[medium jim]

TONEPUB:

Analog recording on 2" tape is/was way more expensive than the current digital. This said, the computer technology back in the late 1970's when the first cd's were produced sound less than desirable.  In the early 1980's the technology and mastering began to improve, yet many LP's literally were converted to cd's using the original analog tapes without remastering.

Eventually time and technology caught up and cd's have become enjoyable.  Yet some the remastered cd's have been priced almost twice as much as their original counterparts.  In my humble opinion, a well mastered Redbook cd is every bit as good as hi rez for 2 channel stereo. Many will disagree, so be it.

Yet those 190gm LP's still cost more and you need to buy stuff to keep them clean and dust free. Get up every 22 minutes to flip them.  With a good tube system you get the glorious soundstage and depth of vinyl with your cd's and the warmth too.  Again, my opinion, yours may be different.

Jim

[JohnR]

You're right, it doesn't. What it does do is make it a claim that can't just be rejected out of hand as completely untestable. If it were really about people's internal states, and not about the causes of them, then, modulo doing brain scans or something, it wouldn't even fall into the realm of empirically testable claims.

Hm, but the claim that the null hypotheses should be expectation bias ignores the possibility that the reason there are no good known engineering reasons (for some subjective observation) is that there is a good engineering reason that is unknown. I mean, I'm not trying to be smart, but...

Engineering is not a black and white thing as people often seem to think. It's about solving problems. Plus, there is a complex interaction with psycho-acoustics. While I think I'm generally in agreement with you, I'm not comfortable with the immediate leap to the conclusion that, if the reason for something is unknown, it is by default expectation bias.

[thunderbrick]

No Brick, I wasn't referring specifically to you. I was being funny. I have experience and training at being funny. You on the other hand don't. So your opinion on funny is worthless, and subject to ridicule. Just the way it is. Those of us with experience and training have to be taken for our word. You're not allowed to disagree...because you don't have experience and training. It's not important that I can't tell you where I got that experience and training, but I got it. You know what I mean?

Doc

Yes SIR!  Doc, SIR!  Bullshit, SIR!       I, too, am trained at being funny, but just as I take your word on your experience and training, you are equally bound to take my word on MY experience and training.  Since obviously neither of our sources of experience and training are accredited, I will lay claim to being taught by the best.

Myself.  Sir! 

[thunderbrick]

Hm, but the claim that the null hypotheses should be expectation bias ignores the possibility that the reason there are no good known engineering reasons (for some subjective observation) is that there is a good engineering reason that is unknown. I mean, I'm not trying to be smart, but...

Engineering is not a black and white thing as people often seem to think. It's about solving problems. Plus, there is a complex interaction with psycho-acoustics. While I think I'm generally in agreement with you, I'm not comfortable with the immediate leap to the conclusion that, if the reason for something is unknown, it is by default expectation bias.

I work at an engineering university, and we have a helluva time getting engineers to agree with each other!   

We also favor a saying that "There's a whole lot more to engineering than just engineering!"   

[totoro]

Hm, but the claim that the null hypotheses should be expectation bias ignores the possibility that the reason there are no good known engineering reasons (for some subjective observation) is that there is a good engineering reason that is unknown. I mean, I'm not trying to be smart, but...

Engineering is not a black and white thing as people often seem to think. It's about solving problems. Plus, there is a complex interaction with psycho-acoustics. While I think I'm generally in agreement with you, I'm not comfortable with the immediate leap to the conclusion that, if the reason for something is unknown, it is by default expectation bias.

As letitroll noted, this was mostly for the sake of argument. My main point was really that there are a lot of people who seem to think that when someone makes a claim about expectation bias, they are somehow denying someone else's subjective experience, and that this is simply not the case. This has come up in this thread, and I've seen the same kind of "how dare you question my experience of reality" meme in other discussions and on other fora for years now. I really never intended to drag us down a rathole of figuring out when expectation bias needed to be accounted for. If I gave that impression, then mea culpa, but it wasn't my intent at all.

