[FFmpeg-devel] [bug/patch] MPEG-TS muxer: PCR not in sync withPTS/DTS

[Alexandre Ferrieux]

(resending, not sure the intended dest got it ;-)

Hi Niobos,

> Niobos wrote:
>>
>> I reviewed this patch since I needed some additional features:
>> * Bugfix: In VBR mode, the SI-table period was calculated on the 
>> muxrate after it had been multiplied by 10. The resulting rates were 
>> 10x to slow.
>> * Extra feature: Specifying a negative muxrate on the commandline 
>> overwrites the calculated muxrate, but still activates VBR.
>> * I added an additional patch which includes some additional 
>> info/clarification and does not cause any changes in the compiled code.
>>
>> This extra feature is needed to produce a VBR MPEG-TS by 
>> remultiplexing a source ("-{a,v}codec copy"), since the bitrates can't 
>> be learned from the stearm.
>>
>> Attached patches are against r20080.


As usual I tested your update. Indeed the fix of the SI table frequency
has a nice effect on hardware decoders :-)

Now I observed a regression wrt your previous patches: when muxing an
audio-only TS, the PCR is wrong by a large factor (between 4 and 6 times
too slow). When muxing video only or both, it is correct.

-Alex