Rationality, objectivity, the scientific method, ABX, and dogmatism
Dec 12, 2004 at 11:57 PM Post #106 of 254
Quote:

Originally Posted by taymat
The biggest difference to sound tone in a hi-fi is room reverb/size and speakers.


Of course, *any* added reverberation will be a distortion of the recording. It's just a matter of how much and is it pleasing, etc.


JF
 
Dec 13, 2004 at 12:46 AM Post #107 of 254
Seems this issue is often brought up. No sweat, sometimes reading through these threads provides more insight when some well-written remarks turn up.

I can understand the sceptics and moreover, have been one in the past and still am one for some things. Yes there are plenty of items that seem to be snake-oil to me (those magic rocks, pixie dust, shiny stickers) but some other things are not.

Much of what markl wrote (as well as MikeS) I agree with. I'm a big proponent of garbage in garbage out. I think detecting nuances (subtle or not so subtle) as well as large improvements often come with a refined ear. This is not the same as some with good hearing.

Take for example wine. Some wine is ludicrously expensive to this poor college student. I mean....50k USD for 750 ml of wine bottled 100 years + ago? Err....no thanks. But to others this is an hour's pay. To the rich, some may buy because they want to appear "refined" while others can truly taste the diffence between an award winning Lafitte red and a boxed red wine. Obviously that difference was purposely dramatic.

I have come to enjoy the difference between a Merlot and a Cabernet Sauvignon (my prefered grapes). Some years are better than others and some companies consistantly pump out better batches. Can my untrained palette detect the difference between a merlot and a cabernet? (or between copper and silver wiring...) why yes it can! Distinct differences in body, nose, texture etc. (similarly for the wire). Can I succesfully detect the differences between a 1000 dollar Merlot and 50k Merlot and a 100 Merlot? Maybe, but I bet it would be slight (I have tasted a very expensive Lafitte and it was awesome but I could have been fooled given the great foods I was enjoying the wine with, all matched perfectly and purposely). However, a true wine connoisseur could in spades describe differences, subtleties, aromas, deviations etc, that I am simply not equipped to supply. Is it because I am inherently incapable? No...(though maybe) but certainly this is because I lack experience.

Experience counts for a lot, the more equipement I listen to the better I am at describing sound and what I like in a musical playback system. Also, the longer one has with asystem the more then get to know the sound (this is assuming a "marklesque" listening, not the background listener or specifically the multitasker). From here it is much easier to pick out differences. I was once told that HD600's sound like HP-1's. That this person could not detect a single difference in the sound. He is young, can describe what he likes and describes it well, but for the life of me I can't understand how he can't tell the difference! Same thing has happened between RS-1's and HP-1's with a bunch of my friends, all of whom have been recently checked for their hearing and all scored well (they are all under 30 for those wondering). They simply lack the exposure. They sat at my system were wowed by the clarity, the resolution but otherwise they were like...wow yeah, they sound the same but cool! Such wonderful music.

Meanwhile, I am mystified. The same? Are they deaf? Not according to science, to their audiologists! Experience and exposure to a great variety of gear will benefit a listener well.

Now will there be a huge difference between 15k copper wire and 100 copper wire? Perhaps given the configuration, insulation etc, but worth the markup? Not to me...but to someone else perhaps. I think once one gets into the wire purity and the configuration of the wire + proper connection to the connectors...well then things should be equal. Thus...the ludicrous markup in cables is yes...oily.

That said, those that have never heard a lamp cord wire next to a well designed cable can't really comment, or better, if they can't hear the difference, perhaps they truly do need to experience the system in a more thorough manner. Seriously, it is not as if all those in audio are ridiculously over indulgent with their money. Moreover, there are PLENTY of folks who buy into the quality cable deal because there is a difference...not because they have more money than brains.

Finally, those who can't hear a difference between a pcdp and a quality player, have simply never heard a quality player. It's that simple. Changes in snap, palpable playback, resolution, coherency, bass, highs, its there and its real. For those that don't believe my doors are open to you. Come on by and we will DBT my cd player with my portable one.
 
