|   Where can I buy biwired cable? Is it worth it? | mtzoanos
Oct 15, 2003 1:16 PM | | Hi,
I'd like to try and bi-wire my front and center speakers (B&W 600 series) to see if there is any acoustic benefit from doing so. I've looked around but I can't find any affordable speaker wire. I don't want to spend more than $1-$2 per foot. Can you recommend some places online or retailers?
Thank you,
Michael Tzoanos |
| re: Where can I buy biwired cable? Is it worth it? | Jetsons
Oct 15, 2003 7:48 PM | | http://www.markertek.com/MTStore/store.cfm?framed=/MTStore/list.cfm?Idx=Audio+Bulk+Cable
Check out the Canare StarQuad 4s11, 4 runs of 14 gauge in a twisted configuration. Very easy to biwire with his cable and inexpensive @ $0.69/ft. I've seen boutique wire shops sell terminated speaker cable using this same wire for far more cash like this, http://cgi.ebay.com/ws/eBayISAPI.dll?ViewItem&item=3052226415&category=14966
Jet |
| re: Where can I buy biwired cable? Is it worth it? | mtrycrafts
Oct 15, 2003 11:32 PM | | Get some 12/14ga at Home Depot and run two sets. Inexpensive.
But, it is a waste of time and effort. No evidence to support it. |
| re: Where can I buy biwired cable? Is it worth it? | Bill Clinton
Oct 16, 2003 11:32 AM | | If you can manage it, get a local hifi store to let you demo some high quality biwire cable and do A/B testing.
Let your ears decide. Given the biwire debate, only you can decide whether you can hear an improvement.
Personally, I'm a convert. I can hear a huge improvement in my system. As for technical proof, this is the best explanation I've seen...
http://sound.westhost.com/articles.htm
You can also order a copy of the Genesis report from QED. It was a seminal thesis on cable design.
As for doing it cheaply, I'm a cable convert in general and think you should splash out on high quality cabling, preferrably silver plated. Though, the new QED x-tube stuff shows real promise. So, do some research and keep an eye on ebay. |
| No biwire article is listed | RobotCzar
Oct 16, 2003 12:23 PM | | Did I miss something? I can't find an article on biwiring on the menu of your reference link.
There is an article on biamplication, a different concept.
There is no logical, scientific reason to think that biwiring could possibly improve the audio quality of your system. If you heard a difference it could be due to simple volume differences between the bi and mono wired cable. If you hear a "huge improvement" it is most likely due to your imagination (i.e., your expectations influencing your hearing). Maybe you were listening to a biamped system?
The original poster should note that there is no need to use special cable to biwire. |
| re: Where can I buy biwired cable? Is it worth it? | mtrycrafts
Oct 16, 2003 12:42 PM | | b I can hear a huge improvement in my system.
Perhaps you failed to completely read your own referenced link, specifically the DBT/ABX links.
You did compare your before and after under level matched, DBT listening protocol, right?
No one has yet demonstrated in a credible DBT audible differences with biwiring. You are not the first and only. So, there are other explanations to your perceptions: biased, flawed protocol, imagined, etc.
b You can also order a copy of the Genesis report from QED. It was a seminal thesis on cable design.
I am sure they have supporting DBT listening data to support any audibility claims, right?
b As for doing it cheaply, I'm a cable convert in general and think you should splash out on high quality cabling, preferrably silver plated.
Yep, lot of gullible people on the planet, in all consumer market place. Audio is not excluded.
Silver plating does nothing at audio frequencies. Just another audio mythology from the gullible ones.
b So, do some research
Oh, I think it is you who is in dire need of better research supported by science, not audio voodoo. |
| A predictably over-zealous response... | Bill Clinton
Oct 17, 2003 7:54 AM | | Despite your passion for bell wire and penchant for stirring up trouble, you should let the man decide for himself. There are lots of people who are not dumb, gullible, or have too much money...who believe it has an advantage.
And yes, the link does address biwiring in what I think is a very unbiased way. The author clearly prefers biamping, but explains the theory behind it very well. He also purports in his other articles that cabling, beyond a certain minimum standard, doesn't make that much difference (a view I only partly subscribe to, but a well presented argument none the less).
The relevant bits are below. And, without responding to mtycrofter's remarks one by one, all care was taken in setting up the a/b testing to ensure a fair comparison including double blind to avoid bias. I suggest the original poster try it out.
If you need convincing that cables make a difference in a system, start with something easier to test, like a high-quality interconnect. The difference between the thin black jobbies you normally get and a properly constructed ic can be monumental.
-----
1.13 - Bi-Wiring Facts And Myths (and More on Passive Crossovers)
Many speaker systems now cater for bi-wiring - running a separate speaker lead from the amplifier to the low and mid+high crossovers via separate terminals on the back of the enclosure. The benefits of this technique are said to be improved imaging due to the reduced interactions of the loudspeakers and their respective crossover networks, since the amplifier acts as an essentially zero impedance source for each section (the speaker cable now has no influence on crossover performance).
Some people equate bi-wiring as a cheaper method of achieving the same gains as one would with bi-amping. This is quite obviously not the case - there are gains to be had, but they are comparatively minor. This is not to say that the "minor" gains are not worth the effort, because as you will see this is not true at all.
For those who may not be sure of how bi-wiring really works, Figure 5 shows the setup. The broken line indicates where the connection would normally be made internally (i.e. inside the speaker cabinet). When bi-wiring is used, this connection is removed (usually with links on the connection panel), and a separate cable is run back to the amplifier.