That being said, the way I've always understood the null hypothesis is as follows. Suppose we are in the situation where we have a claimed difference in sound between two pieces of equipment which our current engineering knowledge would predict sound the same.  We would then have two hypotheses: that the equipment sounds different, for reasons we don't understand, or it sounds different because of expectation bias. Traditionally in statistics, the "simpler" hypothesis is made the null hypothesis. I don't understand how one would claim that the hypothesis "this equipment sounds different for reasons we don't understand" is simpler than the hypothesis "this equipment sounds different because of a well-known phenomenon (expectation bias)". If there really is a difference, it should show up in the tests. If the difference is unexplainable and not reproducible in a manner that accounts for expectation bias, it's hard for me to imagine believing in it.

But as you say, most real world cases are far more complex than this hypothetical one, there are measurable differences between devices under test, these kinds of tests are actually hard to do and expensive, etc.

[JohnR]

I don't understand how one would claim that the hypothesis "this equipment sounds different for reasons we don't understand" is simpler than the hypothesis "this equipment sounds different because of a well-known phenomenon (expectation bias)".

I'm pretty sure that what you are proposing there is a (different) alternative hypothesis, not the null hypothesis. Interesting point though, I need to think about it...

[eclein]

...man I enjoy reading this thread. 

[JohnR]

OK, so the null hypothesis "proposes that no statistical significance exists in a set of given observations. The null hypothesis attempts to show that no variation exists between variables, or that a single variable is no different than zero. It is presumed to be true until statistical evidence nullifies it for an alternative hypothesis."

http://www.investopedia.com/terms/n/null_hypothesis.asp

Reading elsewhere, the null hypothesis is also formulated as something that you wish to show is not true. You cannot prove the null hypothesis, nor can you disprove it, but you can gather statistical evidence to demonstrate that the alternative hypothesis is more likely.

Let's say we took a pair of components that are identical in every way. The null hypothesis is that listeners do not detect any difference between the two. We then conduct a test in which listeners are told that one component costs ten times more than the other. Moreover, they are told which one they are listening to. If the statistical analysis shows that listeners consistently detect a difference between the two components in favor of the "more expensive" one, then the hypothesis that listeners detect no difference, even for identical components, is rejected.

This is an example where expectation bias is the alternative hypothesis. Given sufficient examples of this test, I think it could be concluded that expectation bias is a real effect.

Let's take the case where two components have measurable differences i.e. "good known engineering reasons to sound different." I think that even in this case, the null hypothesis must be that listeners cannot detect a difference, assuming they are not told which is which. If the statistical analysis shows that listeners can detect a difference, then the null hypothesis is rejected (in this test and at whatever level of statistical confidence). But, it should not be hard to concoct tests where the null hypothesis is almost always rejected (put a highpass filter at 1 kHz in one component, or smash it with a sledgehammer), or hardly ever rejected (make a very minor change to frequency response in the bass region, such as changing the low frequency rolloff from 10 Hz to 12 Hz).

So even in the presence of "good known engineering reasons to sound different," this can only be demonstrated some of the time.

I realize this is long-winded, but I'm getting there.

[JohnR]

Suppose we are in the situation where we have a claimed difference in sound between two pieces of equipment which our current engineering knowledge would predict sound the same.  We would then have two hypotheses: that the equipment sounds different, for reasons we don't understand, or it sounds different because of expectation bias.

OK, so, when we go to something like this, the first question is whether we are dealing with a controlled test, where there IS a hypothesis and statistical analysis, or an individual report i.e. hearsay. I'm somewhat leaning to the position that in the latter case it's mildly pointless to attempt to apply concepts like these to an individual observation. "When I was listening to my system I heard aliens discussing the next invasion of the Earth." OK, it seems unlikely but in isolation and with a rigorous approach it's actually not possible to disprove any such claim.

That's an alarming and uncomfortable conclusion for me to reach at this point, but I believe it's correct, if we follow the logic.

[raindance]

I'd like to add a new high end audio rip-off for consideration: tweaking. Let me clarify, as I am a tweaker myself: The rip-off relates to adding boutique high-end components to products where the design is flawed to start with.

For example, a user of the Dared MC7P preamp says it sounds wonderful, but it has too much gain and oscillates just a little. Many users are adding boutique caps and resistors to these preamps, but the design is flawed (if it oscillates, I wouldn't want it near my speakers).

Same with Jolida - and I own a Jolida JD1000a amp myself - none of the commercially available tweaks address circuit design issues, just adding expensive parts to something that could be optimized by addressing design flaws introduced by compromises at the factory.

