Huge WAV files, tiny Opus files
Voice recorders and meeting tools often save WAV, which fills a disk fast: a two-hour interview in CD-quality stereo takes about 1.3 GB. Opus was designed for exactly this kind of material. It keeps speech clear at bitrates where other codecs fall apart, so the same interview fits into 30 to 40 MB with no trouble understanding a single word. For music, Opus at 128 kbps competes with MP3 at 192 kbps or more.
| Recording | WAV size | Opus setting | Opus size |
|---|---|---|---|
| 1 hour lecture | about 635 MB | 32 kbps mono | about 14 MB |
| 1 hour interview | about 635 MB | 48 kbps mono | about 22 MB |
| 1 hour of music | about 635 MB | 128 kbps stereo | about 58 MB |
Choosing the settings
- Mono for voice. Most microphones capture one channel, and many recorders simply copy it to both sides of a stereo WAV. Mono output avoids wasting bits on a duplicate.
- 32 kbps is plenty for a lecture or voice memo. Use 48 kbps when there are several speakers, a noisy room or music in the background.
- 96 to 128 kbps stereo for music. Above 160 kbps the difference is inaudible for nearly everyone.
Opus always encodes at 48 kHz. A 44.1 kHz WAV is resampled during conversion, which is normal and has no audible effect.
Who uses Opus
Opus is the codec behind WhatsApp and Telegram voice messages, Discord, and voice calls in web browsers. Developers convert WAV prompts and sound files to Opus for chat bots, voice assistants and web apps, because files load quickly on slow mobile connections. Archivists of spoken word use it to keep large collections small.
Before you convert
Background hum and hiss take bits away from the voice. Cleaning the WAV first with noise reduction makes low-bitrate Opus sound noticeably better. Setting a consistent level with normalize helps too. Keep in mind that some older devices, Windows Media Player and many car stereos do not play Opus; for those, WAV to MP3 is safer. Text tags are seldom present in WAV, so add them afterwards if you need them.