Acoustic Identity Transport
OAS Voice
Acoustic transport layer for DID-based identity. Transforms identity credentials into FSK-modulated audio signals with encrypted channels, ECDH key exchange, and replay protection.
FSK modulation (48kHz PCM, mono)
Ping/pong challenge-response handshake
ECDH ephemeral key exchange
AES-256-GCM encrypted channels
Reed-Solomon per-chunk error correction
Replay protection (nonce + timestamp + sequence)
Inside OAS Voice.
Every crate in the OAS Voice workspace, what it does, and how it fits together.
oas-voice
8.2K LOC
Acoustic transport layer. ggwave WASM binding, multiple transport backends (WebAudio, Node.js, raw PCM, loopback), 16-byte chunk framing with sequence numbering.
oas-voiceid
— LOC
Identity and encryption layer. Ping/pong handshake, ECDH ephemeral key exchange (Seal), AES-256-GCM encrypted channels, VoiceID passive/active API.
Cryptography
Reference languages
TypeScript
Integration path
Read the acoustic transport contract.
This catalog describes the separate acoustic transport. Upstream repository access is currently restricted; this catalog offers no public npm installation for Voice. Keep transport authentication separate from avatar presentation.
Read the integration guide →Start building with OAS Voice.
Read the documentation or explore the source.