Michael Drons wrote:
> Tom Metro wrote:
>> ...I've observed that the video stream can't keep up when I'm in 3x
>> FF mode (so far only tried with MythTV protocol), which results in
>> the video stuttering, and of course the buffer graphs show the
>> buffers emptying. But I've been dismissing that as an expected 
>> network bandwidth limitation.
> 
> As for FF, you can really only FF for as long as there
> is data in the video buffer. 

So then this is a common problem and an expected limitation?

(I'm cc'ing the user list, as probably others would benefit from this.)


> Try increasing the buffer in the setup menus of the mvp.  I set mine
> to 40000 and seem to get good performance, but it is really only
> about 30 seconds of FF ability.

I had tried that, but didn't really notice any difference. I may not 
have increased the size enough to be noticeable. If that number is 
bytes, then I wouldn't expect even 40,000 to make a difference. I'd 
expect an increase on the order of megabytes to be required.

I just tried setting the MythTV Program TCP Receive Buffer to 2,002,941 
(which took a while), and it had no apparent effect.

I had noticed that different encoding rates had an impact. Low quality 
video streams fewer bits per second, so of course it takes longer before 
the buffers get starved.

What I didn't know is whether there was something particular to my setup 
that was resulting in not being able to sustain the required bandwidth, 
or whether it was just an expected limitation.


> IMO you are better off hitting the skip button to jump 30 seconds or
> hitting the number keys to jump 1%, 2%, 3%, etc into the video.

Of course, but FF capability would be nice too.

I ran some bandwidth tests on several files with various encoding rates 
(all via MythTV protocol) and saw rates ranging from 7.3 Mbps to 11.55 
Mbps. Around 10 Mbps seemed average. If I recall correctly, that's in 
the same neighborhood as what others on the list have reported.

(I did see one reading of 40 Mbps, but that occurred when I did a 
bandwidth test on file B, after having just played a few minutes of file 
A. Is perhaps the byte counter not getting reset?)

During playback my "high quality" files required between 4 to 6 Mbps. 
With only 10 Mbps available, it's understandable that the buffers 
deplete. Observing FF showed the buffers never emptied completely 
(holding a partial frame?), but the video started stuttering after about 
5 seconds.

During playback my "low quality" files required between 1 to 2 Mbps. 
These I'd expect to be able to sustain FF - at least most of the time. 
But instead I observe the video buffer completely depleting after about 
6 seconds and the video stuttering after about 7 seconds.


Out of curiosity, if you're using VLC, which I haven't tried yet, does 
it support doing the FF processing on the back-end so that what is 
streamed to the client doesn't require any additional bandwidth? (This 
approach seems logical, though you still might run into disk and CPU 
limitations on the back-end.)

  -Tom


-------------------------------------------------------------------------
Take Surveys. Earn Cash. Influence the Future of IT
Join SourceForge.net's Techsay panel and you'll get the chance to share your
opinions on IT & business topics through brief surveys-and earn cash
http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV
_______________________________________________
Mvpmc-users mailing list
Mvpmc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/mvpmc-users
mvpmc wiki: http://mvpmc.wikispaces.com/

Reply via email to