Objective testing of speaker wire? Does it exist?

William Lemmerhirt

William Lemmerhirt

Audioholic Overlord
I have read posts and watched videos of people who braided Cat5e and claimed it was transformational for their systems. I'll never get that time back.
Been there done that. Some things you just have to do. Like sticking your tongue on a 9v battery!!!
 
Speedskater

Speedskater

Audioholic General
You test speaker cables objectively with a multimeter. Any method that involves your ears is subjective.
Well no. There can be objective and subjective listening tests:
Objective test. A demanding blind A/BX listening test to see if a JND (Just Noticeable Difference) can be heard.
Subjective test. A blind test of larger differences to determine a preference or to describe the differences.
 
mtrycrafts

mtrycrafts

Seriously, I have no life.
Exactly. Stats are a totally different animal and skill set and far too many ignore it and think they understand it, and while I survived bio stats and epi, what was clear was it would never be an area of expertise for me. Unfortunitluy, as I suspect you're well aware, even among the MDs and related clinical researchers, some dunning kruger effect and ego creeps in, and that's often where they get themselves into trouble. I have referreed in a journal, and always recommend having a bio statistician take a hard look at their numbers.

There does appear to be some lit on the topic, more than I realized:
Ten years of A/B/X Testing
Then, here you go, even a lot more.
http://2eyespy.tripod.com/id5.html

Some AES papers that may be peer reviewed or not like a conference paper as above link is such but still very good one.
There are some statistical papers at AES on this as well as the paper by Les Leventhal, JAES vol 34, No 6 Jun 1986, Type 1 and Type 2 Errors in statistical Analysis of Listening tests.
Others are published in magazines. Tom Nousaine was famous for going around testing claimants. No one had significant outcomes.
Then, there was an interesting test at AVS a number of years ago:
The system of the listener was custom built from foundation on up, a million dollar investment. His system: http://www.positive-feedback.com/Issue16/lavigneroom.htm
The cable cost was 5 digits against a monster 12ga.

The video is interesting but each listener had only 5 chances of guessing. Not statistically valid sample/trial size. Minimum of 10 trials with 9 correct for significance but even that is minimum. One should do at least 16 trials for each listener see how well they can guess.

You have your work cut out for you. ;)
I may not be here to see the end. :)
 
Will Brink

Will Brink

Audioholic
Then, here you go, even a lot more.
http://2eyespy.tripod.com/id5.html

Some AES papers that may be peer reviewed or not like a conference paper as above link is such but still very good one.
There are some statistical papers at AES on this as well as the paper by Les Leventhal, JAES vol 34, No 6 Jun 1986, Type 1 and Type 2 Errors in statistical Analysis of Listening tests.
Others are published in magazines. Tom Nousaine was famous for going around testing claimants. No one had significant outcomes.
Then, there was an interesting test at AVS a number of years ago:
The system of the listener was custom built from foundation on up, a million dollar investment. His system: http://www.positive-feedback.com/Issue16/lavigneroom.htm
The cable cost was 5 digits against a monster 12ga.

The video is interesting but each listener had only 5 chances of guessing. Not statistically valid sample/trial size. Minimum of 10 trials with 9 correct for significance but even that is minimum. One should do at least 16 trials for each listener see how well they can guess.

You have your work cut out for you. ;)
I may not be here to see the end. :)
Great info, thanx! So super simple, and n = 1 test, but take that one person who swears there's a difference, flip a coin, change speaker wire (using their choice if super $ wire to make it "fair" vs say bluejean stuff) using the same track, and run it 20 times or more? How many times with 1 person does it reach stat sig for an acceptable CI it's not due to chance they can or can't identify their wonder cable from say Blue Jeans stuff? 20, 50, 100 times?

Obviously a hugly over simplistic test, but say you wanna put some cable/wire pushing derpa derp in their corner and tell them to put their $ where their mouth is on a challenge?

So, not the scientifically valid test protocol but the get that idiot to STFU protocol you challenge them with protocol.

I have done similar in my own industry with a few scam products (e.g, T boosters and other nonsense) where I offered to put up 10k they had to match, test their product, winner takes all. Of course none to date taken up on it, and I'm not a bluffer and would not offer that kinda $ if was not 100% sure I'd win.

What would be the audio version of that? I'd really like Gene to be on that as the person who would run/referee such a test. ;)
 
mtrycrafts

mtrycrafts

Seriously, I have no life.
...

What would be the audio version of that? I'd really like Gene to be on that as the person who would run/referee such a test. ;)
While Tom Nousain was alive, I suggested Gene to contact him and perform some DBT protoco.
Tom had an ABX box for rapid switching of tested products, best for perceiving small differences, but was not interested. So, I don't have much hope this will happen.

By the way, you do not have to run the same audio track repeatedly. Any track that is agreed upon.
9 of 10 trial meets the p.95, so does 12 of 16. If the person can replicate such significant results then there may be something, unless there was a tell in the protocol.
 
Will Brink

Will Brink

Audioholic
While Tom Nousain was alive, I suggested Gene to contact him and perform some DBT protoco.
Tom had an ABX box for rapid switching of tested products, best for perceiving small differences, but was not interested. So, I don't have much hope this will happen.

