Opus is now an open source, royalty-free IETF standard. See Mozilla and Xiph announcements and congratulations to all involved.

This is a pretty big deal. It seems that Opus is superior to all existing audio codecs in quality and latency for any given bitrate. I will guess that for some large number of years it will be the no-brainer audio codec to use in any embedded application.

Will it replace the ancient (almost ancient enough for relevant patents to expire) but ubiquitous MP3 for non-embedded uses (i.e., where users can interact with files via multiple applications, such as on-disk music libraries)? If I were betting I’d have to bet no, but surely long-term it has a better chance than any free audio codec since Vorbis in the late 1990s. Vorbis never gained wide use outside some classes of embedded applications and free software advocates, but it surely played a big role in suppressing licensing demands from MP3 patent holders. Opus puts a stake through the heart of future audio codec licensing demands, unless some other monopoly can be leveraged (by Apple) to make another codec competitive.

Also, Opus is a great brand. Which doesn’t include an exclamation point. The title of this post merely expresses excitement.

…

I published an Opus-encoded file July 30. Firefox ≥15 supports Opus, which meant beta at the time, and now means general release.

To publish your own Opus encoded audio files, use opus-tools for encoding, and add a line like the below to your web server’s .htaccess file (or equivalent configuration):

AddType audio/ogg .opus

Hopefully the obvious large community sites (Wikimedia Commons and Internet Archive) will accept and support Opus uploads as soon as possible. Unlike their slow action on WebM. Speaking of which the Mozilla announcement mentions “working on the same thing for video”. I can’t tell whether this means submitting WebM (probably more specifically the VP8 codec) to the IETF or something else, but good luck and thank you in all cases. [Update: The proposed video codec charter starts from some requirements not mentioning any particular code; my wholly uniformed wild guess is that it will be another venue for VP8 and H.264 camps to argue.] [Update 20120913: Or maybe “same thing for video” means Daala.] [Update 20120914: Greg Maxwell comments with a precise answer below.]