<div dir="ltr"><div class="gmail_extra"><div class="gmail_quote">On Sat, Dec 7, 2013 at 9:31 AM, Tom Dexter <span dir="ltr"><<a href="mailto:digitalaudiorock@gmail.com" target="_blank">digitalaudiorock@gmail.com</a>></span> wrote:<br>
<blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><div class="HOEnZb"><div class="h5">On Thu, Nov 14, 2013 at 9:39 PM, Jim Stichnoth <<a href="mailto:stichnot@gmail.com">stichnot@gmail.com</a>> wrote:<br>
> On Thu, Nov 14, 2013 at 4:13 PM, Gabe Rubin <<a href="mailto:gaberubin@gmail.com">gaberubin@gmail.com</a>> wrote:<br>
>>,<br>
>> Does this mean recordings affected by this in versions < .27 will be fixed<br>
>> once upgraded to .27 or that seek tables need to be rebuilt?<br>
><br>
><br>
> Seektables will need to be rebuilt as needed, using "mythcommflag<br>
> --rebuild". Pre-0.27 seektables map frame numbers to file offsets for<br>
> keyframes to enable relatively accurate seeking to a specific frame number,<br>
> but do nothing to help with accurate position and duration information, nor<br>
> to help map timecodes to frame numbers for accurate time-based seeking.<br>
> 0.27 adds a mapping from keyframe frame numbers to timecodes (using linear<br>
> interpolation for non-keyframes), which is used for accurate seeking and<br>
> position/duration display.<br>
><br>
> One caveat is that seeking/position/duration aren't yet working properly for<br>
> audio-only recordings, since there are essentially no video frames to sync<br>
> to.<br>
><br>
> Jim<br>
><br>
<br>
</div></div>I just finished upgrading from 0.25.3 to 0.27. Out of curiosity I<br>
just played one of the NBC shows that was behaving this way. After<br>
the upgrade, the behavior changed in that it actually thinks the<br>
recording is in fact 52 minutes rather than an hour, even if I rebuild<br>
the seektable.<br>
<br>
>From what you describe above however, it sounds like this would be<br>
expected and that, if anything, I'll likely see this sort of thing<br>
stop on new NBC recordings. Is that correct, that the new timecode<br>
data you describe would only be available for new recordings?<br><br></blockquote><div>It would be enlightening if you could dump the recording markup with "mythutil --getmarkup markup.xml --chanid <chanid> --starttime <starttime>" and post/attach the resulting markup.xml file.</div>
<div><br></div><div>Jim</div></div></div></div>