[XviD-devel] hey I'm back :)

[Johannes Reinhardt]

Hi,

I will try do a lambda based hvs plugin as a replacement for the AQ
based lumimasking, but I don`t know how much spare time I have the next
days. And there are other effects that could be exploited.

I recently read some papers about foveating and automatic fixation point
detection. Do you think this could be useful?

Johannes



Quoting Radek Czyz <radoslaw at syskin.cjb.net>:

> Hello everyone,
> 
> I have summer vacations now, FINALLY :) and this means I'm back to
> xvid
> poking, with some fresh ideas and motivations.
> 
> First thing I'd like to do is release 1.1 final. How about we do
> this
> without waiting? I was thinking that we commit GomGom's Sparse patch
> (the pointer casting for fdct and idct is fugly, everything else is
> fine), bump version strings and have it behind us.
> 
> Future work I want to do, in no particular order:
> 
> - xvidenc - a commandline executable like x264 has. It turns out
> there's
> a demand for such executable, we already have a lot of code, and
> everything else can be taken from x264
> 
> - pre-pass ME. During my recent work, I discovered that motion
> finding
> over larger distances is often flawed and inefficient, and can be
> largly
> improved by pre-pass. I'm going to try and see what happens.
> 
> - Kopernikus' variable-lambda code. I believe lambda (VHQ and
> trellis)
> is the best way to modulate quality for all HVS applications. I want
> to
> give it a chance ;)
> 
> - multithreaded code. I have several ideas how various parts of XviD
> can
> be multithreaded, and I intend to implement them, if only to learn
> how.
> 
> - everything else that comes to my mind ;)
> 
> I'll start soon :)
> 
> Radek
> _______________________________________________
> XviD-devel mailing list
> XviD-devel at xvid.org
> http://list.xvid.org/mailman/listinfo/xvid-devel
>