Double Blind Listening Tests

[Ray Garrison]

Reading the thread on interconnects prompted me to post this. I was going to just chime in on that thread, but thought starting a new one might attract more responses.

My position. Our senses are interconnected. It is easier to distinguish subtle difference in taste when you can smell something than it is when your nose is plugged. It is easier to distinguish subtle differences in touch when you can see what you're touching that it is when you're blindfolded. Why is it such a stretch to say that you can detect subtle, but real, differences in sound between two things when you can see them, but you can't when you don't know which is which?

For example, take coke and pepsi. (Yes, I know there are chemical differences that are identifyable via analysis, but that's not the point.) Most people have a preference. If you sit someone down, and give them two bottles, one coke and one pepsi, and say Which to YOU prefer? most people will pick one or the other. However (and if you've never tried this, you should) if you pour a bunch of identical glasses, some coke, some pepsi, and ask people to distinguish which is which when there's no identifiable marking indicating such, people quickly become confused and can't score higher than random chance when trying to pick which is which (or rather, which they like better). I did this for a science project in high school with 400+ subjects, almost every one of which insisted that they had a significant preference for one or the other. Does that mean that they really taste the same, and you can't tell which is which under blind conditions?

I don't think so. I think it means they taste different, but you can't identify the differences under blind conditions.

Using the feedback provided by being able to see which is which allows the olfactory / taste organs to pick up subtle differences and clues that you simply can't sense without the reinforcement provided by seeing what you're tasting. IMHO.

So, I guess this means that if you drank cola from unbranded, identical containers it wouldn't matter whether they were bottled by coke or pepsi. But you don't. You know what you're drinking. And that knowing allows you to detect differences that you could not detect did you not know. And as those differences are both detectable and real, you wind up with a preference.

I fully believe the same sort of thing is going on with audio, whether we're talking amps, or preamps, or cables, or interconnects, or whatever. If you can't see or identify two or more different components when comparing them, the differences must be quite large to be able to hear them. Different speakers, for example. But just because you can't identify two different interconnects under double blink listening conditions doesn't mean you can't identify differences given the visual feedback of what you're listening to.

Okay, I have my Nomex shorts on...

[Daddy Dee]

Man, do I ever with I would have thought of this when my daughter's were looking for school science projects. Interesting. No better than random? That is surprising. One would think they could tell the difference between coke and pepsi in blind taste tests.

This is also an interesting point about the relationship of senses and preferences. Good food for thought.

You might want to keep your nomex shorts handy.

[ben.]

I can see the difference between Pepsi and Diet Pepsi. What do I win?

[Guy Landau]

----------------

On 3/29/2004 4:37:04 PM bclarke421 wrote:

I can

see

the difference between Pepsi and Diet Pepsi. What do I win?

----------------

A bottle of Coke.

[Juba310]

I guess thats why that "take the pepsi challenge" thing no longer is running...

[rtaylor]

My position. Our senses are interconnected. It is easier to distinguish subtle difference in taste when you can smell something than it is when your nose is plugged. It is easier to distinguish subtle differences in touch when you can see what you're touching that it is when you're blindfolded. Why is it such a stretch to say that you can detect subtle, but real, differences in sound between two things when you can see them, but you can't when you don't know which is which?

For example, take coke and pepsi. (Yes, I know there are chemical differences that are identifyable via analysis, but that's not the point.) Most people have a preference. If you sit someone down, and give them two bottles, one coke and one pepsi, and say Which to YOU prefer? most people will pick one or the other. However (and if you've never tried this, you should) if you pour a bunch of identical glasses, some coke, some pepsi, and ask people to distinguish which is which when there's no identifiable marking indicating such, people quickly become confused and can't score higher than random chance when trying to pick which is which (or rather, which they like better). I did this for a science project in high school with 400+ subjects, almost every one of which insisted that they had a significant preference for one or the other. Does that mean that they really taste the same, and you can't tell which is which under blind conditions?

I don't think so. I think it means they taste different, but you can't identify the differences under blind conditions.

Using the feedback provided by being able to see which is which allows the olfactory / taste organs to pick up subtle differences and clues that you simply can't sense without the reinforcement provided by seeing what you're tasting. IMHO.

So, I guess this means that if you drank cola from unbranded, identical containers it wouldn't matter whether they were bottled by coke or pepsi. But you don't. You know what you're drinking. And that knowing allows you to detect differences that you could not detect did you not know. And as those differences are both detectable and real, you wind up with a preference.

I fully believe the same sort of thing is going on with audio, whether we're talking amps, or preamps, or cables, or interconnects, or whatever. If you can't see or identify two or more different components when comparing them, the differences must be quite large to be able to hear them. Different speakers, for example. But just because you can't identify two different interconnects under double blink listening conditions doesn't mean you can't identify differences given the visual feedback of what you're listening to.

Okay, I have my Nomex shorts on...

