JPLAY Responds: An Open Letter

[Phelonious Ponk]

I agree. The way the Tinkerbell effect works is that if you believe in something, you start trying to convert other people to your belief, hang around people who share your belief, only hear and read things that support your belief - and very quickly it becomes a self-reinforcing circle, not just for you but for your whole environment. It is almost like an augmented version of groupthink.

Wow. So America is in an era of Tinkerbell politics?



Tim

 

[Julf]

Wow. So America is in an era of Tinkerbell politics?

My wife is American, so I'd rather not comment on that.

 

[Julf]

The problem was, listening blind, I heard the difference as often, statistically speaking, when the audiophile player was not on.

But that is just because your system isn't resolving enough. Or your ears aren't trained enough. Anyway, bind listening is flawed.

 

[Phelonious Ponk]

But that is just because your system isn't resolving enough. Or your ears aren't trained enough. Anyway, bind listening is flawed.

Correct. Knowing when you're listening to what you most desperately want to believe in is much more objective.

Tim

 

[audioguy]

\Anyway, bind listening is flawed.

In some cases I might agree but in this case, it is the ONLY way to determine (objectively) if JPLAY is better than, the same as or worse than JRIVER. I know others might disagree but if on the highest resolving system, one can consistently select one over the other, then there must be a real difference. Set up two identical servers, one running each product with two different input to a DAC and switch inputs. Pretty straightforward.

But I am not naive enough to believe those who claim differences would ever agree to such a test. It is much easier to just say "blind tests are flawed".

 

[Phelonious Ponk]

Blind tests are flawed. Even when they are very carefully designed and executed and repeated enough with enough subjects and subject matter that they are nearly impossible to deny (though the faithful persist in denying even then), they are imperfect.

But they're still better than sighted tests.

Tim

 

[Groucho]

In some cases I might agree but in this case, it is the ONLY way to determine (objectively) if JPLAY is better than, the same as or worse than JRIVER. I know others might disagree but if on the highest resolving system, one can consistently select one over the other, then there must be a real difference. Set up two identical servers, one running each product with two different input to a DAC and switch inputs. Pretty straightforward.

Is there such a thing as "two identical servers"? Their system clocks will be running at slightly different rates relative to the DAC clock. At power-on they will each load up slightly different system times. Their disk drives will spin up at slightly different rates, and the heads take slightly different times to move across the disk. If connected to a network, they will each access it at different times. After a while they would be doing different things even if running the same software. One output will be delayed compared to the other (maybe by a small amount, but how much is too much?) and there will be a glitch when switching over the DAC. Maybe the proximity of the computers to each other, or their cross-coupling via the mains supply will more than swamp any supposed differences caused by JPlay vs. JRiver. Maybe the only real difference is due to acoustic emissions from the PC's power supply, but this test will be running both at the same time so it will not show up. Not so straightforward, I think.

 

[audioguy]

Is there such a thing as "two identical servers"? Their system clocks will be running at slightly different rates relative to the DAC clock. At power-on they will each load up slightly different system times. Their disk drives will spin up at slightly different rates, and the heads take slightly different times to move across the disk. If connected to a network, they will each access it at different times. After a while they would be doing different things even if running the same software. One output will be delayed compared to the other (maybe by a small amount, but how much is too much?) and there will be a glitch when switching over the DAC. Maybe the proximity of the computers to each other, or their cross-coupling via the mains supply will more than swamp any supposed differences caused by JPlay vs. JRiver. Maybe the only real difference is due to acoustic emissions from the PC's power supply, but this test will be running both at the same time so it will not show up. Not so straightforward, I think.

I will quote Tim:

But they're still better than sighted tests.

 

[opus111]

It is always good to suggest new ideas for things that might affect the sound, but the next step is always verifying if the effect is real.

The debate over these things I've noticed often follows this pattern - objectivists start talking about 'real' effects but never define what's meant by 'real'. So to head this off, when you use the word 'real' do you actually mean the effect is not placebo? (I ask because placebo effect is most certainly real).