This diagram also illustrates the composition of a 3-way crossover network. Low frequencies are fed to the woofer via a low-pass filter. The remaining signal is then fed through a high-pass filter to remove the bass energy. This is the mid+high component. To ensure that the midrange loudspeaker does not receive high frequencies as well (which it would otherwise proceed to mangle), a low-pass filter is used to filter out the high frequency component. Finally, to ensure that the tweeter is protected from the midrange signals, another high-pass filter is used.
The cutoff (i.e. crossover) frequencies for the two filter "groups" will typically be in the range of 300Hz to 800Hz for the low/mid+high section, and 2kHz to 6kHz for the mid/high section. Crossover frequency selection is discussed in a following section. (A Short Dissertation On Crossover Frequency Selection)
Figure 5 - Bi-Wiring Connections
As can be seen, the low frequency energy is now separated from the mid+high frequency energy in the cabling. The amplifier must still handle the full frequency range, but each section of the crossover has its own cable feed, which prevents some of the interactions between the crossover sections.
The overall effect is often (or so it has been said) a vast improvement, and largely because of the fundamental imperfections of passive crossover networks. These networks (regardless of their cost or complexity) have a few basic weaknesses which determine their overall performance. Basically, these are:
Input impedance - a properly designed inductor / capacitor (L/C) filter rel |
| A predictably over-zealous response... | FLZapped
Oct 17, 2003 9:30 AM | | Ahh, but the facts and myths are not seperated here:
i 1.13 - Bi-Wiring Facts And Myths (and More on Passive Crossovers)
Many speaker systems now cater for bi-wiring - running a separate speaker lead from the amplifier to the low and mid+high crossovers via separate terminals on the back of the enclosure.
b Fact.
i The benefits of this technique are said to be improved imaging due to the reduced interactions of the loudspeakers and their respective crossover networks, since the amplifier acts as an essentially zero impedance source for each section (the speaker cable now has no influence on crossover performance).
b Myth(factually in error, too)
i As can be seen, the low frequency energy is now separated from the mid+high frequency energy in the cabling.
b Myth, both cables have the same information on them.(see below, why)
i The amplifier must still handle the full frequency range, but each section of the crossover has its own cable feed,
b Fact and places that on BOTH cables.
-Bruce |
| Correction..... | FLZapped
Oct 17, 2003 9:40 AM | | i As can be seen, the low frequency energy is now separated from the mid+high frequency energy in the cabling.
b Myth, both cables have the same information on them.(see below, why)
Darn, missed the word energy the first go. The signal information is on both cables, however the energy distribution is in fact different. My mistake.
-Bruce |
| Hey Bill | pctower
Oct 17, 2003 10:01 AM | | Looks like you got bit by the nefarious AR post limit. When I have a long post, I general compose it in Word and then cut and paste into subsequent replies to my initial post until Ive included everything I wanted to appear in the post. In my case, it also helps a great deal to have my posts subjected to spell-check, which is another advantage of composing in Word.
The important thing when you have a long post is to look at the post in its final form (not how it appears immediately before posting) to ensure that nothing has been cropped from your post.
AR has indicated they are aware of the problem but that a fix would involve too much of a re-write of their software.
BTW, I own Vandersteen Fives (Ive established a cookie jar for my pennies hoping that before I go deaf Ill be able to upgrade to the Five As), and, as you know, Richard claims to build his speakers specifically to enable his first order cross-overs to take advantage of bi-wiring. Ive never tried my speakers (Ive owned Vandys for over 15 years) without bi-wiring. I know Richard well enough to know that he lives his life to produce the most cost effective speakers he can and that he does not arrive at any design decision lightly. So I have more or less blindly followed his advice.
However, Bruces post raises an issue Ive never understood, and that is how simply moving the joinder point from the speakers back to the amp changes anything. Seems to me the circuit involved is still the same. Ive read some technical discussions attempting to explain why bi-wiring makes a difference, but Ive never been convinced by what Ive read. |
| A predictably over-zealous response... | mtrycrafts
Oct 18, 2003 12:15 AM | | b all care was taken in setting up the a/b testing to ensure a fair comparison including double blind to avoid bias.
That is a dubious claim as there is no known evidence to support this claim of audibility of this method of wiring, zero. Maybe he can demonstrate his claims under 3rd party supervision to make sure he is not forgetting something.
b The benefits of this technique are said to be improved imaging due to the reduced interactions of the loudspeakers and their respective crossover networks, since the amplifier acts as an essentially zero impedance source for each section (the speaker cable now has no influence on crossover performance).
Said to be? He doesn't know? Why not? No evidence?
b The overall effect is often (or so it has been said) a vast improvement,
Ah, once again no evidence, just speculating and admitting it. |
| The article is wrong | RobotCzar
Oct 20, 2003 3:19 PM | | "since the amplifier acts as an essentially zero impedance source for each section (the speaker cable now has no influence on crossover performance)".
This statement is incorrect to me. The author makes the same mistake as many biwiring analysts. The two paths to the crossovers are connected to the same amp terminal (not to the amp--ie, through the amp). Perhaps "zero impedance" is the correct phase--in other words, there is no difference in a biwired circuit and a monowired one.
His assertion that the two filters are "separated" in the biwired circuit is clearly wrong as they are shorted at the amp terminal. They certainly "see" each other across the terminal connection. This simple, simple error calls into question the whole rest of his web site. Too bad, he seems to make a lot of good points.
