How Opus Works
Opus is the modern, open audio codec that quietly powers most real-time audio on the internet, and its defining feature is versatility: a single codec that handles everything from a low-bandwidth voice call to high-fidelity music, switching between them on the fly. It achieves this by combining two underlying technologies into one hybrid codec. The first is SILK, a speech codec originally developed by Skype and optimized for the human voice. The second is CELT, a low-latency transform codec from the Xiph.Org Foundation built for general audio and music.
Opus can use SILK alone, CELT alone, or both together, and it decides which mode to use based on the content and bitrate, shifting seamlessly as the audio moves between speech and music within the same stream. This is why a video call can carry both a person talking and background music without sounding broken: the codec adapts in real time rather than forcing one compromise setting for everything.
Its technical range is exceptionally wide. Opus supports bitrates from about 6 kbit/s (intelligible speech at tiny sizes) up to 510 kbit/s (transparent high-fidelity audio), sampling rates from narrowband telephone quality all the way to full 48 kHz, and, critically, very low algorithmic delay, as little as a few milliseconds. That low latency is what makes Opus suitable for interactive, real-time use like calls and live streaming, where a codec that introduces a noticeable delay would make conversation awkward.
Why Opus Beats MP3 and AAC
In independent listening tests, Opus consistently outperforms older lossy codecs at the same bitrate, especially at low and medium bitrates where the quality gap is largest. At 64 kbit/s, Opus music sounds clearly better than MP3 or AAC at the same rate; for voice the advantage is even more pronounced. This efficiency, more quality per kilobit, is the main technical reason it has displaced older codecs in new applications.
It is also open and royalty-free. Standardized by the Internet Engineering Task Force as RFC 6716, Opus carries no licensing fees, so developers can build it into software and hardware without cost or legal friction. The combination of better quality and zero licensing is why it became the default rather than a niche alternative.
Where Opus Is Used
If you have made a voice or video call online in the last several years, you have almost certainly used Opus without knowing it. It is the mandatory audio codec for WebRTC, the technology behind in-browser calling, which means it underpins a huge share of internet voice and video. WhatsApp, Discord, Zoom, Signal, and Google Meet all rely on Opus for their audio. YouTube uses it for many audio streams, and it is increasingly common in podcasting and game audio.
Opus is normally carried inside an OGG container when saved as a file (the .opus extension), or inside WebM for video. So when you see a .opus file, it is Opus-encoded audio in an Ogg wrapper, the modern successor to the older Ogg Vorbis pairing.
Limitations
Opus's one real weakness is compatibility with old hardware. Because it is relatively recent, it is not supported by many legacy devices, older car stereos, basic MP3 players, and some aging software, where MP3 remains the safe universal choice. On the desktop and the modern web, support is essentially universal, but for a file that must play on unknown or old equipment, MP3 or AAC is still the more cautious pick. Like all lossy codecs, Opus also discards data permanently, so it is a delivery format, not an archival one; keep a lossless master in WAV or FLAC if you need to re-encode later.
