On Mon, Aug 06, 2007 at 10:55:56PM +0200, Klaus Schmidinger wrote:
> The way VDR currently handles different tracks is to identify them
> through their PES ids, and I'd like to keep it that way when going
> to HDTV.
In the end, it's an ID...
> But if TS works for you, that's fine, of course.
If vdr can handle both formats I can't see any disadvantage. Most TV h264
recordings "in the wild" are TS with PAT/PMT anyway... So I don't see the
need for repacking it in a non-standard way. A lot of RB customers with
Windows on the PC have problems with that format. A lot of popular editing
tools do not and will not support it, so they need to fiddle with ProjectX
first etc...
> What I am wondering about, though, is: how do you detect the frame
> borders in order to generate the index file? Don't you have to unpack
> the TS to "see" the actual payload, anyway?
All current h264 channels start a frame with the beginning of a new PES
packet. So when the PES-flag is set, the first two TS packets after that are
searched for the access unit delimiter. This optimization reduces the CPU
load considerably. For MPEG2, the RMM-remux already works similar when it
detects this behavior (new frame at the beginning of new PES). Surprisingly,
almost all "usefull" channels can benefit from this optimization. The ones
where it does not work have low data rate anyway...
--
Georg Acher, acher@xxxxxxxxx
http://www.lrr.in.tum.de/~acher
"Oh no, not again !" The bowl of petunias
_______________________________________________
vdr mailing list
vdr@xxxxxxxxxxx
http://www.linuxtv.org/cgi-bin/mailman/listinfo/vdr
