[XviD-devel] Mode decision

[skal]

	Hi Gruel,

On Tue, 2003-03-25 at 12:23, Christoph Lampert wrote:

> hm, from a quick glance at your data I'd say: 
> 
> Sometimes trying as much INTER4V as possible is good (though slow),
> sometimes trying any INTER4V is bad (and slow), sometimes it doesn't
> matter at all (except for being slow). 
> 
> So be honest, this doesn't sound like a "we should take 60% as a
> compromise" (even though this would mean only 2% too large file for
> squares and less than 1% too large for Hollow Men), but it rather sounds
> like "we should find out _when_ inter4v is beneficial and when it's not
> and use that as a trigger".
	hmm, that's somehow what I proposed: trigger a 4v-search when
	max{sub-sad8} is 60% of best-sad16. And after the search, when
	refined sad8 are available, eventually decide to code the block
	4V for real, or not. The trick is that the initial sub-sad8 are
	evaluated with the best motion vector found so far by the 16x16
	search.

> After all, there is not reason why INTER4V
> should make things _worse_ at all in any case, except for a bad
> implementation.  
	or bad criterion: in your orignal mail, the figure you showed
	displayed a strong overlap of blocks wrongly qualified as 4MV.
	So there's *also* room for improving the final choice
	sad16<sum_of_sad8...

	Did I overlook sthg?

		Skal

> If it cannot be detected automatically, we could at least add a switch
> (anime<->reallife) and adapt the quality levels accordingly. 
> 
>