If you do mean not placebo I'd like to make a plea that in future this phrase 'real effect' is dropped. It leads to all kinds of misunderstandings because the next short step from this is that once the effect is established as 'not real' then it becomes 'imaginary' and 'fantasy' and then communication just tends to break down.

<edit> I'm a bit slow, but its just dawned on me that perhaps the (unconscious or conscious?) intention is precisely for the dialog to break down and that's why 'real' is being used...

 

[Imperial]

Blind tests are flawed. Even when they are very carefully designed and executed and repeated enough with enough subjects and subject matter that they are nearly impossible to deny (though the faithful persist in denying even then), they are imperfect.

But they're still better than sighted tests.

Tim

Tim. It is highly unusual to see an American use irony in negative connotation, angling it as a double entendre.
Very cunning, to then flaunt the crux as a denotation... calling the worst choice the preferred one, using "that" as a proof that all others are fallacy.

I can not even say : I beg to differ...
I'm gonna have to go with: When the music no longer causes the eyes to close, the mind to wander, the heart to dream...
Then the next purchase is not far off.

I bet manufacturers would love ABX actually. Then one could prove that the upgrade is better...
It baffles me that they dread it, when the ears tell the listeners that the specs, although different don't sound like the (sum of) cash involved...
Specs don't lie?

Jriver, taking the spec route her, but not wanting to do an ABX has dug themselves a very fine grave...
They should not have dug up the axe...

This is my humble theorem.

Imperial.

 

[opus111]

Blind tests are flawed. Even when they are very carefully designed and executed and repeated enough with enough subjects and subject matter that they are nearly impossible to deny (though the faithful persist in denying even then), they are imperfect.

But they're still better than sighted tests.

Rather like discussing whether a punch on the nose is better than having one's ears boxed.

Speaking as a teacher, I know full well that tests are flawed. Continuous assessment knocks tests into a cocked hat. But I think its too much to ask for this insight to be adopted by the audio world

 

[Imperial]

Phelonious Ponk served up a masterpiece in wordplay, I'll give him that. I've never read a better post on this forum before, that much is true!

Im-pressed-perial.

 

[NorthStar]

Are there "proven" scientific measurements between JPlay and JRiver?

---------

* By the way, the original link from the first post has ongoing comments; up to now (June 17, 2013).
=> http://www.audiostream.com/content/jplay-responds-open-letter

 

[Julf]

The debate over these things I've noticed often follows this pattern - objectivists start talking about 'real' effects but never define what's meant by 'real'. So to head this off, when you use the word 'real' do you actually mean the effect is not placebo? (I ask because placebo effect is most certainly real).

If you do mean not placebo I'd like to make a plea that in future this phrase 'real effect' is dropped. It leads to all kinds of misunderstandings because the next short step from this is that once the effect is established as 'not real' then it becomes 'imaginary' and 'fantasy' and then communication just tends to break down.

You definitely have a point - the association from placebo (or perceptual bias etc.) to "imaginary" and "fantasy" is a negative one, making it sound derogatory. Placebo is a real effect, and there is nothing wrong with the person experiencing it.

What would be a good term to describe "effects not reflected in the actual sound waves entering the ear of the listener"?

 

[opus111]

How about 'expectation bias effects' ?

 

[Julf]

How about 'expectation bias effects' ?

Even "bias" tends to have negative connotations. It would be good to have a term that is really neutral, so that people can acknowledge that it is something entirely natural and common, and invoking it is not an insult.

 

[Groucho]

The debate over these things I've noticed often follows this pattern - objectivists start talking about 'real' effects but never define what's meant by 'real'. So to head this off, when you use the word 'real' do you actually mean the effect is not placebo? (I ask because placebo effect is most certainly real).

If you do mean not placebo I'd like to make a plea that in future this phrase 'real effect' is dropped. It leads to all kinds of misunderstandings because the next short step from this is that once the effect is established as 'not real' then it becomes 'imaginary' and 'fantasy' and then communication just tends to break down.