I indend to write him and seek a correction. |
| The article is wrong | pctower
Oct 3, 2003 9:50 AM | | Your analysis makes sense to me.
However, I recall hearing an argument to the effect that reflected waves from each part of the crossover (perhaps caused by reverse EMF) will be shorted at the amp with bi-wires and therefore not affect the signal in the opposite pair. Setting aside for the moment the fact that no one has apparently ever established the audibility of bi-wiring under normal conditions (reasonably size gauge wire in both runs) using valid control testing, do you think the reflected waves argument has any possible validity? |
| The article is wrong | mtrycrafts
Oct 21, 2003 12:44 AM | | May want to refresh from Steve Eddy
http://www.audioasylum.com/audio/cables/messages/35554.html |
| The article is wrong | pctower
Oct 21, 2003 7:44 AM | | Thanks. I now remember that thread, and of course, Steve's explanation makes sense. |
| Here is another | mtrycrafts
Oct 21, 2003 9:40 PM | | that I was trying to find in my references and missed the forst time around. Dick Pierce knows a few things too well:)
http://216.239.53.100/search?q=cache:abjmuxZAPp4C:www.diyspeakers.net/Articles/Richard%2520Pierce%2520DAMPING%2520FACTOR.pdf+damping+factor&hl=en&ie=UTF-8
got to copy the whole link |
| re: Where can I buy biwired cable? Is it worth it? | FLZapped
Oct 17, 2003 9:38 AM | | The correct link should be:
http://sound.westhost.com/bi-amp.htm#passive_xovers
The references were buried far down in an article on another topic.
-Bruce |
| re: Where can I buy biwired cable? Is it worth it? | FLZapped
Oct 16, 2003 7:05 AM | | If you just want to play around, go over to your local Home Depot(etc) and grab some generic wire. Low-voltage Landscaping wire would be a good choice and make your own.....
Now then, there are a large number of variables(6 minimum) involved in trying to determine if bi-wiring will actually make an electrical difference. So if it actually does, will that be audible? So far there is no evidence to support that it will.
-Bruce |
| Biwired cable? | Swerd
Oct 16, 2003 10:32 AM | | Your question raises two different issues. The first one is whether there is any audible benefit to biwiring. I will sidestep that one because there is no clear answer to it. However, you can try it for yourself, without major cost, by using double lengths of inexpensive zip cord wire. Pick any gauge between 16 and 12.
The words "biwired cable" in your question raises the second issue. I assume by "biwired cable" you mean a bundled set of (at least) 4 wires. Some cable manufacturers sell these bundled sets to minimize the clutter from pairs of normal cable. However, this defies the standard theoretical reason given for the benefit of biwiring, which goes something like this:
Bass signals are much stronger than the signals for the mid and treble ranges. The stronger electromagnetic fields in the speaker cable created by the bass signals can interfere with the weaker upper range signals, degrading their sound. By using two separate cables (and separate crossover networks) the weak upper range signals are isolated from the stronger bass signals, eliminating the possibility of interference.
The strength of an electromagnetic field falls with the square of the distance between the two cables. Therefore, if you want to take any advantage of biwiring, you should avoid keeping the two cables close to each other. A few inches of separation could make a big difference, if you believe the theory.
So try double sets of zip cord. If you like it, great. But don't bundle the cables close together because that would defeat any possible benefit. |
| re: Where can I buy biwired cable? Is it worth it? | okiemax
Oct 16, 2003 1:05 PM | | Many speaker makers are sold on the benefits of biwiring --Polk, Infinity, and Vandersteen --to name a few. Here are statements from the Polk and Vandersteen web sites.
"I was shocked by the difference bi-wiring made with LSi9's in my listening Room." Presumably these were Matthew Polk's words.
This is from Vandersteen: "The effects of bi-wiring are not subtle. The improvements are large enough that a bi-wire set of moderately priced cables will usually sound better than an single run of expensive cables."
Not all speaker producers, however, favor bi-wiring. The following is from the Thiel web site: "Thiel speakers do not offer bi-wiring or bi-amping capabilities. It is our opinion that using one better amplifier or a single, but better, set of cables offers superior performance than using two lesser amplifiers or lesser sets of cables with Thiel speakers."
Using zip cord to experiment with your B&W's sounds like a good idea. Please let us know about your findings. |
| re: Where can I buy biwired cable? Is it worth it? | mtrycrafts
Oct 17, 2003 12:02 AM | | b "I was shocked by the difference bi-wiring made with LSi9's in my listening Room." Presumably these were Matthew Polk's words.
Or not his words but customer feedback which can be unreliable.
b Many speaker makers are sold on the benefits of biwiring
Yes, they don't want to lose any sales to audiophiles who insist on buywiring. A business decision, nothing more.
Paradigm has a comment in their manual. The company is well known for doing DBT research. Try calling them and ask for their DBT data on buywiring audibility.
b This is from Vandersteen: "The effects of bi-wiring are not subtle. The improvements are large enough that a bi-wire set of moderately priced cables will usually sound better than an single run of expensive cables."
Call them and ask for their DBT data. As this comment is sheer speculation.
b Please let us know about your findings.
Do you think he should implement some sort of bias controls? |
| re: Polk says bi-wiring makes shocking difference | okiemax
Oct 17, 2003 12:22 PM | | I checked the Polk web site again, and it sure looks to me like Matthew Polk or someone else in his firm is saying the difference with bi-wiring is shocking. It's not a customer's statement. You can see for yourself:
http://www.polkaudio.com/faqad/questions.php
Note: See Polk's answer to Question #14, "Why are there two sets of inputs on my speakers?"
