EVS (Enhanced Voice Services) Codec - High-Quality Voice & Audio for VoLTE and VoNR

The Enhanced Voice Services (EVS) Codec is a 3GPP-standardized speech and audio codec designed as the successor to AMR-WB. While AMR-WB is limited to wideband audio (up to 7 kHz), EVS supports narrowband, wideband, super-wideband (up to 14 kHz), and fullband (up to 20 kHz) — delivering significantly richer, more natural-sounding voice. Consilient Technologies provides optimized EVS encoder/decoder libraries for mobile, VoIP, and embedded platforms, with the codec already field-proven in VoLTE deployments.

EVS operates across a wide bitrate range from 5.9 kbps (narrowband) up to 128 kbps (fullband). At lower bitrates (5.9–13.2 kbps), it delivers comparable or better quality than AMR-WB at similar rates. At higher bitrates (24.4–128 kbps), it provides near-transparent fullband audio suitable for music-on-hold, rich conferencing, and premium voice services. This flexibility allows network operators to balance bandwidth costs against audio quality for different service tiers.

EVS includes built-in packet loss concealment (PLC) and jitter buffer management that significantly outperform AMR-WB under impaired network conditions. The codec can reconstruct up to 3–4 consecutive lost frames with minimal perceptible degradation. Consilient’s implementation integrates an adaptive jitter buffer that adjusts dynamically to network conditions, making it well-suited for VoLTE, VoNR (Voice over New Radio), and enterprise VoIP deployments where network quality can fluctuate.

Yes. EVS includes an interoperable mode (EVS-AMR-WB IO) that allows it to communicate directly with AMR-WB endpoints without transcoding. This means operators can deploy EVS on new devices while maintaining full voice interoperability with the existing AMR-WB device base. The transition can be gradual — EVS-to-EVS calls benefit from improved quality, while EVS-to-AMR-WB calls fall back seamlessly to the common AMR-WB mode.

The EVS Codec is primarily deployed in VoLTE and VoNR services by mobile operators upgrading from AMR-WB to deliver premium voice quality. It is also used in unified communications platforms that need fullband audio for HD conferencing, embedded systems in automotive and IoT where high-quality hands-free voice is required, and media gateways that bridge between mobile and IP-based voice networks.

Consilient provides EVS codec libraries as portable C source and optimized builds for ARM (Cortex-A, Cortex-M), DSP platforms, and x86/x64 server environments. We support Linux, Android, RTOS (FreeRTOS, VxWorks), and bare-metal systems. The codec is available in both fixed-point and floating-point variants, allowing deployment across everything from resource-constrained chipsets in handsets to high-density cloud media servers.

EVS requires roughly 2–3x the processing power of AMR-WB, depending on the operating mode and bitrate. The additional complexity pays for significantly better audio quality, especially in fullband and super-wideband modes, and improved resilience to packet loss. For embedded targets where CPU budget is tight, Consilient offers mode-restricted configurations that limit the codec to lower-complexity operating points while still outperforming AMR-WB in audio quality.

Consilient provides an evaluation SDK that includes pre-built libraries, API documentation, sample integration code, and 3GPP conformance test vectors (TS 26.442/443). We also provide MOS (Mean Opinion Score) benchmark data comparing EVS performance against AMR-WB at matched bitrates. Our engineering team supports the full integration cycle from initial evaluation through platform optimization and conformance testing.