Mike:

Another example of ABC (With hidden reference ) is attached below. In this
test Bang & Olufsen's spin-off amplifier company "ICE Power" used this
method to determine whether there was an audible difference with a filter
removed from their class D amplifier. They found that listeners could not
tell the difference.

The fact is that very few audio companies do controlled double-blind tests
period... ABX or any other method for that matter. Harman is one of the
very few companies that do valid controlled listening tests, which is the
reason I work here.

Proper valid listening tests cost lots of time and money, money that most
companies feel is better spent on marketing and advertising -- or profit
margin. To a large extent they can get away with it because there are few
checks and balances. For example:

1) 99% of audio equipment review magazines don't do valid controlled
listening tests and their objective measurements of the equipment is often
missing or is incomplete.

2) the consumer is increasingly less able to do a valid comparison of
products in the stores, as audio equipment has become a big-box store
commodity

3) The standard specifications (amplifiers and loudspeakers) given by
manufacturer's are largely misleading indicators of performance and sound
quality.
Making a purchase decision based on these specifications is like rolling the
dice in Las Vegas..


Cheers,
Sean Olive, Manager Subjective Evaluation
R&D Group, Harman International
8500 Balboa Blvd.
Northridge, CA, 91329


If you want the attachments, email me for the .pdf files.