[MPlayer-dev-eng] Patch: Low latency audio enhancements

[Ed Wildgoose]

>less invasive, does that mean that theres still some side effects and if
>so which(sorry im not too familiar with the ao internals)?
>  
>

Not at all.  It was just before I lobbied to change the ao internals to 
raise the latency a bit, ie increase the default buffer size.  There was 
some resistance to that.

However, at least with my 2.6 kernel I seem to be able to easily 
maintain latencies of 2 x 512 samples, while compiling stuff and 
watching a film, so I'm quite happy to leave the code the way it is 
now.  In fact for me it might even be a slight advantage because my card 
doesn't report free space correctly, only whether a buffer is empty or 
full, so the 512/1024 buffer sizes will track the audio and video more 
accurately.

I think the patch as submitted will basically fix a few minor problems 
for people who have low latency cards (they probably saw some wierd 
effects when fast forwarding), and reduces the latency of the main loop 
for people like me with audio cards with small buffer sizes.  It also 
fixes a minor bug in ao code.

I don't expect to see any negative impact from these changes.  However, 
I can at last finally use mplayer!  (Now I think I need to make the same 
changes to Xine and Ogle since they are broken for me as well)

Please apply to CVS

Thanks

Ed W