G722

G.722 is an ITU-T standard for wideband audio coding that aims to improve speech intelligibility in real-time communications by extending the audible bandwidth to about 7 kHz. The codec samples audio at 16 kHz and can encode mono streams at nominal bitrates of 48, 56, or 64 kilobits per second; stereo operation is supported by encoding two channels, effectively doubling the bitrate.

G.722 is widely used in Voice over IP, video conferencing, and telepresence systems where higher audio quality is preferred over traditional narrowband codecs. It provides low latency suitable for conversational speech and is commonly carried by standard real-time transport protocols such as RTP.

Technically, G.722 uses a split-band coding approach that processes the input into sub-bands and encodes them with a light-weight coder to achieve efficient bandwidth usage. This design enables relatively high-quality speech at the modest bitrates it supports, making it suitable for network conditions typical in real-time communications.

G.722 is the original member of the G.722 family developed by ITU-T. Related variants include G.722.1 and G.722.2 (AMR-WB), which address different bandwidths and use cases to widen adoption. The codec remains widely implemented in contemporary VoIP platforms and conferencing solutions, often in combination with common signaling and media transport standards.