audioAMR

audioAMR refers to the Adaptive Multi-Rate (AMR) audio codec family standardized by the 3GPP for speech coding in mobile networks. It comprises two main variants: AMR-NB (narrowband) and AMR-WB (wideband). AMR is designed to preserve intelligibility under varying network bandwidth and packet-loss conditions by supporting multiple bitrate modes that can be switched between in real time. The codecs are based on CELP (Code-Excited Linear Prediction) techniques and operate on short frames, typically around 20 milliseconds, enabling efficient error concealment and fast adaptation to network conditions.

Standardization and deployment: AMR was introduced in the early 2000s as part of 3GPP specifications for GSM and was later adopted in 3G and 4G networks. It has become a common primary codec for voice in mobile telephony, including VoLTE. AMR-WB provides higher-fidelity speech than AMR-NB, suitable for wider audio bandwidth while maintaining low bitrates and robust performance in mobile environments.

File formats and usage: Audio encoded with AMR commonly appears in .amr files, with AMR-WB using .awb in some contexts. MIME types used for transmission include audio/AMR and audio/AMR-WB. The format is widely supported across mobile devices, desktop players, and streaming pipelines, with both open-source and proprietary implementations available.

Licensing and limitations: AMR codecs are patented and typically licensed under FRAND terms through 3GPP patent pools. They are optimized for speech and are not intended for high-fidelity music; performance can degrade with non-speech content. AMR remains a cornerstone of real-time voice communication in modern mobile networks.