I guess what the general public does not understand too well is that the majority of product available to  the consumer includes significant trade-offs either introduced by a lack of design knowledge or manufacturing cost considerations. In my experience designing and tweaking tube gear, a lot of what I see was copied from a 1960's design with no modification for modern, high output sources (CD or universal disc players or DAC's) - therefore too much gain - plus fundamental design issues such as not elevating the heaters in an SRPP preamp or gain stage - or negative feedback schemes that don't address stability.

So, let's take a flawed product, add expensive components to it, and wax lyrical about its abilities without fixing the flaws. Right?

[doug s.]

I think what you're mistaking for people giving up "quality in audio" isn't taking into account what is "good enough" for the average, non audiophile.

If you look back to when the CD came out, for the average person that was listening to scratched records on a few hundred dollar player at worst or maybe a japanese receiver, pair of $400 speakers and a technics or dual turntable (or maybe even pre recorded cassettes), the CD was a huge step forward for that listener.  At the end of the analog run, cassettes were selling more than LP's.

The record stores didn't have a mass exodus until digital downloads took over.  Back in the mid 80's when CD's got popular, record stores actually expanded.  Look at the big Tower, Virgin Megastores and even the major bookstore chains, as well as Best Buy, Tweeter and Circuit city all started selling CD's.

Last but not least, the record companies adopted the CD mainly because they felt analog was too easy to pirate from LP to bootleg cassette.

Kind of wacky if you think about it.

i am in agreement w/most of this.  (except i would still take a technics turntable connected to a good hi-end system over any digital source any day.)  but, my main gripe w/this is that, as you say, digital was for the average non-audiophile, not the audiophile.  which is why digital is a big rip-off in hi-end audio.  audiophiles were thrown under the bus, to cater to the lowest common denominator.  this is really sad, as now, w/computer downloads, mp3's, ipods, etc., it seems that the lowest common denominator is getting even worse, not better.

doug s.

[rollo]

OK, so, when we go to something like this, the first question is whether we are dealing with a controlled test, where there IS a hypothesis and statistical analysis, or an individual report i.e. hearsay. I'm somewhat leaning to the position that in the latter case it's mildly pointless to attempt to apply concepts like these to an individual observation. "When I was listening to my system I heard aliens discussing the next invasion of the Earth." OK, it seems unlikely but in isolation and with a rigorous approach it's actually not possible to disprove any such claim.

That's an alarming and uncomfortable conclusion for me to reach at this point, but I believe it's correct, if we follow the logic.

Personally I believe that is it in a nutshell. Ones perception of the resulting sound is what it is all about. Why one comes to that conclusion is the $64,000 dollar question.
   In recent times the measurements of why we hear what we hear is better than ever before. However not conclusive. We have only our ears after all the measurements were conducted to evaluate the result. So in the end do we like it better than before ?? How we get to that is moot as there are just too many variables in our subjective hobby.
   So for me the biggest rip off  are one's own decisions based on blind tests and measurements only. Listen for yourself cause at the end of the day thats all that matters however you shake it. Yes ? No ?



charles

[Ericus Rex]

Im not even a fan of Audiogon, but seriously, how are they such a rip off, even with their current pricing structure?  So they charge a few percent when it's all said and done. 

They provide you with a world wide network of people to buy your product.  Try selling your power amp in the classifieds.
They don't really charge all that much more or less than Ebay, which doesn't always work for selling high end gear.

If you were fortunate enough to have a local shop that sells used hifi gear on consignment, they usually charge about 15-20 percent (still fair)

So, other than the fact that they used to offer this service for free and now they charge a few bucks, why are they a rip off?  I'm sure Audiogon's cost of operation is 100 times what it used to be when they gave it away.

Why is it that so many of you feel that no one that participates in the world of audio should earn a living for what they do and the services they provide?  This has puzzled me for some time now.