Paradigm is another speaker maker that believes bi-wiring can be an impovement. The following statement is from their web site: "Bi-wiring/bi-amping can improve clarity, detail, and bass response. And since bi-wiring only requires annother set of speaker cables, this can be a very inexpensive option to consider."
http://www.paradigm.ca/Support/TechFAQ.html#Question12
I don't have experience with bi-wiring, so I don't have an opinion on it. In the future, I may own speakers that are bi-wireable, and that is why I am interested in hearing about opinions and DBT results. Bi-wiring is such an inexpensive option, however, that experimenting with it seems like a "no-brainer." |
| re: Polk says bi-wiring makes shocking difference | mtrycrafts
Oct 18, 2003 12:38 AM | | Thanks for th elink. I found it after some effort:)
b According to one theory, by providing each "half" of the signal a separate wire path, interference effects within the wire are reduced producing better sound.
Again, they are just speculating here otherwise they would have have made it a fact with data and not made it ambiguous. No, they have no such data as it doesn't exist.
They offer no real evidence. Neither does Paradigm and Paradigm has the facility to conduct DBT listening. I have called Paradigm engineering. They don't have the evidence. All marketing.It happnes. I had a email from another speaker company stating exactely just that: marketing makes up what you read, not the engineers. |
| re: Polk says bi-wiring makes shocking difference | okiemax
Oct 18, 2003 9:24 PM | | I'm sorry for the elink problems, particularily the Paradigm link, which did't even go to their web site. I would welcome advice.
I have doubts about the use of ABX DBT's for detecting subtle differences in audio components. My main concern is whether these tests are the same as listening at home. I haven't seen the results of a DBT of any components in my hifi system. If such tests existed, and conflicted with my experience, I wouldn't conclude that the test results are right and I'm wrong. I would just think they were two different things. Why should I deny what I hear because of something I have read? |
| re: Polk says bi-wiring makes shocking difference | mtrycrafts
Oct 18, 2003 11:51 PM | | b I have doubts about the use of ABX DBT's for detecting subtle differences in audio components.
That is unfortunate as DBT, ABX being one such type, is th eonly way to detect small and threshold levels of differences. Those are th efacts, indisputable.
b My main concern is whether these tests are the same as listening at home.
It is usually done at the listeners home.
b I haven't seen the results of a DBT of any components in my hifi system.
Is yours unique? Have unique qualities? Specs? Specs do correlate to what you hear, no matter that some will try to tell you otherwise.
DBT listening do show little differences in modern components of competent design.
b Why should I deny what I hear because of something I have read?
Because your sighted listening is unreliable to begin with, even before you read anything about anything of DBT. |
| re: Polk says bi-wiring makes shocking difference | okiemax
Oct 19, 2003 12:04 PM | | How can a DBT be the "only way" to detect small differences? Wouldn't hearing be the same with or without an ABX comparator? If its good at detecting small differences, why are so many tests inconclusive(null hypothesis not rejected)?
If you are saying ABX DBT"S usually are done at the listener's home, do you mean with a hardware-based comparator(the box) or its software equivalent? I don't see how the latter could be used for testing my hifi components. I suppose comparator boxes have been used long-term in some homes, but I haven't seen the results of such tests. I have less than $1,400 invested in my two audio systems, so I am not interested in buying an expensive comparator($715 at QSC Audio's web site) for experimentation.
No one component in either of my sytems is unique. The systems and the settings, however, may be unique. For example, I doubt there is another living room exactlty like mine with exactly the same combination of components.
Short-term sighted listening can be unreliable, which is why I like to evaluate a new component over a week or two before deciding whether to keep it. I feel this is enough time to eliminate any bias on my part from a placebo effect(i.e., I wanted it to be better so I imagined it was).
If the placebo affect is not lasting, how can a long-term ABX DBT give a much different result than a long-term sighted listening test? I think the former usually would tend to confirm the latter. I can imagine myself thinking there might be a small positive difference in a long-term sighted test of a new component, but getting an inconclusive result using an ABX comparator. But If I thought the difference was small, I probably would have returned the component anyway. |
| re: Polk says bi-wiring makes shocking difference | mtrycrafts
Oct 20, 2003 12:04 AM | | b How can a DBT be the "only way" to detect small differences? Wouldn't hearing be the same with or without an ABX comparator?
Acoustics, hearing, perception, brain translations of information, or making it up out of thin air is some of the disciplines one needs to explore.
Hearing is the same. The sound waves travels to the ear. Electrical signal to the brain. Or, no signal to the brain, yet it registers it anyhow, makes it up, fills in missing information.
The ABX box is a tool to select, switch and record your response, not a mandatory component of DBT listening.
n If its good at detecting small differences, why are so many tests inconclusive(null hypothesis not rejected)?
A null is not inconclusive, it is a failure to hear differences. It has meaning. Why must there be audible differences? That is what the test is for to determine differences or lack of it. A null has meaning, just as in medicine and any other disciplines. A null is reached because on could not hear a difference beyond chance events, guessing.
b If you are saying ABX DBT"S usually are done at the listener's home, do you mean with a hardware-based comparator(the box) or its software equivalent? I
Yes, the box is not large. You can also swap cables but your memory will be hindered due to the time it takes to swap cables. Acoustic memory for small differences is very short. Just as sampling a paint chip.
b I don't see how the latter could be used for testing my hifi components.
