From: Fred Maxwell <rockbox_at_anti-spam.org>
Date: Sat, 20 Dec 2003 18:51:50 -0500

Neon John wrote:

> Any non-double-blind test contains natural bias. This is one of the
> fundamental bedrocks of experimental statistics, an area where I am very
> experienced.

I agree with that statement, but that's different than saying that those
conducting the experiments have a bias. They may have no bias, but their
non-double-blind experiments are flawed.

The entire high-end audio industry has fought against double-blind testing
for years. The "audiophiles" don't want a scientific test of their claims
that they can hear the supposed improvements wrought by their idiotic
investments in unidirectional cables, green paint for the edge of CDs,
bricks to sit on top of amps, $200 power cords, and the like. The snake-oil
salesmen that market these voodoo audio products using ads filled with
pseudo-scientific techno babble certainly doesn't want scientific testing.

If anyone (willing to run Windows) wants to do PC-based ABX (double-blind)
testing of audio, there is a program called PCABX which can be downloaded
from www.pcabx.com. It takes two WAV files and lets you compare them. If
you want to compare two compression methods, convert compressed files back
to WAV files and use that as the source for PCABX.

Regards,
  Fred Maxwell
Received on 2003-12-21