[LAU] ABX test (was Re: What is the best MP3 encoder?)

[Peder Hedlund]

Quoting Ralf Mardorf <ralf.mardorf at alice-dsl.net>:

> I agree, an experienced tester btw. does know that the tests that can be
> done are limited. You can't do it 8 hours a day, after making a few
> tests, _nobody_ is able to listen in the most possible neutral way
> anymore, after a few tests the listening is biased.

You don't have to do the ABX audio test in an 8 hour non-stop session.  
You can do two test runs, go to sleep, do three more, go to a movie,  
do one more, play some soccer and then do the remaining four. And if  
you have golden ears each run shouldn't have to take more than 20  
seconds or so.

> Some time ago I ask a friend to burn me an original CD, but he didn't,
> he converted from MP3 and burned a CD. I didn't know that. I started
> playing and stopped it immediately, without knowing that it were MP3s.

We don't know the settings used to make your friend's MP3s so it's  
hard to say anything specific about that test case. He might have used  
"-b 64" or something.

As said before; there are MP3 encoders (at least many of the older  
ones) that produce lots of audible artifacts. Recent versions of lame  
happens to be among the ones that produces the least.
If you can truthfully ABX my files then you're perfectly entitled to  
claim MP3s sound bad to you. At least we won't be able to say that  
*you* can't hear the difference.

- Peder