[mythtv-users] Alternatives to lossless transcode

mark perkins1724 at hotmail.com
Thu Nov 14 05:57:20 UTC 2013


On 14/11/13 07:38, Karl Dietz wrote:

> Your version of mythtranscode (looking at the logfile from the 6th) has
> known bugs in the handling of audio (avf and fifo mode) and video
> (lossless mpeg2 mode) timestamps that got fixed a bit less then 2 days
> ago in fixes/0.27 so it may be worth to update to the latest nighly and
> retest with the recording in question (if you still have it around)
> See
> http://code.mythtv.org/cgit/mythtv/log/mythtv/programs/mythtranscode?h=fixes/0.27
>
Thanks Karl - I see John Finlay's pts patch has been pushed to both 
fixes/0.27 and fixes/0.26 so the manual patch is now no longer required.
>
>
> The delay information from mediainfo appears to be the delta of the
> first audio audio timestamp and the first video timestamp. (found that
> out by searching their bug tracker, documentation is hard to find) As
> audio and video frames have different durations depending on how you
> cut away excess frames at the beginning you'll get different delays.
>

Thanks for the detail. It was probably too much wishful thinking on my 
part that I might have found an easy way to accurately compare audio / 
video sync, but it was a nice thought.

I updated ffmpeg (and mythtranscode) and re-ran Paul's --fifodir script 
and as predicted by Paul there was no change in the audio delay figure.

I also re-ran the updated mythtranscode --mpeg2 with post-processing by 
handbrake and it was completely shot as per before with massive (1sec) 
audio / video out of sync and a log full of errors. So not sure what it 
is about the file but it does not work with --mpeg2 mythtranscode but 
will work fine with --fifodir mythtranscode.

I've gone back and re-looked at a few segments to try and see if there 
is audio out of sync (using --fifodir) - to be honest if there is it is 
very very slight. The ProjectX cut file audio probably leads the video 
by as much as the --fifodir audio lags the video, it is very very 
slight. Probably less of a difference than what is introduced by the 
transmission path and electronic processing circuits of the various 
connected sound system devices.


More information about the mythtv-users mailing list