[MPlayer-dev-eng] -lameopts: conflicting documentation and code?

[D Richard Felker III]

On Thu, Feb 19, 2004 at 01:37:57AM -0300, Diego Biurrun wrote:
> Charles Wilcox writes:
>  > The man page and HTML documentation say that they q= value can be set from
>  > 0 to 9, the equivalent of the same option of Lame.  However, this is not
>  > the case of what happens.  mencoder.c:986 has a line that adds one onto
>  > the value specified before it is sent off to libmp3lame.  This is
>  > complicated by the fact that cfg-mencoder.h:25 says that the input is
>  > indeed in the 0 to 9 range.  Also, the code makes no effort to check to
>  > see if the resulting value is 10, which is an incorrect value to pass on.
>  > 
>  > Considering all the docs imply this option should match up to what LAME
>  > would normally take, could this possibly be a error in the code?  Taking a
>  > cursory glance at the lame code in frontend/parse.c:1626, it does not alter
>  > the value before it calls lame_set_VBR_q.
>  > 
>  > Sorry, I'm just learning how to use Mencoder, but I've used Lame
>  > extensively, and this inconsistency confuses me.  Is there a reason it it
>  > like this?  If so, there is no warning to those who'd assume that the VBR
>  > quality numbers mean the same thing.  (That, and it just doesn't make any
>  > sense to me.)  If this is a mistake... wow, I guess not many people
>  > specify -lameopts to have caught this.
> 
> Sounds like you found a real bug, could you come up with a patch that
> fixes it, please?

IMO this is a difficult issue. I wanted to fix it a long time ago, but
changing the behavior is not acceptible under mplayer policy afaik...

Rich