[MPlayer-dev-eng] comprehensive lavcopts test - in progress.

Roman Gaufman hackeron at dsl.pipex.com
Tue Jun 15 01:56:38 CEST 2004


I'm not sure about qpel, I hear it actually makes everything worse. But I 
havent tested it myself, so cant reply. I will add it to test.

As for mv0, I do run every single test in -lavcopts with mv0 as well as 
without mv0, I also run with mbd=2 and run with trell. So I do test how well 
every opt works with these main ones. As well as without them.

So yes, it is quite a lot of combinations, but really not quite as many as you 
tested :), I also dont see a reason to test that many (atleast for this first 
test).

After I run this test, I will see what performs best and start grouping 
together, which will lead to another test.

On Monday 14 June 2004 16:16, Romain Dolbeau wrote:
> Roman Gaufman wrote:
> > Well, this is the suggestions I want. So you say *cmp depends on qpel and
> > mv0 to work and doesnt work without it?
>
> No. I say that depending on whether or not you use mv0 and qpel,
> then the best possible value for *cmp change. Therefore, you probably
> need to test not only each option on its own, but also the possible
> interactions between options, and that makes a *lot* of combinations.




More information about the MPlayer-dev-eng mailing list