Opus Audio

Opus switches encoding modes per frame. The encoder analyzes each 20-millisecond audio frame and selects SILK-only for speech, CELT-only for music, or Hybrid mode where SILK handles frequencies below 8 kHz while CELT covers 8–20 kHz. A podcast alternating between a host speaking and background music transitions between engines without audible artifact.

The quality advantage is measurable. Controlled double-blind listening tests using the MUSHRA methodology show Opus at 64 kbps stereo matching MP3 quality at 128 kbps — the same perceived audio quality at half the file size. At 96 kbps, Opus reaches transparency for most content, meaning listeners cannot distinguish compressed from original. For speech at 16 kbps, the gap widens further — Opus delivers intelligible voice where MP3, AAC, and Vorbis produce speech-obscuring artifacts. Codec quality is one thing. Ubiquity is another — and Opus became the web's default audio codec faster than anyone predicted.

Every Video Call in Your Browser Uses Opus. Why?

RFC 7874, published in 2016, designates Opus as the Mandatory to Implement audio codec for WebRTC — the technology powering browser-based voice and video calls. Every compliant WebRTC implementation must include an Opus decoder, which means Chrome, Firefox, Safari, and Edge all decode Opus natively.

Google Meet, Zoom's web client, Microsoft Teams in the browser, Discord, and every other WebRTC-based platform transmits audio as Opus. Over 2 billion people use WebRTC monthly, and every one of them relies on Opus without knowing the codec's name. The IETF codec working group chose the hybrid approach over selecting either SILK or CELT alone because no single existing codec performed well across both speech and music at WebRTC's required bitrates.

Browser adoption happened gradually — Firefox added Opus in 2012, Chrome in 2013, Safari in 2017 — but today every major browser on every platform supports it, making Opus the most widely deployed audio codec on the web. Browser adoption is about the codec. The file on disk is a different question — because Opus isn't technically a file format at all.

Opus Isn't Actually a File Format. Then What's Inside a .opus File?

The distinction between Opus and an .opus file matters. Opus is the codec — the algorithm that compresses audio. An .opus file wraps Opus-compressed packets inside an Ogg container, the same container format used by Vorbis and FLAC. The container handles framing, seeking, and metadata; the codec handles compression.

Every .opus file begins with four bytes spelling OggS (hex 4F 67 67 53) at offset 0. The first Ogg page payload contains the OpusHead identification header at offset 0x1C: an 8-byte magic signature, version number, channel count, pre-skip samples the player must discard, original input sample rate, output gain in Q7.8 fixed-point decibels, and channel mapping family. These seven fields are all fixed-length, a single read of 19 bytes reveals the channel count, sample rate, and output gain of the audio stream.

The second Ogg page contains OpusTags — metadata stored as Vorbis Comment key-value pairs in UTF-8. Unlike MP3's ID3 tags, which were retrofitted onto a format without native metadata support, OpusTags are part of the Ogg specification from the start. Audio data follows in subsequent pages, each containing Opus packets encoding 2.5 to 60 milliseconds of sound. The file format details matter less than where people encounter them — and most people encounter Opus files without ever knowing the format's name.

Every WhatsApp Voice Message Is Opus. You Never Knew. Why?

WhatsApp encodes every voice message as Opus inside an Ogg container at approximately 24 kbps — MP3 needs 64+ kbps for comparable speech quality. When you record and send a voice message, WhatsApp captures microphone audio, encodes it with Opus's SILK mode optimized for speech, wraps it in Ogg, and transmits the file. The recipient's app decodes it.

WhatsApp is far from alone. Telegram voice messages use Opus. Signal voice messages use Opus. Discord voice channels stream Opus in real time. WhatsApp, Telegram, Signal, and Discord all use Opus for voice, but because apps encode and decode without user intervention, most users never learn the format's name.

The moment Opus becomes visible is when you try to use a voice message outside its original app. Export a WhatsApp voice message and you get an .opus file that older car stereos, some hardware MP3 players, and Windows versions before 10 (1903) cannot play. Converting to MP3 at 192 kbps preserves voice quality while ensuring universal playback on any device from the last two decades. FileDex handles this by converting Opus to MP3, WAV, or FLAC in the browser, with no file upload required.