Re: Adaptive quantization help--variance masking

Are you absolutely sure that the code is being used?
I didn't bother looking if it has assembler
counterpart, so make sure no MMX/SSE/3DNow! function
is used instead of the C one.
To quickly be sure, compile it in debug mode and place
a breakpoint inside the function. If the debugger
stops there, then you're using it and something is
wrong with the code.
If the debugger never breaks, the code is simple not
being used.
You're sure that no "duplicate library" is used,
however XviD has "duplicate functions" optimized for
different system hardware.

Hope this helps
Dark Sylinc

--- Jason Garrett-Glaser <[email protected]>
escribió:

> I got bored and decided I would port x264's new VAQ
> to Xvid.
> 
> After 15 minutes, I got this patch:
> http://pastebin.com/f609010fc .  It just
> overwrites the lumimasking code.
> 
> However, all my attempts to compile Xvid result in a
> library that just does
> ordinary lumimasking, even though I've overwritten
> the code.  This isn't a
> "duplicate library" issue; I have confirmed that it
> is indeed my compiled
> version of Xvid that is being used.
> 
> What could the problem be--and could any other
> developer try throwing in
> this patch and seeing if it works?
> 
> Note the "strength" parameter can be varied
> arbitrarily.  It does what you
> think it does.
> 
> Dark Shikari, x264 developer
> _______________________________________________
> XviD-devel mailing list
> [email protected]
> http://list.xvid.org/mailman/listinfo/xvid-devel
> 



      Los referentes más importantes en compra/ venta de autos se juntaron:
Demotores y Yahoo!
Ahora comprar o vender tu auto es más fácil. Vistá ar.autos.yahoo.com/