Currently, browsers rely on Flash to analyze, modify, and even create audio data online. Given that rich media seems to be of ever increasing importance, I feel that a set of methods to access audio data would be extremely useful. Creating and modifying audio would also be a nice feature but my primary concern is just audio access.
These concepts were originally suggested under an 'Audio Canvas' concept by Dave Singer in this list: http://lists.whatwg.org/pipermail/whatwg-whatwg.org/2008-July/015334.html
More recently, some of these features have been discussed and added experimentally over at Mozilla, which also served as the inspiration for this request: https://bugzilla.mozilla.org/show_bug.cgi?id=490705
Mozilla has issued a Draft specifications on their Audio Data API work:
Chris, should we close this now that (a) the Media Processing API would supersede this, and (b) the working group has adopted the Web Audio API?