Open GoogleCodeExporter opened 9 years ago
Can this problem be solved if we would bought Flowplayer Support? We already
have a commercial version.
Original comment by vil...@lnk.lt
on 23 Jul 2012 at 7:42
Hi, please contact the ova developers in regards to your issue, you would have
support with them
http://www.longtailvideo.com/support/open-video-ads/ova-for-flowplayer
Original comment by electrot...@gmail.com
on 24 Jul 2012 at 5:15
Actually I've done so at first, but they redirected me here. Here are some
excerpts from the emails between me and OVA support:
Me to OVA support: "I‘m still not sure that it‘s related to
Flowplayer API. At least the workaround should be possible because as you
have mentioned the video starts to play and it IS there under the OVA
overlay. Switching to full screen clearly shows that. Check the screenshot
attached. It seems like OVA overlay just stays there after the ad is
finished."
OVA support: "Ok, but OVA has *no* control of that.
All OVA does is: _player.playInstream(clip)
where "_player" is Flowplayer.
Flowplayer is responsible playback (starting, stopping and closing the
instream clip). Clearly in this case, Flowplayer is not correctly closing
down the instream clip.
Honestly, there is nothing I can do - this issue definitely has to be
logged with Flowplayer."
I can try to get Paul from OVA to comment on this bug if you need more
information.
Original comment by vil...@lnk.lt
on 24 Jul 2012 at 9:01
HI don't see a pre-roll or mid roll happen. Here is a working config for the
basic functionality I am able to seek back and replay the mid roll many times.
flowplayer("player", "../build/flowplayer.swf",
{
onStart: function(clip) {
console.log("onStart: " + clip.url);
},
onFinish: function(clip) {
console.log("onFinish: " + clip.url);
},
onPause: function(clip) {
console.log("onPause: " + clip.url);
},
onResume: function(clip) {
console.log("onResume " + clip.url);
},
onSeek: function() {
console.log("onSeek");
},
clip: {
// baseUrl for both parent and instream clips
baseUrl: 'http://stream.flowplayer.org',
// duration for the parent clip
//duration: 10,
// instream playlist
playlist: [
// in the middle of the parent clip (ie. "mid-roll") starting at
// 5 seconds
{
url: 'honda_accord.flv', duration: 5, position: 2
},
{
url: 'http://pseudo01.hddn.com/vod/demo.flowplayervod/flowplayer-700.flv', duration: 5, position: 15
}
]
}
});
Original comment by electrot...@gmail.com
on 10 Aug 2012 at 11:40
Sorry for the delay. Our ad provider disabled commercials by accident.
I have enabled the ads again, can you try to reproduce it now? I can confirm
that this doesn't happen on HTTP streams but it does happen on RTMP stream
*and* instream API.
Original comment by vil...@lnk.lt
on 13 Aug 2012 at 1:02
Still cannot replicate any issues using rtmp
clip: {
// baseUrl for both parent and instream clips
url: "mp4:big_buck_bunny_480p_h264.mp4",
provider: "rtmp",
// duration for the parent clip
//duration: 10,
// instream playlist
playlist: [
// in the middle of the parent clip (ie. "mid-roll") starting at
// 5 seconds
{
url: "mp4:Elephants_Dream_720-h264-st-aac.mp4", duration: 5, position: 2
},
{
url: "mp4:big_buck_bunny_480p_h264.mp4", duration: 5, position: 15
}
]
},
I'll notify paul to see what is going on.
Original comment by electrot...@gmail.com
on 15 Aug 2012 at 3:55
More information available, please see:
http://flash.flowplayer.org/forum/7/104184
a) The bug only appears with "accelerated: true" and "wmode: direct" settings.
b) I have created somewhat reproducible example without OVA:
http://post.lnk.lt/flow/flow.html
1. Start the video and wait for the mid-roll at 15 seconds.
2. Wait for the mid-roll to finish and main video resumes playing. Memorize
last frame of the mid-roll (It's a guy standing with a briefcase).
3. Seek back to 10 second mark and wait for the mid-roll to start playing
again. You can see that second time it starts playing from the *last* frame of
the mid-roll from step 2 very briefly.
I think some kind of canvas or buffer is not cleared after mid-roll finishes.
The same last frame is not visible without HW acceleration.
Original comment by vil...@lnk.lt
on 24 Oct 2012 at 6:19
Original issue reported on code.google.com by
vil...@lnk.lt
on 18 Jul 2012 at 1:38