Adaptive Differential Pulse Code Modulation (ADPCM)

Like PCM codecs, ADPCM codecs use also waveform encoding. However instead of quantizing the speech signal directly, the difference between the speech signal and a prediction that has been made of the speech signal is quantized. This is based on the fact that the next speech sample can be predicted from the past samples. If the prediction is accurate then the difference between the real and predicted speech samples will have a lower variance than the real speech samples, and will be accurately quantized with fewer bits than would be needed to quantize the original speech samples. At the decoder the quantized difference signal is added to the predicted signal to give the reconstructed speech signal. Adaptive prediction and quantization are used to increase the performance of the encoder. The goal is to adapt the changing characteristics of the speech with the predictor and difference quantizer. The standard G721 has as bit rate of 32 Kbps and gives subjective quality almost as that of PCM, about 4.1 score on the ITU-5-point quality scale. ITU recommendations G721, G726 and G727 codecs which provide 40, 30, 24 and 16 Kbps as bit rates. Each speech sample is encoded by 2, 3, 4, 5 bits using ADPCM codecs. However, the quality of the compressed signal decreases as expected with the decrease of the output bit rate.