Re[2]: [XviD-devel] rrv en/decoding

skal xvid-devel@xvid.org
10 Dec 2002 16:38:12 +0100


	Hi all,

On Tue, 2002-12-10 at 14:27, Radek Czyz wrote:
> Hello,
> 
> Tuesday, December 10, 2002, 9:45:52 PM, Pete wrote:
> 
> > * motion estimation needs modification to support rrv. current the
> > encoder uses the halfpel mv's, which are non-optimal. we need to perform a
> > sad 32x32 and 16x16 search, and scaleup the vectors. is anyone willing
> > to help out here, or give me some "pointers"?
> 
> Sure, I can do that.
> I'll just download the code and see what's it all about. If I won't
> understand anything, I'll ask.

	arghh... seems serious rrv enc is not that easy. 
	I'm sorry i can't give a help right now :(
	Maybe of little help: here's a gcc3-linux binary
	of my encoder/decoder:

http://skal.planet-d.net/coding/tmp4_i86linux_v00.gz

gcc2 version is:
http://skal.planet-d.net/coding/tmp4_i86linuxgcc2_v00.gz


	At least you can generate test bitstreams
	with rrv inside. => The '-red' option will
	alternate P-vop/Reduced P-vop, whilst the
	'-redf' one will simply force reduced
	encoding for all P-VOPs.

	There are also some ref bitstreams with rrv in it:

vcon-A1GE-15-ARTS-L*.bits  => only I-vops (bweuarrk)
vcon-A1GE-16-ARTS-L*.bits  => P-VOPs


	bye,

			Skal


(PS: if you experience greenish color bleeding, it's because
my f/iDct is still weak, regarding precision. Using 
the reference or C implementation ['-cpu c' or '-cpu ref'
options] should solve this)