QUOTE(alive @ Aug 25 2005, 01:16 PM)
The danish shop, Hi-Fi Klubben (Site also available in Swedish and Norwegian languages) sells speaker cables for a price of up to €2681 for two meters.
This is nothing. Here are some of the highest prices :
Transparent Opus MM speaker cable. Lenght : 8'. Price : 23,500 $
http://gallery.consumerreview.com/audio/ga...res/opus-mm.jpg
ISM BI-WIRE The One speaker cable : 3 meters, price : 16,720 $
SILTECH G5 Ruby Mountain. Power cord. Price : 3,299 euros.
QUOTE(henkersmahlzeit @ Aug 25 2005, 04:09 PM)
Do I miss something her?? If people who claim to hear a difference fail the ABX-test this should have no meaning? This is not a esoteric discussion!
Of course it means something. It means that the guy was wrong. But the real meaning must be appreciated in light of the test setup.
QUOTE(Nero @ Aug 25 2005, 07:00 PM)
This is not just a project. This is the full acount of the ABX test, with raw results and all. It failed.
QUOTE(Nero @ Aug 25 2005, 07:00 PM)
These are examples of people who (at least apparently) used double-blind ABX testing to show that there is no (or not a substantial) audible difference between inexpensive cables and the "high-end" stuff. Seems to me the onus would now be on anyone who claims there is a difference to show it with the same type of test.
No, the power cord test that you link was made by voodoo audiophiles in order to prove that power cords do affect the sound. The test conducer, Jason Victor Serinus still claims that power cords are very important, after having himself utterly failed the ABX test.
John Atkinson also claims that very expensive power amplifiers are worth their price after having failed to ABX them against cheap ones. And Rorominator (Toulouse test below) still claims that interconnect cables are important after having failed to AB them (simplified ABX, the listener had to tell wether the cables were AB or BA).
QUOTE(Nero @ Aug 25 2005, 05:45 PM)
There actually WAS a thread that referred to ABX results for audio cables here sometime last year. IIRC, the test wasn't conducted by anyone here, but the test as it was described seemed to have been conducted properly. I'll look for the thread. (Was it Pio that posted the link to the test, perhaps?)
6th topic in the first page of our "listening test" forum
It was in may 2005 and it was conduced by me.
Two voodoo cable audiophiles (Patrice, Marsupilami), one open minded (JC07), and two skeptical guys (Emmanuel Piat and Pio2001). A high end system, a small room. Interconnect cables from 2.30 € to 530.00 €.
We all kept describing differences when the randomization made us listen to twice the same cable. We sometimes thought that the cables were identical when they were different.
At one point, two cables were played at random.
I said "they are the same".
Marsupilami "They are different. I prefer the notes extinction with the second one. There is more volume."
JC07 "They are different, the first one has a more balanced image, the second one leans to the right side".
In fact, they were different. The first one was a TaraLabs (530 €, left hand side on the photo), the second one was made of more than 6 meters of standard cable with an additionnal connection in the middle (1.5 meters 2.30 € standard cable + 5 meters standard extension).
We ended the training session whithout having been able to recognize reliably if the cables were changed or not between two consecutive playbacks.
I took the time to measure the frequency response of the cables with my CD Player as a source. RMAA detected a 0.01 dB roll off at 20 kHz with the standard cable, and about 0.02 dB with the 5 meters extension in addition.
One month later, Rorominator, in Toulouse, France, organized another test. He had closely followed the development of the first one, and proposed to test his own cables in order to overcome two problems with the first test. First improvement : the test was run at home, on his own system, with his own cables, ands his own CDs. Thus the listener was not perturbated by the discovery of an unknown system in an unknown room with unknown CDs. Second improvement, he chose two cables that he knew he could distinguish immediately. A standard one, and an Ecosse one, worth 1,700 €.
The test involved 5 people who gave more or less answers. I was in charge of the maths, and it was a challenge for me to calculate a relevant p value with a number of listeners unknown in advance, with 5 to 20 answers !
Every session consisted in the two cables played in a random order. AB, or BA. The listeners had to tell the right order. This setup was decided after JV Serinus power cords test, and our interconnect cable test, so as to reduce the time decision interval to a minimum (one cable switching), in order to avoid auditive memory problems.
Rorominator got 9 right answers out of 20. Worse than random. However, PhoenixPrt got 10/10. Taking into account the presence of 4 other listeners, this lead to a p value of 0.005, which is significant. Several days later, after having examined the raw results, we discovered a mistake : Phoenix_Prt score was not 10/10, but 8/10, which leads to a p value of 0.25, which is not significant.
It is worth mentionning that 4 people out of 5 claimed to hear the difference when the cables were presented to them just before the test, while in our first test, only Patrice was convinced that we would succeed.
What's interesting in these tests is not the fact that they failed, but the fact that everyone gave wrong answers. This is especially the case in Jason Victor Serinus power cord test, where people rated the "differences heard during the test" from 2 to 4 on a 1-5 scale before they were told that they actually failed ! One of them even said that the test was too easy, because the sound was louder with one cable than with the other ("SPL should have been matched [...] the difference in volume made the cable identifiable") ! His actual result was 5/10.
Here are all the links to the french test discussions : http://www.homecinema-fr.com/forum/viewtop...r=asc&start=255
And let me recall also this ABX success between speaker cables : http://www.hydrogenaudio.org/forums/index....ndpost&p=148461
The frequency analysis showed a 2 dB roll off above 10 kHz !
QUOTE(Lyx @ Aug 25 2005, 07:13 PM)
Anyways, in some of those cases, it may be possible to test transparency without needing a reference-unit. Wouldn't this be possible in the case of amps, by simply picking a recording, then putting it through an amplifter, then recording the output someway and afterwards compare the source and target?
This article says that this have been done by Baxandall in 1977, and Belcher in 1985, proving that a decent power amplifier could be completely transparent to the human ear without needing exotic audiophile conception : http://www.dself.dsl.pipex.com/ampins/pseudo/subjectv.htm
QUOTE(alive @ Aug 25 2005, 06:26 PM)
1: It seems, that no-one can actually disprove that "better" cables give better sound.
We just have to apply the Baxandall cancellation technique in order to prove it.
QUOTE(alive @ Aug 25 2005, 06:26 PM)
2: To increase the probability of the claim that better/more expensive cables increase audible sound transparency is a fallicious claim, a pro-cablevoodo group of people would have to conduct a scientific blind ABX test.
This was done in the power cord and second interconnect blind tests above. In the first interconnect test, Patrice only listened once.
QUOTE(alive @ Aug 25 2005, 06:26 PM)
3: Such a test would require extremely high end audio equipment, as to make the test more convincing.
This was the case in the power cord test and in the first interconnect test at least, exept the CD player of the interconnect test, that was only 1500 €. The second interconnect test was done on a lesser equipment, but still high end enough.
