Digital
A digital secure voice usually includes two components, a digitizer to convert between speech and digital signals and an encryption system to provide confidentiality. What makes ciphony difficult in practice is a need to send the encrypted signal over the same voiceband communication circuits used to transmit unencrypted voice, e.g. analog telephone lines or mobile radios.
This has led to the use of Voice Coders (vocoders) to achieve tight bandwidth compression of the speech signals. NSA's STU-III, KY-57 and SCIP are examples of systems that operate over existing voice circuits. The STE system, by contrast, requires wide bandwidth ISDN lines for its normal mode of operation. For encrypting GSM and VoIP, which are digital anyway, the standard protocol ZRTP could be used as an end-to-end encryption technology.
Secure voice's robustness greatly benefits from having the voice data compressed into very low bit-rates by special component called speech coding, voice compression or voice coder (also known as vocoder). The old secure voice compression standards include (CVSD, CELP, LPC-10e and MELP, where the latest standard is the state of the art MELPe algorithm.
Read more about this topic: Secure Voice