Dec 13, 2004 at 1:22 AM Post #108 of 254
I like to read these discussions, but they can be h-e-double-toothpicks when you are in the middle of them!

I just took some measurements. I'll try to stick to facts without any judgments or inferences.

The Senn HD580 stock cable resistance is .7 ohms.

To put this in perspective, I measured the impedence of the HD580 headphones -- 296.5 ohms (this is quite close to the published specification of 300 ohms) for each channel, .7 ohms of which would be attributable to the stock cable.

I have two Sony detachable headphone cables, one that is supposed to be high performance and one that just comes with some cheaper phones. I was skeptical as to whether the two would measure differently.

I was surprised -- the cable for the cheaper Sony phones measured 1.3 ohms, while the high-performance Sony cable measured .8 ohms. They are the same length.

The Senn HD580 stock cable beats the Sony high-performance cable in resistance by .1 ohms. However, the Sony cable is about 14 inches longer than the Senn cable. Both cables are pretty long (in the neighborhood of 10 feet long, give or take).
 
Dec 13, 2004 at 1:59 AM Post #109 of 254
Quote:

Originally Posted by taymat
BTW in my experience 90% of cables once you get past stock bell-wire/interconnects are snake oil.


I assume what you mean is that the cables are "snake oil" to you, i.e., they did not make any audible difference in your listening and, therefore, it would have been foolish for you to spend your money on such cables. Just because YOU don't hear a difference, however, does not mean others don't hear a difference. And if they do, then the products are not "snake oil" at all.
 
Dec 13, 2004 at 3:12 AM Post #110 of 254
Quote:

Originally Posted by Steve999
The Senn HD580 stock cable resistance is .7 ohms.


Hi Steve,

It seems like you have a meter with pretty good resolution. Resistance readings less than 2 ohms are hard to make accurately. The best way is to use a four-wire ohm meter. Maybe that is what you used. If not, I expect the reading (which is already low) to be even smaller. A four-wire ohm meter eliminates meter lead and lead to connector resistance... Regardless, your measurements are well within the ballpark and will be accurate (either way) for comparative purposes...

And if you wish to make some "judgments or inferences", I think it's safe now... (don't worry if there happens to be a "beep"...)


JF
 
Dec 13, 2004 at 6:43 AM Post #112 of 254
Quote:

Originally Posted by Ross
...
A test which determines whether someone is able to distinguish two slightly different sounds after a short period of time in artificial listening conditions proves nothing other than that the person can (or cannot) detect a difference between two slightly different sounds in a short period of time in artificial listening conditions.



Actually most ABX tests will let the listeners choose the length and number of trials. It doesn't always have to be "artificial listening conditions," other than the fact the visual cues of different equipments are being removed. The reason for short periods of testing time is to accommodate the short sonic memory of human beings which if I remember corrently is pretty good for 2~3 seconds and deteriorate rapidly after. However the listerner doesn't have to hastily jump back and forth between tracks if he/she doesn't want to take that route.

Quote:

Originally Posted by Ross
What the **** does this have to do with music or hi fi?


To see if the improvements heard in hi-fi is something substantial created by the supposedly better equipments or just placebo.
 
Dec 13, 2004 at 7:17 AM Post #113 of 254
Quote:

Originally Posted by swalker
Actually most ABX tests will let the listeners choose the length and number of trials.


The tests that people frequently point to all seem to have short listening times, and my recollection was that the listening times were fixed, as there are typically several listeners performing the test at one time. Can you point to a test where listeners were allowed to choose how long they listened to a particular track before switching? I'd like to read more about those tests.

Quote:

Originally Posted by swalker
It doesn't always have to be "artificial listening conditions," other than the fact the visual cues of different equipments are being removed.


I'd also be interested in any tests that involve the participant's own equipment. Many of the tests that are often referenced seem to involve (1) not the participant's own system, but a system chosen for several participants to listen to, and (2) a series of musical selections that the participants may not be completely familiar with. To me, these are examples of "artificial listening conditions," and I'd be interested in reading about some tests where they addressed these issues.
 
