[FFmpeg-devel] TS muxer issues -- some progress

[Alexandre FERRIEUX - FT/RD/SIRP/ASF/SOFTL]

Tomer Barletz wrote:
> On Fri, Jan 30, 2009 at 12:53 PM, Marc Mason <mpeg.blue at free.fr> wrote:
>> Alexandre FERRIEUX - FT/RD/SIRP/ASF/SOFTL wrote:
>>> ...
> 
>> Last time I looked (march 2008, trunk) ffmpeg PCR generation was a mess.
>> The muxer picked PCR values ignoring PSI insertion, which lead to huge
>> PCR jitter, which (I suppose) makes some hardware decoders throw a fit.
> 
> This was fixed about two weeks ago - check the logs.

Yes and I was excited to try this when Baptiste committed it but it 
didn't solve my problem :(

Can somebody comment on the strange regularity I observed ?


Also, among recent additions there is

  r16607 | bcoudurier | 2009-01-14 23:40:37 +0100 (Wed, 14 Jan 2009) | 1 
line

  warn if dts < pcr which means TS is invalid

and indeed I do observe this warning repeatedly.
But what is the stimulus leading to that situation since we are 
*producing*, not reading, the TS ?

>> Are there any good open source TS analyzers?
>>
>> I used Tektronix TS compliance analyzer to test ffmpeg's and vlc's TS
>> muxers, but I was told it was horribly expensive.
> 
> I would like to know too...

I use dvbsnoop but at a superficial level. What do you MPEG experts 
think of it ?

-Alex