[MPlayer-cvslog] r19162 - trunk/libmpcodecs/ad_hwmpa.c

Nico Sabbi nicola_sabbi at fastwebnet.it
Sun Jul 23 10:48:18 CEST 2006


Reimar Döffinger wrote:

>Hello,
>On Sun, Jul 23, 2006 at 12:14:14AM +0200, nicodvb wrote:
>  
>
>>the value returned  by decode_audio() must be the amount of _decoded_ bytes (finally it's beginning t work: no more toomanypackets... message)
>>    
>>
>
>I really don't think so, it is the number of bytes placed into buf, so
>the old code was actually correct.
>  
>

are you sure? even hwac3 works like the current hwmpa (returning teh 
amount of
"decoded" data corresponding to the amount of ac3 samples passed 
through: 6144)

>Maybe you can describe the problem a bit more? 
>

with the old code decode_audio() was being called far too many times, 
causing
desync and excessive decoding (too many packets in the buffer stuff because
sh_audio->pts was advancing much faster than sh_video->pts); the current 
code
behaves correctly
I tend to believe that decode_audio() in dec_audio.c expects to deal with
decoded samples, not pass-through ones such as mp2 and ac3

>Or how to reproduce without special hardware? 
>

-ao null exhibited the same behaviour

>Maybe it really is impossible to fix without
>doing that kind of "padding", but this "solution" still is a hack...
>
>  
>

perfectly in line with libao2 IMHO, but if you have a better solution 
I'm all ears^W^W eyes





More information about the MPlayer-cvslog mailing list