----------------

I did a 3 way test with my wife and next door neighbor lady who didn't believe I could tell the diff between jif, kroger brand and skippy peanut butter while blindfolded. Nailed all three thank you. Pass the blueberrys, thanks, Randy

[TBrennan]

Nevermind, I ain't a'goin' there.

[jdm56]

I've participated in smaller versions of that experiment and failed too, BUT...I love Coke and loathe Pepsi. So what gives? I think unless you rinse your mouth out with water in between tastes, maybe after having a cracker, your sense of taste will get confused by the intermingling of the two sodas.

With audio, I think preconceived notions and prejudices will influence your opinion. -wire is wire-

[Clipped and Shorn]

whatever you drink for a few days, that will become your preference. it is conditioning. I even tried this with generic and sure enough after a few days pepsi and coke paled by comparison to generic "cola".....

try this and see......

k6

[Allan Songer]

Ah, that's a buncha bull!

If I couldn't tell the difference between Booker's and Ten High I'd be a richer man . . . .

[TBrennan]

Allan---Bookers and Ten High; what are they?

[Allan Songer]

Ten High = Rot Gut bourbon in a 1.75 liter bottle.

Bookers = The best whisky made in North America.

[kuisis]

Allan

What do you think of Blanton's I know it's not 129 proof, but it is mighty tasty. I also know I could tell the difference between that and four queens!

[Allan Songer]

Blanton's is real good whisky!

My back-up brand will always be Wild Turkey 101. Or maybe Ezra Brooks? Perhaps Maker's Mark?

All I know is that Bookers is the king.

[TBrennan]

Anybody ever lose it trying to drink a shot of whiskey with a lip full of snoose? Man, ya had to be careful! Thank God I gave-up snoose.

When Davy Crockett was electioneering for Congress he's go around with a jar of whiskey and a big twist of tobacco.

He'd offer a potential voter a drink and the fella would discard his quid to take the drink. Then Davy'd give the fella a new chew.

"That way I always left a fella in better shape than I found him" said Davy.

[Frank Speaker]

Ray Garrison wrote:

But just because you can't identify two different interconnects under double blink listening conditions doesn't mean you can't identify differences given the visual feedback of what you're listening to.

Here is a suggestion for testing this theory:

Take the interconnects form the above double blind test and see if people can hear a difference between brand Super interconnect and brand Ordinary interconnect. Show them the two interconnects and do an AB test but tell them when hearing Super that its Ordinary and when hearing Ordinary that its Super.

If the results show that people prefer brand Super then it shows that the visual feedback doesnt help identify what they are hearing.

[bkrop]

In double blind tests, neither the test taker or person administering the test knows which sample is what. It is simply A or B, another person knows what A and B really are and validates the results.

[Frank Speaker]

Im not suggesting a double blind test. Ray is interested in whether seeing what is being tested will help identify differences. Im suggesting a way to test that.

[D-MAN]

Ray, point taken.

Here's some other food for thought...repeated listening will cause one to loose the ability to retain short-term memory concerning the previous A/B demonstration.

Repeated listening and/or A/B-ing will have an effect of confusion in that you will loose your "first choice", which in most cases, will be the correct one. So avoid repeated listening. Take frequent breaks.

You cannot repeat the testing back-to-back, as the short term memory of the experience is "forgotten" or becomes confused. Since humans tend to remember the emotional experience of the listening experience in combination with the actual sounds heard, other emotions/sounds will "overwrite" the previous short-term memories of the previous one so objectivity is degraded. This is why the Klipsch demonstration was "useless" except to prove this fact...

1) all A/B interconnect testing should be done on a system that you are intimately familiar with. Otherwise the differences could be from frequency propagation issues in the environment, or equipment performance here-to-fore unrecognised. Too many variables will confuse the test. The source material should be extremely familiar.

2) Since our ears belong exclusively to ourselves, and may feature different abilities and/or deficiencies, then all judgments are relative. There are no wrong choices.

3) the things to look for are the presence of information previously unreproduced and/or previously unrecognised...if the test gear is upto the test. Typically this will be revealed as an expected aspect of the recording being presented to the ear in a different way so it exceeds the expectation. Certainly we have all experienced the "I never heard that before and I've listened to this 1000 times", etc.

One will never truly be able to observe an event without rendering it into a completely personal interpretation.

So in the case of perception, opinions are ALL WE GOT...

DM

[Dflip]

Blanton's is real good whisky!

My back-up brand will always be Wild Turkey 101. Or maybe Ezra Brooks? Perhaps Maker's Mark?

All I know is that Bookers is the king.

-----------------

North of the 49th whiskey is Wisers Finest or Oldest served with ginger ale or coke, or an imported Single Malt from Scotland. No corn up here, other than what's in a can of niblets.

If you want Canadian made rot gut, Seagram's Five Star. Even better would be a bottle of Newfie Screech. You could also consider anything pretending to be rye that is made in the US, has been aged for three months and sold in a jug.