Something to read

PolkThug

Somebody has to film this and make a Polkumentary.

jdhdiggs

Originally posted by Polkmaniac

2 - You can test other people, as many as you want...that's fine. But, when I am a perfect 20/20, I don't want to hear anything about being lucky or any of that stuff (I'm saying this now because you've already mentioned it in your post). IF there's still doubt after a 20/20 success, then we'll do more until there is no doubt.

20/20 would be beyond lucky. Testing 6 people and having a one guess 4 or 5 more times could be lucky. 20/20 not happening. Remember, in this test even if you get every question wrong, you win since you did successfully ID the wires. Stats would say that out of 6 people, the range would be 10-13 correct at random chance, so having one at 15 could still be considered random chance. 20/20, no way.

my goal is similar to yours. Prove that at least someone, somewhere can tell the difference.

I would prefer to have the rules set before the test. Because, as you said, if it is a "there is a difference" then someone might claime that one person got "lucky". If there isn't a difference, then I don't want to have some self-proclaimed golden ear saying that our equipment was sufficient.

unc2701

you don't want a coin flip- randomized block designs come out way ahead & guarantee ballance.

I'd go with CDP-> benchmark DAC->tested IC's->Amp->speakers

Only speakers are in room with subject, person switching IC's is in another room and never in view of subject. play the same 30 secs of CD each time, with equal length pause between each playing.

AsSiMiLaTeD

Man, I'm so excited! I'm definitely glad now that we've pushed this thing back to September, because April looks like crap for me.

And I definitely have to get some type of award for this!

jdhdiggs

Originally posted by unc2701
you don't want a coin flip- randomized block designs come out way ahead & guarantee ballance.

I'd go with CDP-> benchmark DAC->tested IC's->Amp->speakers

Only speakers are in room with subject, person switching IC's is in another room and never in view of subject. play the same 30 secs of CD each time, with equal length pause between each playing.

Only issue is the room, that's the problem. I'm sure mark doesn't want to put holes in his walls. You also missed the switchbox. I'd put it between the IC's and amp. CDP to DAC is in the digital domain and will have not interference.

The reason I don't want to have a preconcieved order on the changes is that it is recorded ahead of time and provides oppurtunity for cheating. Even if it comes out 13 A and 7 B, that's fine for this test. Same 30 seconds. Gack! I agree that would be ideal, but get really old for the tester.

AsSiMiLaTeD

Originally posted by unc2701
you don't want a coin flip- randomized block designs come out way ahead & guarantee ballance.

I'd go with CDP-> benchmark DAC->tested IC's->Amp->speakers

Only speakers are in room with subject, person switching IC's is in another room and never in view of subject. play the same 30 secs of CD each time, with equal length pause between each playing.

I think we may be making this too hard. Here's how it should work:

1 - Give me about a minute to listen to each pair of ICs before we start the test
2 - Put me and one other person in the room with the rig - and blindfold me (as long as that other person isn't Mark because he's always talking about taking his clothes off:D )
3 - That person takes a pair of ICs, plugs them in, and starts the music
4 - I tell you which pair of ICs they are
5 - We repeat this process, changing ICs at ramdon until everyone is satisfied that this isn't luck and that, statistically, there isn't a difference.
6 - Just so there's no cheating, the persom can hold up the IC for everyone else to see...so we all know which IC is in the player and there is no doubt.

If all you 'stats' guys wanna do it another way...then fine by me, this way just seems easiest and the most foolproof.

jdhdiggs

I was just thinking the changer would be behind the speakers and a blinder would be between the person being tested and the switchbox

unc2701

I do this for a living (ok, with pharmaceuticals instead of IC's), so here's the stats for 20 tries:

Assuming that he's FOS and can't tell a difference, the probability of him guessing correctly 5 times or less is about 2.1% and the probability of him guessing 15 times or greater is 2.1%. So, by pure dumb luck, he'll prove us wrong 4.2% of the time... this is the usual "statistical significance" cut point for this particular N.

On the other hand, if he can tell the difference 99.9% of the time, the probablilty of him getting ANY wrong out of 20 is 2.0%... So really, we can move our cut points up to, say >=17 and <=3, to cut back on his dumb luck chances, without really infringing on his claim.

jdhdiggs

Alright, sounds good.

