[FFmpeg-devel] Integrating the mod engine into FFmpeg - what is the best design approach?

Sebastian Vater cdgs.basty
Tue Aug 10 00:07:07 CEST 2010


Ronald S. Bultje a ?crit :
> Hi,
>
> On Mon, Aug 9, 2010 at 8:46 AM, Michael Niedermayer <michaelni at gmx.at> wrote:
>   
>> A Audio decoder should return PCM. Doing something else is requireing all
>> applications that use libav to be changed. I dont see the point in going this
>> way
>>     
>
> PCM what? Float is useless if the soundcard supports only int. The
> idea was to use SAMPLE_FMT_something and wrap into audioconvert.c to
> do the actual sequencing to (and from?) PCM. Applications would use
>   

Could you clarify a bit the part between SAMPLE_FMT_S32 <=>
SAMPLE_FMT_SEQUENCER?

> audioconvert.c to convert between SAMPLE_FMT_* anyway, so it would
> still work without modifications to ffplay.c or API changes. Apps
> would not have to be aware of the libavsequencer API, and in fact we
> (mentors) have said repeatedly that he shouldn't focus on
> sequencer-specific stuff, but on things like transmuxing and so on
> first. I guess that didn't quite come across...
>   

Do you mean that SAMPLE_FMT_SEQUENCER simply takes a AVSequencerContext
as input and use the mixer to about that to SAMPLE_FMT_S32? Sorry, just
a bit confused about this now.

-- 

Best regards,
                   :-) Basty/CDGS (-:




More information about the ffmpeg-devel mailing list