[MEncoder-users] High quality encoding of anime and broken mencoder options

[Jorge Peixoto de Morais Neto]

On Fri, 07 Mar 2008 08:57:57 -0800
RC <cooleyr at gmail.com> wrote:

> On Thu, 6 Mar 2008 17:51:24 -0300
> Jorge Peixoto de Morais Neto <please.no.spam.here at gmail.com> wrote:
> 
> > In my case, the mbcmp=2 option is helpful, 
> 
> mbcmp=<0-2000>
> Sets the comparison function for the macroblock  decision, *only used if
> mbd=0.*
Then the documentation is wrong or there is a bug in mencoder. I have
tested it more than once. One of my early results:

vcodec=mpeg4:psnr:keyint=300:vmax_b_frames=2:vb_strategy=1:b_sensitivity=10:mbd=2:trell:v4mv:last_pred=3:dia=-1 with vbitrate=512: bitrate  512.661 , PSNR 36.01 in 46.661s
with vbitrate=640: bitrate  642.454 , PSNR 37.38 in 48.186s
bitrate for 36.4 PSNR would be 546.677799298

vcodec=mpeg4:psnr:keyint=300:vmax_b_frames=2:vb_strategy=1:b_sensitivity=10:mbd=2:trell:v4mv:last_pred=3:dia=-1:mbcmp=2
with vbitrate=512: bitrate  511.034 , PSNR 36.15 in 49.523s
with vbitrate=640: bitrate  640.731 , PSNR 37.52 in 51.028s
bitrate for 36.4 PSNR would be 532.566981604

> > but setting any other comparison function to 2 only makes things
> > worse. 
> 
> I hope you aren't basing your decisions just on PSNR.
Yes I am. I know that PSNR numbers are not a perfect way of
measuring visual quality, but I don't know of any other metric.

-- 
Software is like sex: it is better when it is free. --Linus Torvalds