On Sun, 2005-07-31 at 20:11 +0100, Ed W wrote: > >RingBuffer.cpp > > > >There are a few places where it waits for a full 256K before doing > >anything. 30 frames may take as little as 100K so this can account > >for about 2-3 seconds. And some of the timeouts on waiting for the > >buffer to fill are LONG, and there is no mechanism for flushing to > >disk based on time rather than buffer fill... You might also want to > >start live TV at 90% of real speed and speed up to 100% when the > >excess buffer has grown to at least 2-3 MB ahead. > > > > > > I think also there is a lot of sync stuff waiting for playback to pause > and audio to pause, etc, etc. I have tried a few patchsets on this in > the past and also reduced the buffering and you can drop the wait times > by a couple of secs easily. I haven't submitted anything because I > don't really understand it all that well so it's a bit of a hack > > I hope someone who understands this better will do some optimisation > because it's currently really wasting a lot of unneccesary time. > > Finger crossed... > > Ed W > > P.S. I can't see any reason to sync on the audio pausing...
I spent some time playing with RingBuffer.cpp, but I couldn't affect channel change speed at all (just managed to screw up myth in various ways) - even a hack would be good - any suggestions as to what to change ? Can I see your hacked file ? Bye - Joel.
_______________________________________________ mythtv-dev mailing list [email protected] http://mythtv.org/cgi-bin/mailman/listinfo/mythtv-dev
