Quote Originally Posted by George47 View Post
1) You can minimise these effects by not carrying out ABX tests with 12 sec samples but by listening for much longer time periods and getting the real measure of the item and then trying the other component. Doing blind tests is an awful way to test components and I have not seen any statistically significant data that shows SS amp sound different. But they are.
Indeed, although it's only "awful" in the sense of considering the results obtained as conclusive, which they aren't, simply because like every other test that involves the human brain and hearing, it's fallible! The likelihood is, whenever you try and *force* yourself to hear a difference, because you're under pressure to do so, in the way of ABX testing or similar, you run the risk of obtaining a false result, as most people don't respond well to being tested in that way.

Subtle differences in sound between components, which generally is what you're testing for, if they exist, are more likely to be revealed naturally, over time, by listening to the component(s) concerned with a variety of familiar recordings, in a relaxed environment, than via any 'quick fire' ABX testing.

Therefore, by all means undertake such tests, with whatever components you like, and use it as ONE valid means of determining differences, but the best way of obtaining meaningful results, is as you've said, by listening for much longer time periods and getting the real measure of the item, and then trying the other component.

Of course, like ABX testing, that process is also fallible, for a variety of obvious reasons (NO test is perfect), but at least you've applied some reasonable judgement criteria, in order to determine the existence of any differences. Therefore, I'd contend that some form of objective *and* subjective testing (as outlined) is required, to get a proper handle on what's happening, not just 10 minutes spent getting jiggy with a blindfold!

Marco.