- From: Patrick Martin <patrick.martin.r@gmail.com>
- Date: Thu, 17 Oct 2013 10:50:17 -0700
- To: public-audio@w3.org
- Message-ID: <CAN_rxo1D8hjmCXS-kOJvPfw5Ywy9OdEc3gYnX2XOgekhR5-ekQ@mail.gmail.com>
I'm wondering what happened to the discussion on the audio artifacts (snapping, clicking, popping etc.) encountered when attempting to schedule audio buffers consecutively of differing rates. The specification/implementations still do not address the issue. The reason why this is so important is that most modern game-engines require these features in order to provide varied and streaming audio, the current solution it seems is to provide pre-canned audio files which are played back when events occur but this results in games sounding horrible, you get a few prerecorded audio pieces in a loop. The ability to continuously play back audio that has been synthesized in chunks is essential, calling a JavaScript node continuously is inefficient in a large application. The specification needs a way to schedule exact, and precise streaming playback of consecutive audio buffers. Explanation of the issue/test case: https://dl.dropboxusercontent.com/u/40949268/Bugs/beep.html Relevant emscripten Code: https://github.com/kripken/emscripten/blob/a08253a00a0c37563ffabb3163de8088bca23933/src/library_sdl.js#L1481 Original topic: http://lists.w3.org/Archives/Public/public-audio/2013AprJun/0546.html
Received on Friday, 18 October 2013 09:53:17 UTC