Why not? How are you different from anyone else who has undertaken a DBT at home?
b I suppose comparator boxes have been used long-term in some homes, but I haven't seen the results of such tests.
There have been some, one published some time back:
"The Great Ego Crunchers: Equalized, Double Blind Testing", Shanefield, Daniel, Hi-Fidelity, Mar 80, pg 57-61.
This one lasted several months.
Long term listening is just another audio mythology, bs not real meaning for differences.
b I have less than $1,400 invested in my two audio systems, so I am not interested in buying an expensive comparator($715 at QSC Audio's web site) for experimentation.
More reason to accept what has transpired in the past and avoid all the audio hype, voodoo and bs.
Audible differences between cables and components are way over hyped to sell, sell, and sell even more and the need for constant upgrade. How else would they sell their products?
b For example, I doubt there is another living room exactlty like mine with exactly the same combination of components.
You are right. But, your room will not let you hear differences while other rooms don't. Does it have a noise floor close the the threshold of hearing? I doubt it and will be closer to 35-40dB above it.
n I feel this is enough time to eliminate any bias on my part from a placebo effect(i.e., I wanted it to be better so I imagined it was).
No. Absolutely not. Bias is not controllable. It is sub conscious, it comes and it goes, unpredictable when. That is why DBT is used.
However, over time using a component will let you get familiar with its features. Maybe you need something it doesn't offer.
b If the placebo affect is not lasting, how can a long-term ABX DBT give a much different result than a long-term sighted listening test?
Placebo, bias is unpredictable when it comes or goes away, for how long. You have no control over it. That is why it is so important to control for bias/placebo if audible differences is what you are basing your buying decisions. If not, it doesn't matter then. |
| A Good Reply | okiemax
Oct 21, 2003 12:58 AM | | Thanks for a thorough and well thought-out reply. I will see if my public library has a Mar 80 Hi-Fidelity archived.
Perhaps my saying an ABX DBT was inconclusive if the null hypothesis could not be rejected was not the best way of putting it. However, I said "inconclusive" because a null result means I can neither conclude that the two components sound different or sound the same. A null might suggest they sound the same or nearly the same, but I can figure that out through long-term sighted listening.
The placebo effect on choice of audio components is an interesting subject. I suspect I have at times been influenced by expectations. But I have rejected too many components to believe the effect is lasting. I see it as just another "its in you head" explanation, the reason given when nothing else is known. I'm not dismissing the placebo effect completly, but I believe its importance is overblown. |
| A Good Reply | mtrycrafts
Oct 21, 2003 9:55 PM | | b I will see if my public library has a Mar 80 Hi-Fidelity archived.
Most likely not but you can request an inter library loan or have a copy made and sent. Not expensive. Besides, it is continuing education, not free:)
b However, I said "inconclusive" because a null result means I can neither conclude that the two components sound different or sound the same.
It cannot do both. How can it? If you didn't reject the null it means you didn't hear a difference to be statistically sure it was a real difference. People have a hard time accepting the null hypothesis, but, after a while with nothing but nulls, it is prudent to do so. One can always change their mind with better data in the future.
b A null might suggest they sound the same or nearly the same, but I can figure that out through long-term sighted listening.
No you cannot. Sighted listening is unreliable, short or long term.
b But I have rejected too many components to believe the effect is lasting. I see it as just another "its in you head" explanation, the reason given when nothing else is known. I'm not dismissing the placebo effect completly, but I believe its importance is overblown.
Then medical research would be all sighted as it is long term, years, if you read the research.
Its imprtance is not over blown at all. Not when people respond as hearing differences 75 % of the time when nothing is really changed but the same component is presented twice.
Your next research asignment is to check out some psychology research into human bias, placebo, human testing protocol and why it is necessary. We do know that it is real, uncontrolable, not predictable. |
| My take on the null | okiemax
Oct 22, 2003 12:16 PM | | What I said about the null hypothesis is consistent with what is said on Carlstrom's ABX web site which you suggested to me several weeks ago. You quoted the following statement by me regarding my understanding of what it means when the null can't be rejected: "However, I said inconclusive because a null result means I can neither conclude that the two components sound different or sound the same." How is my statement different in meaning than the following quote from the ABX site (ABX Statistics, P.9):
"Note that no matter what score is achieved, A=B cannot be proven. That is the ABX Double Blind Comparison can never be used to prove two audio components sound the same. The notion that ABX can prove components sound the same is a common misconception about ABX."
I still maintain that long-term sighted listening will tell me more about an audio component than a short-term ABX DBT. It just doesn't seem reasonable to me that the placebo effect would bias me long-term. Perhaps the influence on some listeners is lasting, but I haven't experienced it that way. I have some thoughts on the placebo effect which I posted earier today as a new subject. |
| I find it interesting that many people | Pat D
Oct 24, 2003 8:14 AM | | seem to ignore the interpretation of a positive result in a DBT, as well. Here's both together:
" This literally means that the score (20/30) is probably not random except for a 0.049 chance that it is random. Thus the result of the experiment is that the null hypothesis is not true except for a 0.049 chance it is true, or in audiophile terms, A sound different from B except for a 4.9% chance that they are the same."
"Note that no matter what score is achieved, A = B cannot be proven. That is the ABX Double Blind Comparison can never be used to prove two audio components sound the same. The notion that ABX can prove components sound the same is a common misconception about ABX."
