[mythtv-users] BBC HD test card

John Pilkington J.Pilk at tesco.net
Fri Mar 22 18:58:12 UTC 2013


On 22/03/13 11:46, John Pilkington wrote:
> On 22/03/13 10:18, David Watkins wrote:
>>> I just recorded the "BBC HD audio sync test" (about 2 min) and the
>>> "BBC HD
>>> test card & 5.1 audio channel check" (about 8 min) from about 0902
>>> and 0952
>>> on DVB-T2 BBC HD.  I cut them from a continuous recording with
>>> MythDVBcut
>>> Around 220 MiB in total.
>>
>>
>> I've recorded these also.
>>
>> Doesn't cutting them out of a longer recording pose a risk of altering
>> the audio/video sync, and hence make the value of the audio sync test
>> a bit dubious?  I know that loss of audio sync is one of the problems
>> I've had when trying to transcode recordings.
>
> MythDVBcut in its UK HD mode just cuts the recorded stream at keyframes
> but doesn't modify it in any other way.  The display device should treat
> it just as it would a new start, and the TS format is, AIUI, designed to
> cope with transmission glitches.
>
> I think transcoding goes to PS;  open-source TS muxers seem thin on the
> ground.
>

What I said there isn't entirely true.  There's a one-packet offset (188 
bytes) between the keyframe and the cutpoint.  It seemed to work 
marginally better like that with Project-X but it may not be a good idea 
for the simple cut.  It's easy to change at the 'cat PXadj... > 
bytelist' line.

The main source of hiccups with a cut like this is the built-in A/V 
offset.  I just tried running MythDVBcut on an SD recording that I had 
experimentally cut in USEPJX=false mode.  Project-X inserted several 
audio frames at each cut, giving what might sometimes be a more 
civilised audio transition, but there's no overall loss of sync in 
either mode.  Unfortunately I don't think Project-X works with my HD 
recordings.

John P



More information about the mythtv-users mailing list