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

[Tobias Diedrich]

Wayde Milas wrote:

> Unless we run the test we will never know :)

Right. :-)

> Also the test was 128k mp3 against 256k mp3. I bet if they did it with
> 128k mp3 agains raw 16 bit 44.1k (raw cd) there would be alot more that
> would hear the difference.

You misread that one.  It was 128k against CD and 256k against CD.

Because of timelimits the took 1 minute of each song.  The testee would
first get to thear the cd version, then 128kbit,256kbit and cd in
unknown order.  That's 68 minutes per person.

If he was able to tell apart 128kbit and cd he would get one point.
If he was able to tell apart 256kbit and cd he would get another two
points, however if he failed to tell apart 128kbit and cd he would get
zero points.

Thus you could get either 0, 1 or 3 points per song (as in the table).


BTW, regarding the psy-model I thought of a better way of putting it
into words:

By using a psychological model the encoder assumes the the chain
between decoder and listener is able to reproduce the decoder output
100% accurately.  This holds much more true for a high-end setup than
for a low-end setup.  So for a low-end setup the psy-model should work
less well and it should be more easy to tell apart uncompressed and
compressed version.

-- 
Tobias						PGP: http://9ac7e0bc.2ya.com
He said he hadn't had a byte in three days.
I had a short, so I split it with him.