So, neither a positive nor a negative result is absolute proof, only probability.
http://www.pcavtech.com/abx/abx_p9.htm
Your beliefs about long term sighted listening don't make it any more reliable. |
| I find it interesting that many people | okiemax
Oct 24, 2003 1:34 PM | | I'm not sure I understand your point, so I hope my comments here relevant. I am not an expert on statistics, but I think you may be equating two types of errors that are different.
You raise the possibility of the risk of a Type I Error(rejecting the null hypothesis when it's true).This can be avoided by selecting a lower significance level -- instead of 0.05, select 0.01 or something even lower if you wish. On the other hand, it may not be possible to avoid the risk of making a Type II Error(failing to reject the null when it's false), altough the possibility can be reduced by increasing the sample size.
I think of a long-term listening as increasing the sample size, whether it be sighted or double blind. I believe the results of any long-term listening tests I do, and any subsequent decisions to purchase or keep a component, will be about the same with either method. |
| I find it interesting that many people | pctower
Oct 24, 2003 2:24 PM | | Your point regarding Type II errors is well taken and one that I have pointed out before. However, I fail to see how you can equate long-term sighted listening (which doesn not eliminate bias) with increasing the sample size in a DBT.
It seems to me that long term listening may disclose differences that cannot be detected in shorter tests, but to my knowledge no long-term listening tests have ever been conducted under control conditions, so whether such differences could be detected in long-term listening seems to me to be purely speculative.
I am certainly no authority on statistics either, but it seems to me that most of the audio DBTs I have read about ignore the Type II error problem by employing too few samples. On the other hand, as the number of samples increase listener fatigue has to set in and it may be impractical to conduct DBT audio tests with a sufficient number of samples to produce reliable results.
These type of issues don't seem to get discussed much on the audio boards because the exchange rapidly devolves into a battle of dogmas, rather than dispassionate consideration. I've asked before on this board and others for someone who has a firm grasp on statistics to discuss this type of issue, but have yet to see one step forward.
Personally, I choose my components and cables based solely on what "sounds" good to me in sighted auditions. But I cringe when people attempt to elevate judgments based on sighted auditions to the level of scientific "proof". If someone wants to design equipment or buy equipment based solely on sighted testing, then, as far as I'm concerned that's their perrogative. But unless someone can report an experience with different cables or components where bias or ABE's have been eliminated, seems to me they need to carefully qualify any claim they might make as to the reliability of such experience from a scientific viewpoint. |
| You should check Mtry's lists of references occasionally. | Pat D
Oct 24, 2003 6:52 PM | | b t seems to me that long term listening may disclose differences that cannot be detected in shorter tests, but to my knowledge no long-term listening tests have ever been conducted under control conditions, so whether such differences could be detected in long-term listening seems to me to be purely speculative.
Mtry has mentioned this several times. You might take notes! One article is by Daniel Shanefield.
"The Great Ego Crunchers: Equalized, Double Blind Testing", Shanefield, Daniel, Hi-Fidelity, Mar 80, pg 57-61.
mtrycrafts 10/20/03 1:04am |
| Question | pctower
Oct 26, 2003 6:14 AM | | Since you apparently are familiar with this reference, is there any reason you couldn't provide a short synopsis? Cables tested, number of samples, test conditions, etc. To the best of my knowledge, mtrycrafts has never done this either.
If your goal is to prove I'm lazy, you've succeeded. If your goal is to educate, you are falling short IMHO. |
| Question | mtrycrafts
Oct 26, 2003 7:00 PM | | Most likely I have not posted that info especially since he compared amps over a number of months.
He compared the Dynaco 400, 200 watts /ch and Lafayette LR 3030A receiver, 30 watts per. Thorens 125TT with SME 3009 arm ADC XLM cart,
5 listeners over time.
1 had 90 switching/trials over 3 weeks. another 40 switchings/trials
19 of 40
20 of 40 for 2 listeners.
Not as easy to interpret as the Greenhill paper witht he bar charts. |
| Question | pctower
Oct 27, 2003 7:13 AM | | Thanks. Appreciate you taking the time. |
| Question | mtrycrafts
Oct 28, 2003 11:37 PM | | Not a problam. A direct question:), a direct answer:)
But best to read it first hand:) |
| I find it interesting that many people | okiemax
Oct 24, 2003 11:07 PM | | I agree that increasing the amount of time devoted to a sighted evaluation is not exactly the same as increasing the sample size in a DBT. However, I think the two are similar.
One way to increase sample size in a DBT is to increase the number of listening trials performed by each participant. Unless listener fatigue becomes a factor, raising the number of trials should reduce the chance of an incorrect conclusion. I believe I am doing something similar by increasing the number of my sighted listening sessions(trials) when comparing two components. More listening sessions should give me a better chance of reaching a correct conclusion. |
| I find it interesting that many people | FLZapped
Oct 26, 2003 3:46 AM | | You seem to miss earlier points. Any sighted methodology is invalid, period. This has been verified retpeatedly and consistantly in multiple disciplines.
-Bruce |
| I find it interesting that many people | okiemax
Oct 27, 2003 2:14 AM | | I appreciate the difference between sighted and double-blind testing. However, I do not believe the possibility of bias inherent in sighted listening makes it useless for me in comparing audio components. Besides,it's hard to do DBT's at home. Go to my next reply for more on this subject. |
| I find it interesting that many people | mtrycrafts
Oct 26, 2003 7:04 PM | | b raising the number of trials should reduce the chance of an incorrect conclusion.
Yes.
b More listening sessions should give me a better chance of reaching a correct conclusion.
