[XviD-devel] rrv decoding

[Marc FD]

> ive just commited skals reduced resolution code, and some decoder
> modifications. great work skal, your code really rocks ;-)
> xvid now supports r.r.v. decoding for ivop and pvops.

damn !! the first time XviD can decode something that he is unable to encode
!!!

> i will now investigate encoding support.

cool. lemme know if i can help.

to choose between i/ri/p/rp/b vops, that's many possibilites.
is it possible to know the following frames (i think 2-3 frames will be
enough)
to fast scan them (with something like AnalyseME) and to choose rrvs
(more or less interleaved) when frames will change a lot ?

BTW, i wonder what would rrvs row do against really reduced resolution.
and how rrvs ivops looks, ect....

> btw, if anyone's bored, and would like to do some easy assembly
> grunt work, consider mmx/sse2'ing transfer16x16_copy() and
> transfer32x32_copy(). these funcs are located in
> xvidcore/src/utils/mem_transfer.h.

i'm not bored, but if i grab 20 minutes tomorow, i'll do it.

Cheers,
MarcFD