where e ( n ) is the residual signal and n is the index of the sample in the frame.…”

“…In forward linear prediction, the optimal predictor coefficients are initially estimated for each block by the autocorrelation method as shown in Figure 6 . This method, which is based on the Levinson-Durbin algorithm in the MPEG-4 ALS standard, has the advantage of providing a simple means to iteratively adapt the order of the predictor [ 8 , 11 ]. After calculating LPCs, the values are used for short-term prediction and are generally implemented using a finite impulse response (FIR) filter.…”

“…where R yy is the auto-correlated signal, a i k corresponds to the i- th coefficient of the k- th filter order, and E is the calculated prediction coefficient. Direct quantization of the predictor coefficient is inefficient because small quantization errors may result in large deviations [ 8 ]. For this reason, the coefficients are quantized based on the following function: …”