Yes, if bias controlled, unless your basis are other than sound alone.
b I believe I am doing something similar by increasing the number of my sighted listening sessions
No matter what you, you cannot improve on sighted listening. It is unrelaible, period, for small differences. Incontrovertible. |
| I find it interesting that many people | okiemax
Oct 27, 2003 1:22 AM | | I'll try to more fully explain why I think long-term sighted listening comparisons in my home are a good way to evaluate audio components I am considering for purchase.
As I have indicated previously, I do not believe imaginary differences in components will survive long-term sighted listening evaluations. Why do I believe this? I have rejected too many components to believe otherwise. The piece being evaluated was audibly about the same as what I already owned or was inferior, so I returned or sold it.
You may disagree that the imaginary difference or placebo effect is temporary. However, if it is permanent or it comes and goes, the listener so affected would be under some enduring delusion, with each additional purchase giving even more pleasure. I don't know anyone that deluded.
I have thought about how interesting it would be to have an ABX comparator to use at home for double-blind testing. A comparator certainly would encourage trying out a lot of new components. But there may be a Catch-22 risk here. What if I buy a comparator to have fun looking for differences, but find almost everything sounds the same through the comparator? I would stop having fun. Is that what you want? |
| I find it interesting that many people | pctower
Oct 27, 2003 7:22 AM | | b I would stop having fun. Is that what you want?
You've hit the very core of why this is such an emotional issues. The yeasayers believe that is the motive of the naysayers. The naysayers believe they are correct about after-market cables being a waste of money and that people would have more fun if they spent their money elsewhere. When their advice is rejected, they believe the person rejecting that advice has been "brainwashed" by the cable "cult".
From that point on, any reasonable discourse becomes impossible.
MonstrousMike once observed that yeasayers and naysayers don't mix very well. Certainly, most prefer to stay on their own board and not "mix" with those that disagree with them. Some even impose rules to prevent the other side from participating. That generates even more animosity. |
| I find it interesting that many people | okiemax
Oct 27, 2003 3:10 PM | | Yeasayers and naysayers probably find it difficult to empathize with each other. If I like to experiment with listening to different amps, cables, etc., I may seem odd to someone who has no interest in doing the same. Rather than being seen as having an unusual but socially acceptable hobby, I may be seen as suffering from a component swapping disorder. At the same time, I may be tempted to see naysayers as overly judgemental people who are so rigid in their beliefs that they are unwilling to even try anything that might alter these beliefs. Both views are unfair. |
| I find it interesting that many people | mtrycrafts
Oct 29, 2003 12:15 AM | | b I may seem odd to someone who has no interest in doing the same.
Oh, but that is not the real story. We are interested in comparing when necessary, but with the proper, meaningful method. Why would you do math if you didn't follow the prescribed rules? You'd be wasting time.
b I may be seen as suffering from a component swapping disorder.
Well, if you do it for the wrong reason, perhaps yes:)
b At the same time, I may be tempted to see naysayers as overly judgemental people who are so rigid in their beliefs that they are unwilling to even try anything that might alter these beliefs.
Not so. We are after the truths. We came to these consclusions by careful experimentation that DBT is the holy grail for reliability in detecting small audible differences. We just know what works reliably and what doesn't.
However, when you publish your case conviningly to be accepted by the community where it matters, we do change. These boards are not the right place for such profound changes to accepted protocols.
Why would I follow astrology? Psychics? Holistic healers? Homeopathic medicines? Have a reading to find out what my dead family has to pass on from the grave?
All are relevant to this issue. |
| How do you keep up? | okiemax
Oct 29, 2003 12:42 PM | | You participate in far more of these threads than I do, and I have trouble keeping up.
I get the impression most naysayers have never tried the audiophile cables they criticize. If they do, they don't talk about it here. The naysayer position seems based on theory and DBT experiments they have read about rather than first-hand experience. The theories may not cover all there is to know, and the use of a DBT in an experiment doesn't in itself assure valid findings if the study is otherwise flawed. |
| I find it interesting that many people | mtrycrafts
Oct 29, 2003 12:03 AM | | b I'll try to more fully explain why I think long-term sighted listening comparisons in my home are a good way to evaluate audio components I am considering for purchase.
If it doesn't involve detecting small audible differences, by all means, any comparison is valid. Nothing further to discuss.
b I do not believe imaginary differences in components will survive long-term sighted listening evaluations.
Yes, you do believe this. People do believe all sorts of false beliefs all the time in all aspects of life.
b I have rejected too many components to believe otherwise.
Not evidence for small audible differences, if that is one of the criteria.
b You may disagree that the imaginary difference or placebo effect is temporary.
And a whole lot of much smarter people in human psychology do as well. I don't set scientific reality.\
b However, if it is permanent or it comes and goes, the listener so affected would be under some enduring delusion, with each additional purchase giving even more pleasure.
Oh, but that is the whole point, you have no idea which way bias and placebo is affecting you as time progresses. It is not constant, linear, but random, unpredictable. Hence your supposition is false.
b I would stop having fun. Is that what you want?
Why would I? Have fun. Don't find out what really is at work. No difference for me. But, when your belief is passed on to influence others, even if you don't intend this effect, some may challenge it.
But, if you did this DBT protocol, even if you didn't use the box but did some switching boxes and found no difference, then, maybe, you could just enjoy the music without worrying about the components or their supposed faults. It is about the music, not the component, isn't it? |
| I find it interesting that many people | FLZapped
Oct 26, 2003 3:39 AM | | A sample, is a sample. Making the test time longer does not makei it more samples.
