User-Agent: Mozilla/5.0 (X11; U; Linux x86_64; en-US; rv:1.9.1b4pre) Gecko/20090401 Ubuntu/9.04 (jaunty) Shiretoko/3.5b4pre Build Identifier: Mozilla/5.0 (Windows; U; Windows NT 5.1; en-US; rv:1.9.1) Gecko/20090615 Firefox/3.5 I encoded a video multiple times with multiple commands in ffmpeg to test it on firefox. The following command produces a video which stops playing when the libvorbis encoded sound of the video ends, plus does not fire the ended DOM event. It just freezes: ffmpeg -i largeExampleVideo.wmv -t 11 -ss 530 -acodec libvorbis -vcodec libtheora sdaf.ogv It seems that the theora encoded video is some ms bigger than the sound encoded. On Shiretoko beta in my Ubuntu linux I've no problems. The video stops properly even if the sound has stopped earlier. On my Vista firefox v3.5; when I move the time scroller some ms after the position where the audio stops, the streaming (play) finishes properly and fires onended. This is why I believe the shorter audio is the reason the problem happen. When I encode the video without sound it plays off course :) Reproducible: Always Steps to Reproduce: Try to watch an .ogv (with shorter encoded audio than video) by using the new <video> tag in Windows Firefox Actual Results: The streaming will freeze when the audio ends Expected Results: The streaming should end when the longer duration one of video and audio ends Am using the original theme. Latest Versions of Web Developer and Firebug.
Would you provide an example video?
I've tried to encode an Ice Age Trailer with the hope that theora would re-create that bug combined with vorbis, but I was wrong. Maybe the trailer was too small, maybe decoding a .mov isn't like decoding a .wmv, maybe I was too damn unlucky.. Anyhow, here is the original video with the sound shorter than the video: http://depositfiles.com/files/i4kl6fmg2 ( to be downloaded only by adults please, it's just an example for the best of open source ;) )
Just installed firefox 3.5 on Ubuntu, same problem..
The linked testcase is now a 404. We fixed something like this ages ago in bug 493673, but as far as I can tell, that fix would've been included in Firefox 3.5.0 already. Do you still have a way to reproduce this?
Resolving incomplete. Need a testcase to reopen.