[MPlayer-users] RFC: docs update for "how to create a high quality DVD rip"

[ivanova]

On Tue, 08 Jun 2004 19:29:40 -0500
Wayde Milas <wmilas at rarcoa.com> wrote:

> On Tue, 2004-06-08 at 18:18, ivanova wrote:
>
>> While I agree with the general point you are making (I also use
>> Vorbis because it degrades better than mp3 at lower bitrates) I
>> think you are exagerating a bit with regards to mp3's quality. There
>> is little difference between them and then only if you listen
>> _realy_ hard. But don't take my word for it. Proper listening tests
>> have already been done. Go to www.hydrogenaudio.org to find them.  
> 
> Those listening tests are a farce. There is no real mention made on
> how the signal was reproduced, what environment it was listened to in,
> and what hardware was used. Unless everyone listened in a studio
> setting, with studio gear, there is no way a valid comparison can be
> made.
> 
> If everyone stood next to a deisel truck while doing listening tests,
> of course its going to all sound the same.
> 
> If everyone used a cheap $10 pair of powered speakers, its all going
> to sound the same.
> 
> Its a joke. Plain and simple.

If all those tests were performed in a high quality studio setup it
won't be very useful to anyone but people using a high quality studio
setup. I don't think those tests are meant to prove that encoder x will
sound better than encoder y when listening in studio conditions, and if
they are that is clearly not right.

What they do prove, is that the average listener will not notice that
big a difference between encoder x and y. Throw in the fact that we are
talking about movie soundtracks which _generally_ needs less bits and
you are meant to watch a movie while listening. Also the differences
heard by the people doing the listening test are heard after
specifically trying to find differences.

So unless you are watching the movie with your eyes closed, intensely
listing for artifacts on your $10000 speaker system, you probably
won't hear a difference. I say probably because I have to allow for
the fact that there will be cases where some people will hear
differences more easily. But those will remain rare cases. 

If you don't have average speakers and can hear artifacts that make
you want to climb the walls - first make sure that you aren't
imagining it by doing a proper abx test and if you still hear a
difference then you clearly do not fall under the header of average
listener and you have to adjust your encoding preferences accordingly.