One of the problems with long-term listening is that our auditory memory is short, so then what happens is your brain starts filling in the gaps.
Short, rapidly switched(between DUT A and B) is the best method of hearing differences. The closer together you can make the switchover, the better.
-Bruce |
| I find it interesting that many people | okiemax
Oct 26, 2003 9:29 PM | | I'm not sure how you are using the word "sample." Do you mean a listening trial? If I listen one hour each day, every daily listening session is another trial, and the more I do, the longer it takes.
Rapid switching seems like a good way of hearing differences in a DBT. I'll have to plead ignorance on the subject, since I have never participared in a DBT. Would I use rapid switching in an ABX DBT to jump back and forth between A,B, and X when trying to identify X? |
| I forgot to add that -- | okiemax
Oct 26, 2003 9:44 PM | | I should have added the following sentence at the end of the first paragraph: If I decide, for example, that a total of 15 trials will be done, these 15 represent the sample. |
| I forgot to add that -- | FLZapped
Oct 27, 2003 8:39 AM | | Without seeing the whole thing layed out, I'm going to go out on a limb and say yes.
-Bruce |
| I find it interesting that many people | mtrycrafts
Oct 29, 2003 12:25 AM | | b Would I use rapid switching in an ABX DBT to jump back and forth between A,B, and X when trying to identify X?
You will switch as you see fit to try to find the differences. However, people who have done this will come to the conclusion that you switch rapidly between A and X, to see if there is a difference, then B and X or between A and B. Only the switching itself is rapid. You don't have to switch back to A, B or X rapidly.
A trial happens when you conclude an answer for a set of comparisons, regardless how many times you compare A to X, B to X or A to B. So, if you listen for an hour each day for a week and come to an answer at the end of a week, different or not, that is one trial.
You set the number of trials before the test, and correct your responses after all the trials. Just like a test in school. |
| Thanks | okiemax
Oct 29, 2003 11:10 AM | | Thanks for clearing that up. I think I understand how it works now. |
| My take on the null | mtrycrafts
Oct 24, 2003 8:57 PM | | Yes, that is what the web site states. However, I don't think I said A=B a mathematical expression. Audibly the same is not A=B, just that it is not audible to human subjects.
b I still maintain that long-term sighted listening will tell me more about an audio component than a short-term ABX DBT.
Yes, it will tell you if you like the components attributes over time not related to sound. Sighted listening regardless what time period we are talking about is still a sighted listening and is unreliable. Time alone cannot give it validity or drug testing would not be bias controlled as it is done over very long time periods, years in many/most cases.
b It just doesn't seem reasonable to me that the placebo effect would bias me long-term.
I can understand this that to you it is not reasonable. However, you are wrong though. |
| Both your links are dead(nt) | FLZapped
Oct 18, 2003 4:32 AM | | |
| Both your links are dead(nt) | okiemax
Oct 18, 2003 9:29 PM | | I apologize for the faulty links. I'll try to learn from this. |
| Thanks for the replies | mtzoanos
Oct 17, 2003 11:45 AM | | Hi,
Thank you all for your replies. I will try to bi-wire using separates runs of cable for now to see if there is an audible difference. While I really don't understand how cables exacly affect sound, I have found that sound can vary greatly across speaker and equipment manufacturers, setup, room dimensions and furnishings. I am curious to find if bi-wiring will make a difference with my setup. I'll post my findings.
Thanks,
Michael Tzoanos |
| Thanks for the replies | Bill Clinton
Oct 17, 2003 5:07 PM | | Good luck to you.
If you get a chance, I'd see if you can demo some high quality cables, though (not a cheap double run). I understand everyone's arguments here and a year ago, I was firmly on the 'how can cables make a difference' camp.
But, I had just upgraded to a new Quad setup and wasn't happy that it didn't sound as good as it did in the store. That's when I borrowed some cables and began playing around and it made my system shine. All of a sudden there was tons of detail and a step change improvement over my old system. But on this issue, it's every man for himself.
Again, good luck |
| Thanks for the replies | mtrycrafts
Oct 18, 2003 12:40 AM | | If you are after determining audible differences, you'd better employ bias controls or your outcome will be unreliable. |
| re: Where can I buy biwired cable? Is it worth it? | Thomas A
Oct 18, 2003 11:51 AM | | I have never seen any evidence of audible differences in cables, but biwiring may make som electrical difference.
E.g. a 4 x 1.5 mm2 EKK crosswired will reduce inductance and resistance by half, if the jumpers at the speaker binding post is attached.
With the speaker post jumpers detached the current will be different in the cables. There will be at least two electrical effects, all minor.
- The back-emf of the drivers need to travel through the cable instead of the jumper.
- Also when high currents run through a stranded cable, the strands tend to move closer to each other because magnetic(?) forces (correct me if am wrong here). This could in turn theoretically increase electrons to jump from one strand to the other. Separating the currents as in biwire may or may not have an effect on this.
All said, the effects may be in the -150 dB region or something. Completely inaudible.
T |
| re: Where can I buy biwired cable? Is it worth it? | Norm Strong
Oct 20, 2003 12:12 PM | | It's remarkably easy to test bi-wiring. Hook up a speaker biwired and listen to it. Now reconnect the straps between the 2 sets of binding posts on the speaker and you're monowired. It can be done almost instantaneously. You won't hear any difference unless the wires are #24 or thereabouts. (Then you'll hear a difference--in favor of monowiring.) |
| |
