On Mon, 10 May 2004, Matto Marjanovic wrote:
> The latter ("hasn't been implemented yet").
> There is not (yet?) a routine for "4:2:0-MPEG2 --> 4:4:4" in subsample.[ch],
Ah, ok. It works with 420JPEG - wasn't sure if 420mpeg2 was treated
as 420jpeg or not (not correct but until recently nothing really
bothered distinguishing between the two ;)).
> In lieu of a dedicated multitap MMX-accelerated supersampling routine,
> there is always "y4mscaler -O chromass=444"....
Actually my preference would be for an Altivec-accelerated version
of y4mscaler since I seem to be spending more time on the shiny new
dual G5 system than the IA32 systems <grin>.
Thanks for the quick reply!