Originally Posted by Shaun B
Sighted bias is probably best modeled as a random variable, so cable cost need not be considered to be an influence with just one possible direction
What do you mean random?
The effects of sighted bias on perceptions of sound quality is random. It may be positive or negative for each component and attribute being compared.
So sighted bias can result in random and non-random variables?
The effects of sighted bias on the perceptions of sound differences in sound quality is more consistent - a sound quality difference will probably be perceived whether there is one or not.
How would you know if it was non-random?
Close to 40 years of experience doing DBTs.
So then everyone is just hearing differences under sighted testing regardless of what you're listening to that has nothing to do with the equipment.
When it comes to things like amplifiers cables and DACs that generally don't sound different, the differences that people perceive have nothing to do with the equipment. The casual nature of the circumstances surrounding the listening evaluation is the source of the differences that are perceived. That is pretty reliable and predictable. Which amp, cable, or DAC is perceived to sound better is
Obviously other tests such as those involving loudspeakers and room acoustics can involve differences that are so large and unmistakable that they may be overcome the biases in sighted testing. However people like Dr. Floyd Toole and Dr. Sean Olive have proven themselves to be correct when they do blind tests involving those things. Blind tests of speakers are similar to blind tests of other kind of audio gear in that they often show that inexpensive equipment is strong sonic competition for very expensive equipment. Two examples of this are these comparisons of inexpensive and pretty expensive speakers,
and this comparison of $400 a pair speakers with a very elegant, highly regarded and complex speaker system running about $20K.
The proof of this is that if you make the tests substantially and effectively less casual, the differences go away. One problem with sighted evaluations is that there is nothing like a built in check sum on the test. In ABX tests there is a kind of a built in check sum. The listener's reactions become random and that is detected in the statistical analysis. It is not a true test unless you can fail.
Sighted listening tests involving digital cables are very instructive because the audible differences that they can possibly cause fits into a narrow category.
Cable problems cause fairly gross problems in a narrow range such as no sound or sound with clicks and pops or loud static and the like.
Digital cables can't possibly cause problems with tone, soundstaging, sonic detail or the like unless the listener misinterprets the clicks, pops or loud static as being problems with tone, soundstaging, sonic detail or the like.