6-14- no Difference
4,5 or 15,16 - Inconclusive
1-3 or 17-20 Conclusive difference.

I think we have a ballgame, assuming Doro allows us to pull this off in one of his rooms...

unc2701

I'd also recommend that cables are "switched" or disconnected everytime, even if you have back-to-back randomiztion to use the same IC. Otherwise, he'll know that the cable's haven't been swapped which means you don't have independent trials, which screws everything. Ideally the only person in the room should be the subject... people's reactions could bias it.

Wow, we should write a protocol & publish this, regardless of outcome.

AsSiMiLaTeD

Originally posted by unc2701
I do this for a living (ok, with pharmaceuticals instead of IC's):

Funny...I'm in the PBM industry...work for Caremark as a project manager...

And yes, of course we'll switch the cables on every run...or else I'd 'know' which one I was listening to.

Man, you guys are going overboard with this 'if he gets 17 out of 20 its conclusive' or whatever stuff...you're wasting your time because I'll get all 20. In my mind, if I don't score a perfect 20/20, then there's still some doubt and room for debate...but that won't happen.

I doubt many people even read far enough in this thread to see what we're doing, so I'm going to create a new thread specifically about this test.

James, I'm listing you as the one who'll be setting all this up.

jdhdiggs

Originally posted by unc2701
I'd also recommend that cables are "switched" or disconnected everytime, even if you have back-to-back randomiztion to use the same IC. Otherwise, he'll know that the cable's haven't been swapped which means you don't have independent trials, which screws everything. Ideally the only person in the room should be the subject... people's reactions could bias it.

Wow, we should write a protocol & publish this, regardless of outcome.

No way, that's the point of the passive pre. No processing and not cable changing. Back to backs would be 1 second apart. We aren't leaving the "accoustic memory is short" hole in this test...

jdhdiggs

Originally posted by Polkmaniac
In my mind, if I don't score a perfect 20/20, then there's still some doubt and room for debate...but that won't happen.

I doubt many people even read far enough in this thread to see what we're doing, so I'm going to create a new thread specifically about this test.

James, I'm listing you as the one who'll be setting all this up.

Thanks for the vote of confidence.

As for 20/20, have you smelled Russ's farts? They're so damn strong that they might affect your hearing...

unc2701

Originally posted by jdhdiggs
No way, that's the point of the passive pre. No processing and not cable changing. Back to backs would be 1 second apart. We aren't leaving the "accoustic memory is short" hole in this test...

Ok, that's fine, but we need to always "switch" so if IC1 is on input A and and IC2 is on input B and input C is empty, we always switch to input C before going to the next trial. So if it goes IC1, IC1, IC2, we'd have input A (30 secs), input c (1 sec) input A (30 secs), input C (1 sec), input B (30 secs). This prevents cross contamination.

2+2

Man Polkmanic, you are elevating our expectations....I dont think I can hear distinctions between interconnects unless I can sit with them for a few minutes with familiar music so that I know what to listen for.......I'm rooting for ya.....:D

jdhdiggs

Originally posted by unc2701
Ok, that's fine, but we need to always "switch" so if IC1 is on input A and and IC2 is on input B and input C is empty, we always switch to input C before going to the next trial. So if it goes IC1, IC1, IC2, we'd have input A (30 secs), input c (1 sec) input A (30 secs), input C (1 sec), input B (30 secs). This prevents cross contamination.

That's what I was planning. Always reset so that if you do go back to the same cable, all "clicks and whirs" won't give away the switch.

John K.

Okay, I read the linked article. The points made appear to be essentially correct and simply represent what we've learned(at least those of us willing to learn)from applying basic principles of audio engineering which have been well-known for years and in some cases decades. The use of the term "lies", instead of for example "myths", might be questioned since although sellers and others who know better sometimes deliberately misstate the facts, many sincerely believe these things.

As to "golden ears" who often incessantly repeat the silly mantra of "just trust your ears", the irony is that when they for the first time have to actually trust them alone, i.e. in a blind test where all variables have been closely controlled, their ears of gold seem to undergo a sort of reverse-transmutation and are miraculously turned into tin. The results of such a test using a large number of subjects and samples follow the usual bell-shaped curve, with a few very low or very high results by chance, but averaging close to the 50% result from a random process.

