The Great Interconnect Cable Test 1 (1 Viewer)

[Brian_C]

Over the weekend, my friends and I completed a series of blind tests. We were curious as to whether or not expensive cables made that much of a noticeable difference. Posted below are the results from our testing. To sum everything up, we did NOT find a difference with different cables. If you refuse to believe we could have come up with that result, do not read anymore.
The results below are FACTS, and the conclusions I draw are my opinions based on the facts. I am not trying to start a flame war, so please don't do it. This weekend was very educational for me, as I hope you all will find the information below interesting. Multiple tests were conducted and they are all summarized below.
There were 7 people present, and I would consider in this group there are 2 audiophiles, 1 enthusiast (me), and 4 casual listeners. All tests had at least the 2 audiophiles and myself present.
We were at my friend's place, so I'm not too sure of his equipment. He had a Sony CD Player, a NAD CD Player, NAD Preamp/Amp, Energy E3 (or something like that) speakers. Cables coming from the amp to the speaker costed $12 a foot, so were pretty decent. Interconnects were the cables tested. I can get exact model/makes if anyone is really interested. He had a bunch of other equipment, but they weren't used so I won't list them.
The CDR Test
It all started from a friend that claimed he was able to hear the difference from an original cd versus a copied CDR. So the bet was on and a date was set. We decided to conduct other tests as well since we were going to be there.
Five original CDs known to the listener were brought, and 1 song from each disc was copied to a CDR with Nero software. The listener will be allowed to listen to all tracks, and then all copied tracks prior to testing. For testing, the original track will be played, and then either the same original or copied track will be randomly played. The listener must pick whether the second track played was the original or the copy. The listener must be correct at least 8 out of 10 tries.
So before testing we played the originals and the copies. I thought I was going to lose. Almost everyone in the room could hear the difference. Two out of the five songs sounded very obvious they were different.
So testing began. We tried as best as possible to make the test double blind. An original track was played, and a coin was flipped to determine whether the original or copied track was to be played. After 4 runs, we stopped because the listener had MISSED 3 out of 4 guesses. Two of those were songs we thought were obviously different.
The CD Player Test
My audiophile friend wanted to know whether he wasted his money or not on the NAD Cd Player so we did a blind test on this. So before testing we played tracks on both players and the NAD sounded a bit better. More clarity and sharpness. So blind testing started and we all could hear the difference again with the NAD showing more clarity and sharpness.
The Interconnect Cable Test
Two sets of cables were tested at once. Each set was connected to one of the two sets of outputs from the CD Player. This way, the Preamp could switch between the two by pressing only a button.
The cables tested are listed below:
$300 THX Certified cable (don't know the brand/make) 5'
$3 El Cheapo cables 3'
Custom made RG6 cables 5'
Custom made Category 5 cables 15' (not a typo, 15 feet)
So the $300 cable and the El Cheapo cable were played for reference. Two out of three of us could not hear any difference, the third (the owner of the cable) heard a slightly clearer treble and better defined bass. So the testing began.
For about 2 hours we did many tests on different songs and cable matchups. There were up to five people in the room and everyone came up with the same answer. No differences in sound, not even the 15' Cat5 cable! We also switched speaker cables to be sure the speaker cables aren't the limiting factor, and again no differences were heard.
I can post pictures of the custom cables if anyone wants to see them. I forgot them at my friends place, so it will take a few days for me to get them back.
MY Conclusions
x I absolutely do not trust non-blind tests anymore. I have witnessed first hand how much of a difference it makes when people know before hand which one they are listening to.
x More expensive components generally yield better sound, but the difference is only slight and you generally lose in the bang-for-your-buck department as prices go up.
x Expensive cables didn't make a difference in my friend's setup and I doubt they will in mine. YMMV. Even if they did make a difference, it is so slight that the cost cannot justify it unless you have money to burn.
x I'm going to keep on using and making my own custom cables. They look nice.

 

[Bob McElfresh]

Brian, sounds like a fun weekend.

But let me ask you a question: When you compared the 2 sets of interconnects doing A/B, was there any change in volume?

 

[Lee Scoggins]

Brian,

This is great! I like seeing folks do actual listening tests and going with their own gut evaluations.

Now here are the problems:

1. You used non-audiophiles in your test. It has been shown in respected science journal that people can improve their ability to listen with practice. As a result, we get "golden eared" audio reviewers who can more clearly delineate differences in tempo, tone, texture, soundstage, midrange purity, etc. The composition of your group probably influenced the results by mixing listening abilities.

2. While the system in question is very good, it is likely not high resolution enough to properly accentuate the cable differences.

3. As Bob pointed out, level matching is critical.

In any event, your test provided further evidence that expensive cables can make a difference. On my home system, even a power cord change can be dramatic. Some friends of mine will choose certain speaker wire and interconnects to balance a "brightness or dullness" in the rest of their system. It seems wild that cable can have such an influence but it does.

Lee

 

[Brian_C]

Bob: There appeared to be no differences in volume.
Lee: In terms of 1 and 2, I think the vast majority of us will fit in that group. I wish I had very high end equipment for further tests, but I don't. Most likely I'll never own equipment of that caliber, but there's those of you who do... (hint, hint)
After the results of our testing, it's going to take a lot for me to think power cables will mean something. Perhaps cable test 2...

 

[Bob McElfresh]

Brian: Something is strange.

You used 5 different sets of interconnects, including some home-made ones.

Audio interconnects can be made with different impedence coax (and the RCA plug has an effect). When you switch from a 50 ohm coax (the expensive THX guy) to a 75 ohm coax (the RG6), there should be a bit of a reduction in overall volume. You found none.

And that home-made interconnect. Somehow you managed to make that cable exactly match the impedence of all the others? Enough so there was no volume change?

