Re: setting up ABX tests at home

Sonnova wrote:
> On Sun, 10 Feb 2008 07:53:33 -0800, Arny Krueger wrote
> (in article <fon6lt01bbe@[EMAIL PROTECTED]
>):
> 
>> "Greg Wormald" <greg.wormald@[EMAIL PROTECTED]
> wrote in message
>> news:folrst02p34@[EMAIL PROTECTED]
>>
<snip>
> 
> This is correct. At audio frequencies, signal levels, and source
impedances, 
> wire is wire. Now, perhaps at microwave frequencies, the size,
composition 
> and routing of cables makes a difference, but from 20 Hz to 20KHz it
matters 
> not one whit unless you are running extremely long runs (like 30 ft or
more) 
> of coax carrying a line-level signal, then one might experience some
small 
> amount of high-frequency roll-off, but this is not what we're talking
about, 
> here. 
> 
> Try this. Take a 1-meter length of Radio Shack coax, solder some cheap 
> tin-plated RCA plugs on each end and use it for one channel of a run
from, 
> say, your CD player to your amplifier (or preamp) and use a 1-meter
length of 
> Nordost Valhalla (at $4000/meter) or any other "high-priced spread" for
the 
> other channel. I DEFY you to hear any difference whatsoever between the
two 
> channels (all else being equal). Cables are snake oil pure and simple.
There 
> is simply no known characteristics of physics, electronics, or music
that 
> would cause any two interconnects to have any sound whatsoever. My rule
of 
> thumb is to buy cables on build quality for reliability and don't spend 
> another penny above what is necessary to get decent, gas tight
connections at 
> the cable/component interface, and adequate strain relief. 

This is a good example of an invalid test.
Or at least a highly flawed test that is so limiting that the results are 
meaningless - and almost always will conform to the expected results.

The amount of acoustical 'cues' available in a mono signal are
significantly 
lower than from a stereo signal. Put another way, it is much much harder
to hear 
differences in mono sources than with stereo sources (assuming a stereo
set up 
that 'properly images'). Therefore a test with left/right channels is
doomed to 
yield results that don't mean very much about anything.

Stereo is a prerequisite for such tests. Not sufficient, but a
prerequisite.

            _-_-bear