[MPlayer-users] variable bitrate, fixed quality

Hans Meine hans_meine at gmx.net
Tue Jul 8 00:26:59 CEST 2003

Nikolaus Rath <Nikolaus at rath.org> writes:
> >> I want to encode a video with a completely variable bitrate but with a
> >> fixed quality. I don't want to use a fixed quantizer too, because this
> > 
> > The bitrate is what you are looking for. Using 2-pass encoding will try to 
> > achieve constant quality. If you don't want to constrain the quality, specify 
> > an enormous bitrate (recall that bitrates above some value are not kbits 
> > anymore, but bits, so add some more zeroes) and lavc will do as good as it 
> > can.
> This doesn't work. I tried it and lavc compressed nearly all frames
> with a quantizer of 3, even if most of them would be perfect with 6. I
> have the impression that lavc wants to reach the given bitrate at all
> costs. If this isn't the case, I need an option to adjust the
> quantizer selection (=the quality), because lavc uses to low
> quantizers for simple frames.

Hmm, maybe you overestimate what a computer can theoretically do in
this context. Since it has no intelligence (and that won't change in
the next decades), it can not really know what you find "good quality"
or "too low quantizer". All it can do is take the bitrate you give it,
and try to distribute it between all frames.

If it takes too low quantizers, I'd try setting a lower bitrate?! If
that increases the quantizers at the wrong places, there are the
following possibilities:
- the configurable limits for the quantizers are too narrow (fixable
  by you)
- mencoder does not vary the bitrate as much as it should (maybe
  fixable by some pro(s) in months of work)
- or it can not tell good quality from bad quality because it's a
  computer program after all (and there's no hope for you in that

I find your usage of the term "the quality" suspicious here, since
there are many different definitions of "quality" a computer
understands, but none of them matches our perception (*)
unfortunately. :-(

Ciao,  /  /
     /  / ANS

*: even if there are good tries, but those are too complicated to be
   put into ffmpeg's codec I fear..

More information about the MPlayer-users mailing list