This system reads a speech signal from a file, and encodes it at two bits per sample using adaptive differential pulse code modulation with a feedback-around-quantizer structure. It then reconstructs the signal from the quantized data.

Listen to the original sound and the processed sound.