Re: [transcode-users] Audio extraction issues.

[Jacob Meuser <jakemsr@xxxxxxxxxxx>]

On Fri, Apr 21, 2006 at 07:12:51PM -0300, Rodrigo Severo wrote:
> On 4/21/06, Francesco Romani <fromani@xxxxxxxxx> wrote:
> > On Wed, 19 Apr 2006 19:14:39 -0300
> > "Rodrigo Severo" <rodrigo@xxxxxxxxxxxxxxxxxxx> wrote:
> >
> > [...]
> > > I am trying to rip some clips from a DVD to raw format but the audio
> > > is getting distorted.[...]
> > [...]
> > > The original audio is really high but both xine and mplayer can play
> > > the original source without distortion and ffmpeg can extract this
> > > same material without distortion.
> > > Is there some extra option I should provide or is it a bug in transcode?
> >
> > Maybe option -s (see manpage) can be useful,
> 
> Already tried it. It didn't change the distortion.
> 
> > as well as a preprocessing loop
> > using -J astat.
> 
> I will take a look at it. I don't know option -J astat yet.
> 
> > I can't exclude a problem in transcode itself since it bundles a (old)
> > private version of libac3 that, I fear, is getting outdated.
> 
> I really believe this should be treated a bug as 3 different
> aplications deal with this same audio without problems, only transcode
> distorts it. I haven't tried yet but I bet regular DVD players will
> play this audio well also.

how about some more information then?  all we know is there is some
distortion.  what kind of distortion?  wrong speed?  crackles?
can't understand it at all?

this also seems rather bizarre to me, since audio is almost always
decoded to "raw" and then reencoded.  if this were a general bug
in transcode, then there would be many more problem reports.

as far as I'm concerned, until there is more to go on, this is some
type of user error ... and seeing as "-y raw -F I420" is not even
a valid transcode option combination ...

my guess is that the AVI header is probably busted somehow.  the
audio stream is probably OK, but the header does not match what
the audio data is.  what does tcprobe say about the resulting AVI?


<rant class="notTryingToBeRude_justTheWayItIs>

please, if you really feel this should be treated as a bug, do a 
little basic debugging of your own, OK?  that does not mean to tell
us that some other programs do not exhibit the same problems with your
source material.  it means that you should give us some kind of
diagnostics to prove to us that it is _not_ user error.  why should
we even look into such a crappy problem report if in the end we find out
that our time has been wasted and the problem was indeed user error?

and what's with asking if we want a sample?  just make it available
and tell us where it is, OK?

sorry, but it's really annoying to do stuff for free, and then have
people bitch about what they got for free and not do a damn thing
(not even a clear description of the problem) to help figure out
whatever it is they are bitching about.

</rant>

-- 
<jakemsr@xxxxxxxxxxx>