[MPlayer-users] Is PSNR a definitive way of judging the quality of encoded files?

[RC]

On Sat, 24 Sep 2005 13:28:37 +0100
VJ <vijay.s.gill at gmail.com> wrote:

> I found that with mbcmp=4,cmp=2,subcmp=2 the PSNR was the highest
> among various values of cmp and subcmp. 

*cmp=2 may have a high PSNR, but it looks quite terrible, IMHO.
Incredibly blocky.  Great on animation, though.

> Increasing the value of cmp and subcmp beyond does not make much
> difference but reduces the speed of encoding.

Umm, what?  The values of the *cmp options aren't in any order
(greater values don't mean better/worse), they're all completely
different methods. Check the man page for mbcmp.