Menu
Latest
12 meetings · Prague, Jeju · 12–16 Oct
Reports
RAN1#126-bis · SA2#177

Glossary term · Services

EVS codec

Enhanced Voice Services codec

Services →

EVS codec is a 3GPP Enhanced Voice Services speech codec with an EVS Primary mode and an AMR-WB interoperable mode, delivering high-quality voice and bitstream interoperability with AMR-WB without transcoding.

Introduced
Rel-12
Where
Services › Codecs
Specifications
2 specs
EVS codec Description Purpose What changed Neighbours Specifications

Description

The Enhanced Voice Services (EVS) codec is the 3GPP speech codec specified in 26.445 and used in IMS/MTSI telephony as profiled in 26.114. It was designed to supersede AMR and AMR-WB for mobile voice, offering significantly improved quality and efficiency across a wide range of bit rates and operating conditions. EVS operates in two distinct operational modes that share a common framework but serve different interoperability needs.

The EVS Primary mode is the native, full-capability mode. It supports a wide range of bit rates (from low rates suitable for capacity-constrained conditions up to high rates for enhanced quality), multiple bandwidths (narrowband, wideband, super-wideband and full-band), and advanced features such as discontinuous transmission, comfort noise generation, error concealment, and jitter-resilient operation. Primary mode is used end-to-end when both endpoints support EVS, delivering the best voice quality and efficiency.

The EVS AMR-WB Inter-Operable (IO) mode provides backward interoperability with the AMR-WB codec. When EVS operates in AMR-WB IO mode, the speech frames it produces are bitstream-interoperable with AMR-WB: frames generated by an EVS AMR-WB IO encoder can be decoded by an AMR-WB decoder, and frames generated by an AMR-WB encoder can be decoded by an EVS AMR-WB IO decoder, both without transcoding. This is a crucial deployment property: it allows EVS-capable and AMR-WB-only endpoints to communicate directly, avoiding transcoding resources and the associated quality loss and latency.

In network architecture, EVS is negotiated via SDP within IMS using the mechanisms of 26.114 (MTSI). Codec mode, bit rate and bandwidth are negotiated per session, and the media gateway or UE selects appropriate modes. The dual-mode design means an operator can deploy EVS while retaining full interworking with the large installed base of AMR-WB devices and networks. EVS also supports features relevant to VoLTE and VoNR, including robust handling of packet loss and adaptation to radio conditions.

Key components include the codec's source coding and decoding algorithms, its mode-switching logic, and its interoperability layer with AMR-WB. The codec's flexibility — variable bit rate, bandwidth and mode — lets it adapt to network conditions and terminal capabilities, which is central to delivering consistent voice quality over mobile networks.

Purpose & Motivation

EVS was created to raise mobile voice quality and efficiency beyond what AMR and AMR-WB could deliver, while preserving backward compatibility with the large installed base of AMR-WB equipment. AMR-WB improved on AMR by extending to wideband audio, but operators needed a codec that could deliver super-wideband/full-band quality, operate efficiently at low rates under poor radio conditions, and still interoperate with existing networks. EVS addresses this with its Primary mode for best quality and its AMR-WB IO mode for seamless interworking.

The dual-mode design solves a classic migration problem: how to introduce a superior codec without forcing transcoding or stranding legacy devices. By making EVS AMR-WB IO frames bitstream-interoperable with AMR-WB, 3GPP ensured that EVS and AMR-WB endpoints can communicate directly, avoiding media gateway transcoding, extra latency and quality degradation. This made EVS deployable incrementally across VoLTE and later VoNR networks.

What changed in its specifications

The latest release-to-release comparison of specifications that define EVS codec. Each summary describes the whole specification, not only this term.

TS 26.114Rel-18 → Rel-19

Rel-19 adds a reference to the IVAS fixed-point C code specification in TS 26.114. It also lists that specification among supported specifications for MTSI immersive audio clients.

Full comparison in TS 26.114 →
TS 26.445Rel-17 → Rel-18

Rel-18 removes the EVS channel-aware coding clauses and several EVS codec internals from TS 26.445. It also adds new definitions for frame, EVS codec and its operational modes, plus abbreviations.

Full comparison in TS 26.445 →

Release Timeline

Evolution Across Releases

An overview written from the defining specifications, not a verified change history. For the exact text, open the specifications listed below.

Rel-12 Initial

Initial definition: EVS introduced with two operational modes — EVS Primary mode and EVS AMR-WB IO mode — with IO-mode frames bitstream-interoperable with AMR-WB, allowing decoding without transcoding in both directions.

Definition carried over unchanged; EVS continued to be specified in 26.445 and profiled in 26.114.

Definition carried over unchanged.

Definition text was reformatted and reworded (e.g., quotation style around mode names and a changed reference number for the AMR-WB codec), but the technical substance — two modes and bitstream interoperability without transcoding — remained the same.

Definition carried over unchanged from the Rel-12/13/14 form.

Definition text reverted to the reformatted style (quoted mode names, updated codec reference), with no change in technical meaning.

Definition carried over unchanged.

Definition carried over unchanged.

Definition text again presented in the reformatted style (quoted mode names, updated codec reference), with no change in technical substance.

Defining Specifications

3GPP specifications that define or reference EVS codec, with the latest known release. Sourced from the 3GPP document catalog — see methodology.

SpecificationTitleRelease
TS 26.114 vk11 IMS Multimedia Telephony Client Rel-20 read
TS 26.445 vj10 EVS Codec Specification Rel-19 read