You presume far too much about me, Tonepub.  Audiogon should be compensated for their efforts.  I said this much in my second post on the subject.  Maybe you didn't see that one.  Audiogon used to charge a flat $6 insertion fee for each classified (when was it ever free?).  This was fair and certainly gave them a nice profit considering their traffic.  Now there is a consignment charge in addition to a flat $4 fee.  But they are doing no more work nor are they promoting your $10k preamp any more than Joe's $150 Bottlehead yet they will charge you $197 more than Joe for exactly the same service.  This is like going to Jiffy Lube and being charge $24.95 to change the oil in your Camry and the next day being charged $99.95 to change the oil in your Porsche.  Or a painter charging you a different hourly rate after he sees your house is in an affluent neighborhood.  I think in that case, you'd probably call the Better Business Bureau on those campanies instead of defending them on a forum.

A shop should charge a higher consignment fee.  For one, unlike Audiogon, they don't make a single dollar until the item sells.  This means they will  be more likely to actively promote your item.  On an individual basis, a shop will do tons more work than Audiogon to promote and sell your product.  With Audiogon, you take the pictures, you write and post the ad and chances are a living breathing Audiogon employee will never even touch it.  So exactly what service is your 2% consignment fee paying for?

And, yes, ebay is a rip-off as well.  Doubly so if you accept Paypal as payment.

[JohnR]

And, yes, ebay is a rip-off as well.  Doubly so if you accept Paypal as payment.

Triply so if you have to accept payment in a foreign currency!!

Although, I don't mind eBay, at least when it's sold it's sold - most of the time. Bid or begone.

[medium jim]

You presume far too much about me, Tonepub.  Audiogon should be compensated for their efforts.  I said this much in my second post on the subject.  Maybe you didn't see that one.  Audiogon used to charge a flat $6 insertion fee for each classified (when was it ever free?).  This was fair and certainly gave them a nice profit considering their traffic.  Now there is a consignment charge in addition to a flat $4 fee.  But they are doing no more work nor are they promoting your $10k preamp any more than Joe's $150 Bottlehead yet they will charge you $197 more than Joe for exactly the same service.  This is like going to Jiffy Lube and being charge $24.95 to change the oil in your Camry and the next day being charged $99.95 to change the oil in your Porsche.  Or a painter charging you a different hourly rate after he sees your house is in an affluent neighborhood.  I think in that case, you'd probably call the Better Business Bureau on those campanies instead of defending them on a forum.

A shop should charge a higher consignment fee.  For one, unlike Audiogon, they don't make a single dollar until the item sells.  This means they will  be more likely to actively promote your item.  On an individual basis, a shop will do tons more work than Audiogon to promote and sell your product.  With Audiogon, you take the pictures, you write and post the ad and chances are a living breathing Audiogon employee will never even touch it.  So exactly what service is your 2% consignment fee paying for?

And, yes, ebay is a rip-off as well.  Doubly so if you accept Paypal as payment.

E.R.:

Your analogies are not on point.   Fact I own a Toyota 4Runner that just requires a basic oil change every 5K, but my Cadillac uses full synthetic and 9.5 quarts, do you think Jiffy Lube charges me the same? 

What AudiogoN does is give the user worldwide exposure that does cost them money and is worth 2% to me for that reason alone.  Plus it is well visited and I dare say rather a safe place to do business.  Is it perfect, no. 

If I sell my Bottlehead Foreplay, I should pay less that if I sell my Marantz 9's and have no complaint as I will be reaching a dedicated marketplace for each. 

As for sellers being able to charge 3% to cover the PayPal fees, any seller that does so is just hurting themselves as it really makes them look greedy. 

To classify AudiogoN as the biggest rip off is absurd to me, maybe not to you and will respect your opinion.

Jim

[JohnR]

Yes, well, I admit that it's probably not really possible to classify something as a "rip off" if you know in advance what it will cost.

Sometimes its "just da cost of doing business"

[medium jim]

Yes, well, I admit that it's probably not really possible to classify something as a "rip off" if you know in advance what it will cost.

Sometimes its "just da cost of doing business"

rip off verb
Definition of RIP OFF
transitive verb
1a : rob; also : cheat, defraud b : steal
2: to copy or imitate blatantly or unscrupulously
3: to perform, achieve, or score quickly or easily <ripped off 10 straight points>

To me a Rip Off" in the Audio World is any product where the seller, maker, designer, makes claims that are patently false, misleading or otherwise not true.  It has nothing to do about the cost of the item, per se.   

In other words, if it cannot do what it is said to do, it is a rip off, pure and simple!

Jim

[doug s.]