The things that make a difference in audio are the quality of the original recording process, together with the mixing and mastering applied to it, the quality of the speakers and the effects of the listening room acoustics. These are what we should concentrate on, not waste our time and money about things(e.g. players , receiver/amps and wires)which nearly all provide audibly flawless reproduction when operated as designed.

bikezappa

John K.

I agree with everything you so clearly put.

I might add Tuners to the list of items worthy of testing. A great tuna that can play the dynamic range of an uncompressed NPR FM station is a joy. Jazz and classical are best. Old tuners appear to work the best.

It's all about the music.

Music is the best.

Peter

F1nut

Originally posted by John K.

These are what we should concentrate on, not waste our time and money about things(e.g. players , receiver/amps and wires)which nearly all provide audibly flawless reproduction when operated as designed.

Well, that statement proves without a doubt that you don't know a damn thing about audio equipment and that YOU ARE A CRACKPOT TROLL!!!

Come over here any time, bring your POS gear with you and I'll back up every word I've ever said about you.

dragon1952

I'd like to dispute the misconception that everyones hearing is the same. Why should hearing be the same when vision and taste aren't? What's so special about hearing? My wife could hear a song with words a million times but would fail to recognize the same song as an instrumental no matter how many times she heard it. There's hearing and then there's listening comprehension.....two different things. Just like vision. There's seeing and then there's comprehending what you see. My daughter has 20/20 vision yet when she gets done in the kitchen I can find many things she missed like drawers and cupboard doors left partially open, potholders left out, crumbs on the counter, grease on the stove. If I asked her to come back and look around she misses every one. To hear differences in cables it doesn't matter how great your hearing is . It's a matter of listening comprehension. To some people music is just background noise. They couldn't tell the difference between a boom box and a high-end system. That doesn't mean they don't have equal 'hearing'. You could be tone deaf and have great hearing. My wife, as an example again, can't tell a guitar from an organ but she can hear just as well as I can. She absolutely cannot distinguish instruments from one another. So why is it so hard to believe that different people have different listening comprehensions. It has nothing to do with your score on the hearing test.

2+2

dragon, I wholeheartedly agree with you except to add that vision it self varies...not every one has 20/20..some people have better, some worse....same with hearing...regardless of comprehension......and for all the naysayers, just because you cant hear it doesnt mean the sound difference is not there. I told my wife yestday about these cable discussions and she says "how can one not hear the difference?" and that hearing a difference is actually a curse cuz we have to spend more money......;)

dragon1952

Well the point I was trying to make about vision is that you can have 20/20 vision and still not have very good 'vision comprehension'. Same with hearing....yeah you may have 100 people with identical hearing test scores but that doesn't mean they all have the same 'hearing comprehension' or the ability or desire to detect nuances in a musical recording.
Some people listen to a song as a whole. They don't necessarily disect it and listen to various parts. I like to sometimes totally concentrate on a bass line or a background instrument etc, and when you do this alot, especially the same track over and over you can immediately detect a difference a cable might make. If you tend to just listen to the recording as a whole and not focus on one particular thing it is not that easy to detect a subtle difference, generally speaking and IMHO.

Mazeroth

If you guys are going to go all out on this test, I think you should throw another one in for fun. Not to see if you can distinguish between cables, but rather which one you prefer. Maybe have a four wire lineup. Do an A/B and have the listener pick which sounds better. Then do a C/D and let the listener pick. Then do an X/Y with the two winners, and a 1/2 with the two losers. This will give you a 1-2-3-4 order list of which they preferred. Add up all the scores and you'll see which people preferred. Use something like:

Coat hanger
16 gauge wire, cut into 5 pieces, soldered together
14 gauge Home Depot wire
Expensive wire

Maybe just use 2 crappy and 2 good ones, but make sure the crappy speaker wires are REALLY crappy!

You can do the same for interconnects. Use 2 really good ones and 2 crappy ones. Maybe for one of the crappy ones use 50 cent RCA brand 24 gauge and for the other I can whip up a pair utilizing phone cable or something. Should be fun!

steveinaz

Mister, haven't you stirred up enough ****?? LOL!!!

jdhdiggs

All out? Man, some people want to hang us up by our thumbnails to even do this simplistic test and you want to make it more complicated? Hell no...