[mythtv-users] An option for watching tv
daworm at comcast.net
Fri Dec 1 17:08:26 UTC 2006
Brad DerManouelian wrote:
> Maybe you should replace your wife and see if that solves your slow
> channel-change issues.
That, sir, was uncalled for, and you know it.
> No one has said there is no problem. Just that there is no problem
> for them. The work-arounds are acceptable to most people. And quite
> frankly, changing channels on my DirecTivo took just about as long as
> it does with MythTV, so I've just come to accept the fact that
> changing channels takes that long. MT Cox Digital box took longer to
> change channels than MythTV with a DirecTV D-11 STB.
MythTV is my first PVR, and a first for my wife as well. Comparing to a
DirecTivo isn't of any use to me. As far as I know, DirecTivo was a
piece of crap. It must have been, you dropped it in favor of MythTV,
didn't you? (See, if you want to trade insults, I can play that game,
but I'd suggest taking that portion of the conversation off list.)
What I do know is this:
a) Changing channels with the ivtv utilities on a PVR-250 while
displaying video with MPlayer is near enough to instant as to make no
difference. This says that there is nothing in the capture, playback,
or channel change process that takes any significant amount of time.
This leaves the buffering process as the only possible source for slow
channel change times.
b) Buffering video to a file on Linux, and reading back that buffer on a
modern processor should not take 3 to 10 seconds to accomplish on a
combined FE/BE. Maybe it would over the network for a split system, but
with a 100BaseT network, even that is a stretch. The amount of data
isn't that large, and the bandwidth of the IDE channel, even if poorly
tuned, certainly isn't that low. Writing ten frames or so of video to
provide a buffer, then reading them back again to start playback, should
be well under a second on a local machine, and probably via a network as
well. So somewhere in this process, either many more than ten frames of
video is being written, or something else is eating up a lot of time.
If it is more video, then someone who knows the code (read: a dev)
should be able to find where this amount is specified, lower it,
recompile and test to see if channel changes are indeed faster. Your
average non-programmer can't do this. Many of us get our Myth systems
from a repository such as ATRPMS and can't do this. I don't run SVN
either, and won't have time to make this check until at least after the
> This is not a public soap box. Someone pays for the bandwidth to
> communicate. That someone has made his opinion on this subject clear
> that he does not want to "live with" topics coming up over and over
> again with the same complaints and no one offering solutions.
I've only responded to this thread for the first time today, and I
avoided the previous thread that spawned this one entirely. If a
legitimate concern for a non-trivial number of users is off topic, then
many more threads than this should be cut short as well.
I haven't smoked for 3 months, 2 weeks and 12 hours,
saving $479.54 and not smoking 3,196.94 cigarettes.
More information about the mythtv-users