[MPlayer-dev-eng] [PATCH] new libvo driver: vo_ivtv [ask for review]

[Benjamin Zores]

> Well, it's your code of course, but somehow I prefer putting function
> type and name on one line as well as keeping the opening { on the same
> line as if and while constructs.  It makes the code more compact.  It's
> your decision.  I just looked over it superficially, all I found was
> some trailing whitespace, IOW just some very minor nits..

hum ... gnu coding standards or not, that's the question.
don't wanna start a new flamewar :p

>> +  if test "$_mp1e" = auto ; then
>> +    # we don't need mp1e
>> +    _mp1e=no
>
> Heh, mp1e..  Does this thing still work nowadays?  I thought it was
> looong obsolete..

copy/pasted from DXR3 configure.
have no idea whether it is useful or not.

By the way i do have one more problem with this vo after having spoke with
ivtv driver authors.

As previously stated, the card needs a few MPEG frames to be received
before displaying them. This produce an A-V out of sync that can't be
handled by MPlayer unless using -delay which is not accurate.

According to the authors, there's no way to handle this issue but to send
MPEG PES data with audio and video inside (i.e writing from a vo and one
from an ao won't work).

How can I solve this problem ?
I need to send a muxed MPEG A-V stream.

Ben