By the way, you do not have to run the same audio track repeatedly. Any track that is agreed upon.
9 of 10 trial meets the p.95, so does 12 of 16. If the person can replicate such significant results then there may be something, unless there was a tell in the protocol.
To add a layer to that, maybe run the 10 trial 3 times in a row with some time between them. That gives 3 seperate trials but without obvious excuses of listening fatigue. Without the ABX box coin flip may have to do, but time between cables would be an issue there.
 
mtrycrafts

mtrycrafts

Seriously, I have no life.
To add a layer to that, maybe run the 10 trial 3 times in a row with some time between them. That gives 3 seperate trials but without obvious excuses of listening fatigue. Without the ABX box coin flip may have to do, but time between cables would be an issue there.
Yes, that is possible but there is no reason not to break after any number of completed trials to rest, recuperate and reflect upon what just happened. :) No need to do all x number at once.
Yes, swapping cables manually is a killer timewise even though many feel it is not an issue as that is how they hear differences in a sighted test, manual swaps of who knows how long.
Memory for small differences are long gone by then.

DBT is a must in any event.
 
mtrycrafts

mtrycrafts

Seriously, I have no life.
And, if a break is taken after 5 trials, lets say, don't let the administrator or rather the person who is keeping the answer log reveal anything until all the agreed upon trials are finished. So, you can run that 30 trial test as one test with all the needed breaks.
There has been month long DBTs when the ABX box was used as the answer is locked up inside the memory until all finished. Legitimate.
 
mazersteven

mazersteven

Audioholic Warlord
And, if a break is taken after 5 trials, lets say, don't let the administrator or rather the person who is keeping the answer log reveal anything until all the agreed upon trials are finished. So, you can run that 30 trial test as one test with all the needed breaks.
There has been month long DBTs when the ABX box was used as the answer is locked up inside the memory until all finished. Legitimate.
Could always hire Tom Bergeron. He's looking for work. I'm sure they could get him cheap.
120316tom-bergeron1.jpg
 
lovinthehd

lovinthehd

Audioholic Jedi
Because a bunch of relatively loony consumers and those willing to feed off them, it becomes needed to do wire listening tests to fend them off? Really?
 
highfigh

highfigh

Seriously, I have no life.
Obviously a hugly over simplistic test, but say you wanna put some cable/wire pushing derpa derp in their corner and tell them to put their $ where their mouth is on a challenge?

So, not the scientifically valid test protocol but the get that idiot to STFU protocol you challenge them with protocol.
I know a sales rep for audio equipment who has heavy speaker cables with cable stands, some kind of black boxes, heavy power cords, a really stupid record cleaner with twirling paint rollers (literally, the 1" diameter rollers with a colored, spiral thread) for drying the LPs and some nice equipment. In his last house, which is Mid Century Modern, the back wall was random stone (good for diffusion), the back wall was at an odd angle, the left wall was drywall and the right was floor to ceiling windows. Apparently, someone told him that the right wall was too reflective and that he needed to use insulating panels, so he bought a couple of pink foam panels and hid them behind the thin curtains. When I listened to the second set of speakers, the left wall was reflecting badly, so I put a pillow on the top of the sofa backrest and that helped, but the room still needed help.

Every time I would dump on some kind of audio hocus pocus, he would just say "It's experiential". We would change speakers on occasion (he lost one job and had to return his demo speakers), went to another company and put his new demo speakers in the same place. He was surprised when I told him that each design needs to be in its own best position.........

Also told me that the new power cords reduced amplifier warm up time to about twenty minutes, from two hours. Used to be that sales reps had some level of product and technical knowledge but that ship has sailed.
 
highfigh

highfigh

Seriously, I have no life.
If any of you have trouble getting your head around what I mean by a good positive control, here's an example. Once, at a DIY speaker meeting, a guy proudly presented his 2-way speakers that were mainly of his own design. The drivers were very expensive, a SEAS Excel 6½" mid-woofer and SEAS Excel Milenium tweeter. He also told us, proudly, how he used only teflon-insulated silver wires, silver solder, very expensive capacitors, and expensive flat-wire inductor coils within the cabinets. He assured us that all those materials significantly improved the sound of the speaker.

When he demo'ed the speakers, someone (Dennis Murphy) politely suggested that he might have mistakenly wired the tweeter with polarity reversed from the intended design. The builder doubted that. Later that same day, the speakers were measured acoustically, and they were indeed mistakenly wired. Those acoustic measurements looked similar to this other 2-way design:

Wired with the correct polarity
View attachment 39863

And with the tweeter with reversed polarity
View attachment 39864

The difference looks pretty large on these graphs, but at roughly 2.5 kHz, were surprisingly easy to overlook. The take home lesson: If the builder could not hear the difference between the correct and incorrect wiring polarities, what can we make of his claim about the improved sound qualities of the silver wires and other expensive crossover components? They must have been much larger than the suckout at 2.5 kHz. That is a positive control – in this case, an unintended positive control. The guy's understandable late night mistake undermined his claims about silver wires and exotic crossover components.
I know someone who had a recording studio and at the time, I worked for a car audio dealer that had an Audio Control RTA. He called to ask if it would be possible for me to bring it to the studio because he had a problem with the sound that he couldn't eliminate, so I did. I had him play something through the monitors and told him that he had a phase cancellation in the mids, but he couldn't think of a reason for that, so we fired up the RTA. We saw a deep notch in the crossover region, so I asked if they had replaced any of the drivers and he said that his partner in the studio had replaced a horn diaphragm. He checked it and sure enough, the wires were reversed.
 
newsletter

  • RBHsound.com
  • BlueJeansCable.com
  • SVS Sound Subwoofers
  • Experience the Martin Logan Montis
Top