Dec 13, 2004 at 7:26 AM Post #114 of 254
Quote:

Actually most ABX tests will let the listeners choose the length and number of trials. It doesn't always have to be "artificial listening conditions," other than the fact the visual cues of different equipments are being removed.


Unless you're secretly camping out in someone's living room and switching cables while they are not looking, I can't see how these tests can be anything other than artificial.

Quote:

The reason for short periods of testing time is to accommodate the short sonic memory of human beings which if I remember corrently is pretty good for 2~3 seconds and deteriorate rapidly after. However the listerner doesn't have to hastily jump back and forth between tracks if he/she doesn't want to take that route.


While differences are obvious within a short space of time, the evaluation of those differences (conscious or unconscious) can take weeks or months. I don't believe that any experiment which switched cables at frequencies less than 1 month would be valid, because it does not represent real world listening conditions. Tests involving seconds or minutes or even hours of listening will tell you nothing of value. No one listens to their hi fi system to determine if A is different to B, but they do know over time whether they are enjoying A more than B.

Quote:

To see if the improvements heard in hi-fi is something substantial created by the supposedly better equipments or just placebo.


But that is exactly what you are not testing by the experiment you have just described. You are testing only a single criterion - the ability to distinguish A from B. This says nothing about "improvements" or whether something is "better". If you were trying to measure whether something was an improvement (rather than just a difference), you could not do this using an ABX experiment unless it measured completely different things - and mostly subjective things - rather than whether subject 1 can tell the difference between A and B.

I'm sure we have all had experiences where the most significant and obvious changes - eg a new pair of speakers - produce unmistakeable and immediately obvious superficial differences, but make little difference to our enjoyment of music. But sometimes the reverse is true: it is the subtle, and sometimes barely perceptible differences, which over time determine - often at a sub-conscious level - whether we continue to enjoy listening to music.

All of which is to say that your crude ABX experiments tell you nothing of real value to hi fi or music, which require far more subtle tools than the jackhammer experiments to which you want to subject them.
 
Dec 13, 2004 at 1:47 PM Post #115 of 254
Quote:

Originally Posted by PhilS
I'd also be interested in any tests that involve the participant's own equipment.


Headphone Cable Listening Test (take 2)

1) Ask a friend for assistance. Ready two pads of paper, something to write with, and a coin.

2) Position your comfortable chair in front of your audio setup, facing the center of the room. Allow room behind for your friend. Find a second comfortable chair for friend.

3) Remove the headphone cables from your headphones and amp.

4) Select whatever music you wish to listen to and get comfortable in the chair. (Don't peek.)

5) [Friend's task] Flip a coin (heads = A = aftermarket cable; tails = B = stock cable). Secretly, mark down the result. Connect the cable. Position headphone on listener. Start music. (Ask to turn up or down, change tracks, anything you require...)

6) Listen for as little or as long as you wish. When you are comfortable with a choice, mark down either A or B.

7) [Friend's task] Remove headphones. Stop the music. Remove cable from headphones and amp.

8) Go to step 5. Repeat steps 5-7 for a total of 25 times.

9) Compare results.


I'm not a statistician, but think that 25 times is statistically valid. And believe that you will need to 17 (or more) correct choices for a positive result. I also think it's important to remove the cable between each listening so that the actions are the same.


JF
 
Dec 13, 2004 at 1:57 PM Post #116 of 254
Quote:

Originally Posted by JohnFerrier
Headphone Cable Listening Test (take 2)

1) Ask a friend for assistance. Ready two pads of paper, something to write with, and a coin.

2) Position your comfortable chair in front of your audio setup, facing the center of the room. Allow room behind for your friend. Find a second comfortable chair for friend.

3) Remove the headphone cables from your headphones and amp.

4) Select whatever music you wish to listen to and get comfortable in the chair. (Don't peek.)

5) [Friend's task] Flip a coin (heads = A = aftermarket cable; tails = B = stock cable). Secretly, mark down the result. Connect the cable. Position headphone on listener. Start music. (Ask to turn up or down, change tracks, anything you require...)