E.R.:

Your analogies are not on point.   Fact I own a Toyota 4Runner that just requires a basic oil change every 5K, but my Cadillac uses full synthetic and 9.5 quarts, do you think Jiffy Lube charges me the same? 

What AudiogoN does is give the user worldwide exposure that does cost them money and is worth 2% to me for that reason alone.  Plus it is well visited and I dare say rather a safe place to do business.  Is it perfect, no. 

If I sell my Bottlehead Foreplay, I should pay less that if I sell my Marantz 9's and have no complaint as I will be reaching a dedicated marketplace for each. 

As for sellers being able to charge 3% to cover the PayPal fees, any seller that does so is just hurting themselves as it really makes them look greedy. 

To classify AudiogoN as the biggest rip off is absurd to me, maybe not to you and will respect your opinion.

Jim

jim, e.r.'s analogy may not be on point, but neither is yours.  take any two cars into jiffy lube that use the same grade and quantity of oil, and do you think the cost will be different?

i do think agon has some issues regarding its fees, but it is one of the costs of partaking in this hobby.  and, if agon didn't get so greedy, perhaps it wouldn't be having the issues it is now having.  it is not the only game in town.  charging a percentage fee does not sound unreasonable to me, but perhaps 0.5% would be more reasonable?  at some point, things will shake out - agon is driving folks to look at other opportunities for trading their gear...

i wish more competition would happen for paypal & ebay - their fees are truly off the charts these days.  technically not rip-offs, as you have knowledge going in... but, where are the wiable alternatives?   

doug s.

[TONEPUB]

PayPal is no different than any other credit card processor.

If you are in business and want to accept credit cards, it's going to cost you a couple of percent.

John R's point about international transactions is true.  I only sell the token bits of vintage gear for our vintage column when we are done with them, but I've stopped taking foreign transactions for this reason.  Once Paypal takes their 6%, Ebay their three, and the additional hassle and time with customs, its just not worth it to ship international.

But again, so much of it is the cost of doing business these days.   

[totoro]

OK, so the null hypothesis "proposes that no statistical significance exists in a set of given observations. The null hypothesis attempts to show that no variation exists between variables, or that a single variable is no different than zero. It is presumed to be true until statistical evidence nullifies it for an alternative hypothesis."

http://www.investopedia.com/terms/n/null_hypothesis.asp

Reading elsewhere, the null hypothesis is also formulated as something that you wish to show is not true. You cannot prove the null hypothesis, nor can you disprove it, but you can gather statistical evidence to demonstrate that the alternative hypothesis is more likely.

Let's say we took a pair of components that are identical in every way. The null hypothesis is that listeners do not detect any difference between the two. We then conduct a test in which listeners are told that one component costs ten times more than the other. Moreover, they are told which one they are listening to. If the statistical analysis shows that listeners consistently detect a difference between the two components in favor of the "more expensive" one, then the hypothesis that listeners detect no difference, even for identical components, is rejected.

This is an example where expectation bias is the alternative hypothesis. Given sufficient examples of this test, I think it could be concluded that expectation bias is a real effect.

Let's take the case where two components have measurable differences i.e. "good known engineering reasons to sound different." I think that even in this case, the null hypothesis must be that listeners cannot detect a difference, assuming they are not told which is which. If the statistical analysis shows that listeners can detect a difference, then the null hypothesis is rejected (in this test and at whatever level of statistical confidence). But, it should not be hard to concoct tests where the null hypothesis is almost always rejected (put a highpass filter at 1 kHz in one component, or smash it with a sledgehammer), or hardly ever rejected (make a very minor change to frequency response in the bass region, such as changing the low frequency rolloff from 10 Hz to 12 Hz).

So even in the presence of "good known engineering reasons to sound different," this can only be demonstrated some of the time.

I realize this is long-winded, but I'm getting there.

Right. We seem to be in agreement on this at this point. And I agree that there are possibly going to be situations where there are going to be good engineering reasons for two components sounding different to people where they don't. In such a case, certainly the result would need some explaining, and using that engineering difference as a reason to market a component based on sound quality would seem like a pretty odd thing to do.

And we can end up in cases where the null hypothesis is not disconfirmed, but this is due to problems with the experiment, for reasons of sampling error, technical glitches, or something else.