Laudably balanced, tolerant, eminently sensible. But... if you think that placebo is "real" and understandable and that everyone is prone to it, why do people in these forums and elsewhere virtually never qualify their sighted 'tests' with the caveat that they may have experienced a placebo effect? Instead, furrows are repeatedly ploughed, asserting that feedback is bad, digital is harsh, Class A is wonderful etc. because listening tests show it.

I think most objectivists consider "real" to be that which is designed and built by proper engineers, and that all the other fluffy stuff - aesthetics, marketing etc. - to be very lovely, but not central. If all audio systems depend on placebo effects to work, it is hard to see how or why anyone should study audio engineering at all; they should simply buy existing hardware and concentrate on making a very nice box and telling a good story in the brochures. And all debate in forums like these is rendered rather silly, because people think they're discussing how air molecules are being moved when really they're talking about their inner response to advertising.

 

[opus111]

Laudably balanced, tolerant, eminently sensible. But... if you think that placebo is "real" and understandable and that everyone is prone to it, why do people in these forums and elsewhere virtually never qualify their sighted 'tests' with the caveat that they may have experienced a placebo effect?

Its an interesting question. I can only speculate but perhaps they don't acknowledge the placebo effect or imagine that while it exists, they are in a privileged position and therefore it only exists for other people? Its quite normal for people to imagine they are somehow 'privileged' compared to others - if you take a poll of drivers, the majority usually claim to have above average driving ability don't they? This is all well set out in an excellent little book by Cordelia Fine:

http://www.amazon.com/Mind-Its-Own-...636/ref=sr_1_1?s=books&ie=UTF8&qid=1371545701

Instead, furrows are repeatedly ploughed, asserting that feedback is bad, digital is harsh, Class A is wonderful etc. because listening tests show it.

I'd say that listening shows that many feedback amplifiers don't give a satisfying sound. Audiophiles do tend to jump to conclusions that feedback is somehow responsible but to me this is a category error. You can't listen to feedback because its a concept, not a thing. But perhaps that's just a distraction in the current debate?

I think most objectivists consider "real" to be that which is designed and built by proper engineers ......

But they haven't to my knowledge given a falsifiable category definition of 'proper engineer' and they're never likely to because it defies scientific investigation. So this is a religious hypothesis in that it can't be falsified. Ditto 'competent designer'.

If all audio systems depend on placebo effects to work, it is hard to see how or why anyone should study audio engineering at all;

To me pretty much all marketing depends on placebo, not just audio. Here for example we have Haagen Dazs ice cream, whenever I ask my students 'how does it taste?' they say 'just so-so' yet its marketing ensures it continues to sell, the high price compared to local options being a marketing feature rather than a bug. Ditto Bose for various audio systems. Two examples which aren't even in the world of audiophila. So why would audiophile marketing be any different?

 

[opus111]

Even "bias" tends to have negative connotations. It would be good to have a term that is really neutral, so that people can acknowledge that it is something entirely natural and common, and invoking it is not an insult.

A fair point so drop the 'bias' part and just leave 'expectation effects' ? Or even 'due to expectation' ?

Its important to note (but almost universally not noted, just as subjectivists rarely acknowledge expectation in sighted tests) that expectation effects apply just as much to experimental design as to sighted listening. So objectivists pretty consistently design experiments according to their own expectation - that of no audible difference. And its not at all surprising that most experiments do indeed return this result.

 

[Julf]

To me pretty much all marketing depends on placebo, not just audio. Here for example we have Haagen Dazs ice cream, whenever I ask my students 'how does it taste?' they say 'just so-so' yet its marketing ensures it continues to sell, the high price compared to local options being a marketing feature rather than a bug. Ditto Bose for various audio systems. Two examples which aren't even in the world of audiophila. So why would audiophile marketing be any different?

For me the classic illustration is always Pen & Teller: BS: truth about bottled water. The relevant part starts at 2:00.