6) Listen for as little or as long as you wish. When you are comfortable with a choice, mark down either A or B.

7) [Friend's task] Remove headphones. Stop the music. Remove cable from headphones and amp.

8) Go to step 5. Repeat steps 5-7 for a total of 25 times.

9) Compare results.


I'm not a statistician, but think that 25 times is statistically valid. And believe that you will need to 17 (or more) correct choices for a positive result. I also think it's important to remove the cable between each listening so that the actions are the same.


JF



Unfortunately your method won't convince the hardcore objectivist.
It isn't doubleblind, that means the friend knows what is what, and therefor he could influence you on a subconscious level.
An even greater fault is the neglection of the usually different weight of the cables.
 
Dec 13, 2004 at 2:08 PM Post #117 of 254
Quote:

Originally Posted by cosmopragma
Unfortunately your method won't convince the hardcore objectivist.
It isn't doubleblind, that means the friend knows what is what, and therefor he could influence you on a subconscious level.
An even greater fault is the neglection of the usually different weight of the cables.



Yep. Hardcore objectivist are a tough bunch, but I'd be satisfied. I think subconscious messages are buried in noise. Weight may be a small issue. I can't compare cables, but trust someone like Phil on this. If a person finds the need to move your head around to judge the weight of the cable, then I hope the person would realize that this isn't fair... (And my experience with listening tests is that my mind does start racing for clues to which is which...)


JF
 
Dec 13, 2004 at 5:52 PM Post #118 of 254
Quote:

Originally Posted by JohnFerrier
Headphone Cable Listening Test (take 2)

1) Ask a friend for assistance. Ready two pads of paper, something to write with, and a coin.

2) Position your comfortable chair in front of your audio setup, facing the center of the room. Allow room behind for your friend. Find a second comfortable chair for friend.

3) Remove the headphone cables from your headphones and amp.

4) Select whatever music you wish to listen to and get comfortable in the chair. (Don't peek.)

5) [Friend's task] Flip a coin (heads = A = aftermarket cable; tails = B = stock cable). Secretly, mark down the result. Connect the cable. Position headphone on listener. Start music. (Ask to turn up or down, change tracks, anything you require...)

6) Listen for as little or as long as you wish. When you are comfortable with a choice, mark down either A or B.

7) [Friend's task] Remove headphones. Stop the music. Remove cable from headphones and amp.

8) Go to step 5. Repeat steps 5-7 for a total of 25 times.

9) Compare results.


I'm not a statistician, but think that 25 times is statistically valid. And believe that you will need to 17 (or more) correct choices for a positive result. I also think it's important to remove the cable between each listening so that the actions are the same.


JF



To perform this test under conditions that I think are meaningful and "real world" would take approximately six months, assuming one week's time to listen to each cable.

P.S. Also, I woudn't mark down A or B. I don't care about being able to identify cables. I care which one sounds better. Thus, I would mark that and at the end of the six months, it could be determined whether my selection of what cable sounded better correlated with A or B a certain percentage of the time.
 
Dec 13, 2004 at 6:05 PM Post #119 of 254
i tried this little test with my friend on saturday and it turns out i prefer my AudioQuest Diamondbacks over his Monster Cables after all. Take that!
 
Dec 13, 2004 at 6:15 PM Post #120 of 254
Quote:

Originally Posted by PhilS
To perform this test under conditions that I think are meaningful and "real world" would take approximately six months, assuming one week's time to listen to each cable.

P.S. Also, I woudn't mark down A or B. I don't care about being able to identify cables. I care which one sounds better. Thus, I would mark that and at the end of the six months, it could be determined whether my selection of what cable sounded better correlated with A or B a certain percentage of the time.



So, you mean that the differences are so small, that it's virtually impossible to tell a difference. Of course, I agree. It is impossible.

-

You can't play one song with one cable, the same song with another cable and tell which cable is which?


JF
 

Users who are viewing this thread

Back
Top