Opus Audio Bandwidth: Narrowband vs Fullband

The Opus audio codec is a highly versatile, open-source format designed for both speech and general audio transmission over the internet. One of its most powerful features is its ability to scale across five different audio bandwidth settings: narrowband, mediumband, wideband, super-wideband, and fullband. This article explains the primary differences between these bandwidth options, detailing their frequency limits, sampling rates, and ideal use cases to help you choose the right setting for your audio application.

Understanding Opus Audio Bandwidth Options

In audio processing, “bandwidth” refers to the range of audio frequencies that a codec can reproduce, which is directly tied to the sampling rate (based on the Nyquist theorem, the maximum reproducible frequency is half the sampling rate). Opus dynamically adapts to network conditions and user preferences by offering five distinct bandwidth modes:

• Narrowband (NB): Captures an audio frequency range up to 4 kHz using an 8 kHz sampling rate.
• Mediumband (MB): Captures up to 6 kHz using a 12 kHz sampling rate.
• Wideband (WB): Captures up to 8 kHz using a 16 kHz sampling rate.
• Super-wideband (SWB): Captures up to 12 kHz using a 24 kHz sampling rate.
• Fullband (FB): Captures up to 20 kHz (the limit of human hearing) using a 48 kHz sampling rate.

Key Differences Between Narrowband and Fullband

The most dramatic contrast within the Opus codec exists between its lowest setting, narrowband, and its highest setting, fullband.

1. Audio Quality and Frequency Range

• Narrowband limits the audio spectrum to 4 kHz. This range is sufficient to understand human speech, but it discards the high and low frequencies that give voices natural warmth and clarity. The result is the classic “tinny” or muffled sound associated with traditional landline telephone networks.
• Fullband extends all the way to 20 kHz, capturing the entire spectrum of human hearing. This includes the deep bass notes and the high-frequency “air” and brilliance of musical instruments and vocal nuances, resulting in a rich, high-fidelity listening experience.

2. Bitrate Efficiency and Data Consumption

• Narrowband is designed for extreme efficiency. It can deliver intelligible speech at incredibly low bitrates, starting as low as 6 kbps. This makes it ideal for networks with severe bandwidth constraints, high latency, or packet loss.
• Fullband requires more data to maintain its high fidelity, typically operating best at bitrates from 32 kbps up to 510 kbps. While Opus can encode fullband audio at lower bitrates, higher bitrates are necessary to prevent compression artifacts in complex audio like multi-instrument music.

3. Primary Use Cases

• Narrowband is used almost exclusively for voice communication in low-bandwidth environments. Examples include emergency radio services, satellite communications, and VoIP calls over unstable mobile networks.
• Fullband is the standard choice for high-quality music streaming, competitive gaming voice chat (where spatial awareness and clarity are crucial), podcasting, and professional remote audio production.

Choosing the Right Bandwidth

When configuring Opus, the choice of bandwidth depends entirely on your audio source and network limitations. For pure speech applications where network conservation is a priority, wideband (16 kHz) is often considered the “sweet spot” because it provides clear, natural-sounding voice quality without the heavy data footprint of fullband. However, if your application involves music, ambient sounds, or requires professional-grade audio, fullband (48 kHz) is essential to ensure no audio detail is lost.