Are blind tests bogus? Examples of blind tests with positive results. - Page 2

Quote:

Originally Posted by BlackbeardBen 

The infamous Carver Challenge is worth listing here, although it's a bit different in premise from the typical blind test:

http://www.stereophile.com/features/the_carver_challenge/

Priceless...

Thanks for the thread.

I don't think blind tests involve trickery or deception very often, but they do encounter statistical error or come to some weird conclusions.

I think when looking for subtle differences clearly you need to know what to look for, sometimes only 1 in 10 people or less are capable of scoring a positive result, yet they are lost in the pool of data and it's overlooked as chance. 

Looking for striking differences (like speakers) is easier and there are less factors to consider, right?  Also, speakers are at the very end of the chain, which helps.  With other components we have to listen to them with speakers, so the speaker or IEM used can influence the result.  To me that's very clear from my experience with IEM's, that some are more transparent than others, some - even expensive models - will tend to sound 'the same' irrespective to source, which is a positive or negative depending on how you look at it.

I think the rapid-switching time aligned ABX technique may possibly have some errors too, in how the human mind perceives that switching, so I'd prefer to see more tests with a 6 second pause between excerpts, or non time aligned, etc.

The listener should notice which technique they're more sensitive to and be able to choose, if the tests forces a time aligned 0.1 second or less switch then it's imposing unnatural conditions on the listener.  I believe that volume and FR changes are much easier to pick up in 0.1 second switching but that doesn't mean all sounds are across the board, considering how the mind works.

Just as an example, if you can identify someones voice on the phone, would it be easier to identify a very similar voice with ABX, like their twin?  Or do you think, if you listened hours apart, your sonic memory would pick up the very subtle differences in their voices equally well either way?

How does it even matter ?

IF there is in fact a difference 'like night and day, a veil fell from my ears ..etc etc'

they should hear the difference anyway, shouldn't they ?

And if there ISN'T a difference, they should also hear that !

It is no more deceptive than how our hearing works .

You could do a AB-test with two identical wires but tell the subjects that 'A' is a coat-hanger and a portion of them will magically start to hear the differences !

Also, speakers/transducers have no place in this discussion, no objectivist ever claimed they sound the same, clearly they don't .

Incredibly, this is reflected in measurements of speakers ! So, I guess you CAN measure speakers reliably, but electronic devices ?

No, because they posses 'the magical synergy-factor' that you can't measure .

Regarding the ABX  documentation linked in the OP http://home.provide.net/~djcarlst/abx_pwr.htm

Yes, people CAN reliably distinguish between a 10 watt Valve-amp and a 400 watt transistor-amp !

Therefore we can conclude that all gear sounds different and blindtests showing otherwise are faulty ?

Quote:

Originally Posted by kiteki 

I think the rapid-switching time aligned ABX technique may possibly have some errors too, in how the human mind perceives that switching, so I'd prefer to see more tests with a 6 second pause between excerpts, or non time aligned, etc.

You can go ahead and use slow switching if you want but it will just give you even more more fails.  The reason rapid switching is preferred is because it gives you the very best chance of discerning differences.  For up to about 200ms humans retain a very precise echo-like auditory memory that stores most of the information your ears actually heard and is far less susceptible to conscious shifting of attention.  Anything after that is heavily influenced by conscious shifting of attention to different parts of the sound as well as unconscious biases and expectation.

This picture form Audioskeptic's blog sums it up rather nicely.

The human brain cannot store or or even process all the data it receives from the senses.  Everything you consciously see or hear has already undergone what amounts to lossy compression in order to reduce the torrent of raw data to something your conscious mind can handle.  When you change your focus and concentrate of another aspect of a sight or sound you're essentially messing with the compression settings and changing which data is discarded.  If you listen to one part of the sound you can't listen to the others.