Abstract
A new method of transform encoding is presented which yields good quality voice at 9600 bits/sec. This method is far simpler to implement than similar coding techniques. The method incorporates the ears' insensitivity to the phase of the short-time spectrum. The log magnitude of the Discrete Fourier Transform is truncated and encoded using adaptive PCM. The phase is separately companded and encoded using PCM.
Original language | English (US) |
---|---|
Pages (from-to) | 1102-1106 |
Number of pages | 5 |
Journal | Conference Record - International Conference on Communications |
Volume | 2 |
State | Published - 1983 |
All Science Journal Classification (ASJC) codes
- Computer Networks and Communications
- Electrical and Electronic Engineering