<div class="gmail_quote"><blockquote class="gmail_quote" style="border-left: 1px solid rgb(204, 204, 204); margin: 0pt 0pt 0pt 0.8ex; padding-left: 1ex;">I got the STB shortly before SA4200 support was added to MythTV SVN, and<br>
by the time I got my cables all sorted out it was working. Initially I<br>
had some stability problems: every week or so, a firewire recording would<br>
fail, and I would have to power-cycle the STB as well as re-start the<br>
backend to get things working again.</blockquote><div><br>I have the SA4250 and have similar problems. <br></div><blockquote class="gmail_quote" style="border-left: 1px solid rgb(204, 204, 204); margin: 0pt 0pt 0pt 0.8ex; padding-left: 1ex;">
<br>
<br>
This was all taken care of after enabling BusReset in<br>
"libs/libmythtv/linuxfirewiredevice.cpp" (back when this was not enabled<br>
by default), and in subsequent SVNs this has been very reliable without<br>
changes. Thanks to whomever fixed that!</blockquote><div><br>as I am on mythdora 4, does the firewire_tester reset command do same? I haven't had any luck with firewire tester testing or resetting (I do have firewire tester svn from around December ).<br>
</div><blockquote class="gmail_quote" style="border-left: 1px solid rgb(204, 204, 204); margin: 0pt 0pt 0pt 0.8ex; padding-left: 1ex;"><br>
<br>
I have two remaining problems. The first, I don't know how to debug:<br>
I've only run into it twice. The backend claims to have recorded a show,<br>
yet the file simply does not exist. There were no unusual messages in the<br>
mythbackend log, just notations that the recording was started & stopped<br>
at the expected times. I've recently upgraded to SVN r16382, and will<br>
post again if I ever see this specific behavior.</blockquote><div><br>I have a file, but it's 0 bytes (I think this is a symptom of it just not working as above? )<br></div><blockquote class="gmail_quote" style="border-left: 1px solid rgb(204, 204, 204); margin: 0pt 0pt 0pt 0.8ex; padding-left: 1ex;">
<br>
<br>
On the SA4200HD, no matter how fast you enter the channel numbers, channel<br>
changes are slow. This is due to the box itself, not MythtTV. What can<br>
be a problem is the fact that MythtTV begins streaming before the channel<br>
change finishes.<br>
<br>
What I wind up with then is a file that has a brief section of HD<br>
material, followed by the SD stream that I'm trying to record. Since I<br>
cannot display HD material, I get a "Failed to Reinit Video" message and<br>
cannot watch that program.<br>
<br>
Occasionally I get "lucky" and will be able to sneak a "fast forward"<br>
command in before MythTV finishes the screen init, and I can then watch<br>
the SD programming, but without sound. I'm apparently skipping past<br>
something in the stream that is required to detect or set up the audio<br>
tracks.</blockquote><div><br>With HD recordings in order view to them on my SD tv, I have to view the recording on vnc and cut out the first few seconds before the channel change and then let the transcoder do it's thing to SD size. If I don't cut out the first few seconds from the SD channel, transcoding fails. Lastly, with the transcoded file, sometimes I get audio, sometimes I don't.<br>
</div><blockquote class="gmail_quote" style="border-left: 1px solid rgb(204, 204, 204); margin: 0pt 0pt 0pt 0.8ex; padding-left: 1ex;"><br>
<br>
What I've settled on is this change to<br>
"libs/libmythtv/firewiredevice.cpp":<br>
<br>
in function FirewireDevice::SetChannel, ~line 222, I added a 3 second<br>
sleep after the channel change, just before the function returns.<br>
<br>
It may be that a shorter sleep duration would work.<br>
</blockquote><div><br>Again, not being on svn, I can't do this. Hopefully this could be in the next release or some other method to improve it.<br><br><br><br><br><br></div></div>