Dont take this the wrong way, but are you sure you were switching between the cables? Did you by any chance have a optical connection hooked up as well and the pre-processor auto-selected the optical feed?

I'm not a fan of expensive copper so I'm not trying to refute your findings. But...having all 5 cables sound identical without even a overall volume shift as you went A/B/A... that sounds a little funny.

 

[Lee Scoggins]

Brian,

Good point. I believe most would fit in the first or second category. In rereading my post, it appears a bit elitist although that was not my intention. I want to improve audio for everyone at every price point. I believe that many HT buffs buy equipment by brand and often overpay when better stuff for same or less money is available at "audiophile run" independent shops.

You see I feel strongly about cable quality having worked as a recording engineer and seen day in-day-out how important this is and how many skimp on this in even the best recording studios. I point out the three points as some things that would enhance the test on the next go around.

I truly believe that you could hear a difference with the right set-up. You may also want to use very pure recordings with a deep soundstage (3D instruments, etc.) Try XRCD2s or audiophile labels like APO ot Chesky and listen to the same tracks over and over.

As far as custom cables go, there are some excellent ones out there, but reliability is all over the map by my own experience. Termination and quality of connectors is very important. George Cardas makes probably the best connectors in the world; see if you can have some made with his components.

Finally, use a contact cleaner/conditioner on your stereo connections. You won't believe what a difference this can make.

In any event, it's all about the music. Have fun!

Lee

 

[Brian_C]

Bob, I absolutely know what you're saying. I was rather disappointed myself when no one could distinguish the 15 foot cat 5 rca cable. This is a cable I put together myself so that I'd have a control test to be able to tell the difference. I intensionally made it long to increase the chances for variance.

We did make sure we were making valid tests as we disconnected one set of cables and attempted to switch to it, resulting in no sound. Repeated with the other set, verifying we were hearing signals from those sets of cable.

I'm not too sure on this, so don't quote me, but I read some where that the impediance in terms of a line level signal means nothing and the only figure that may make a difference is the DC resistance of the cable. That's why it's ok to use video cable (RG6) as audio cables.

 

[Bob McElfresh]

I'd agree that at audio frequencies, impedence is more of an issue for the speaker wire (power) than the line-level signals. (At video frequencies, it's an issue for line-levels).

Strange, but I'm glad you brought over friends and had a fun day of playing with the equipment.

So whats next, Video cables?

 

[Chu Gai]

L. Scoggins stated:

1. You used non-audiophiles in your test. It has been shown in respected science journal that people can improve their ability to listen with practice. As a result, we get "golden eared" audio reviewers who can more clearly delineate differences in tempo, tone, texture, soundstage, midrange purity, etc. The composition of your group probably influenced the results by mixing listening abilities.

Listening tests in question (recently performed by Wright of Northwestern) do indicate people can be trained trained to improve their hearing, but dealt with single tones not musical passages. This involved people listening through headphones in a room relatively free from any sort of distractions. The thinking that this study lends credence to the concept of "golden ears" is false and is not supported either by the research or the researcher.
While the study performed by Brian is interesting, I would welcome further clarification of the entire process. To only perform 4 tests has neglible statistical significance and introduces an enormous amount of error into the results such that conclusions are difficult to determine.
As far as 'sufficiently revealing' systems, I throw out the question, what constitutes such a system? In what sort of environment should it be assembled and what then should be the auditory criteria (after all, as we get older, our hearing ability generally decreases especially with higher frequencies) to be able to be considered a suitable candidate to participate in such a test? Personally, I find Brian's mix of participants reasonable.
Along the lines of interconnect testing, there was a test (details not forthcoming) that listed the results and can be found at Link Removed
Bob is quite correct on the importance of volume matching as generally people prefer that which is louder.

 

[DanKaps]

I'd be happy to offer up some cables to be used in another cable "Pepsi Challenge".
Dan Kaps
www.heartlandcables.com

 

[Brian_C]

Chu: I'm assuming you're referring to our CDR tests where we only went 4 trials. Originally, the wager was that the listener must correctly identify 16 out of 20 tries in order to win, but this was reduced to 8 out of 10 in order to leave time for the rest of the tests. So out of 10, the listener was only allowed to miss 2 times.
Since the listener missed 3 out of the first four trials, it became statistically impossible for him to win, thus we ended the test early and got our pizzas.

 

[Chu Gai]

well would've been nice to see the trial carried out all the way, but pizza can become compelling.

 

[Earl Simpson]

ummmm, Pizza, chocolate!

 

[Lee Scoggins]

Chu,

I know of several people who can hear more deeply into tapes from the recording studio. Many musicians will agree that their training often gives them an edge for hearing small details in the music. Whether or not there have been formal, academic studies on music is irrelevent to me in the face of such experience.

As far as a high resolution system goes, I am thinking of an audiophile system that can accurately portray the midrange and high frequencies which have a lot to do with evaluating cable.

Lee

 

[Ian Montgomerie]

Even worse is when people claim that they can hear the difference between different kinds of digital cable, where the supposed difference is "the bass is better defined" or something like that.

Just goes to show you how much peoples' expectations and so forth can lead to odd perceptions.

It's funny because of how a digital signal works. Basically the whole point with digital is that there is no such thing as one cable leading to better sound than another. A signal that is, electrically speaking, "mediocre" still carries exactly the same information as a perfect, powerful signal. All that matters is that when it gets to the other end, the receiver can differentiate the range that means "1" from the range that means "0". (And even if it can't perfectly differentiate, enter error-correcting codes which will recover perfectly from small errors).

And if there actually is a problem with the signal, you won't hear it as a bit inferior quality, it will be unmistakable as nasty static, audio drops, or pops and clicks. In encoded audio (DTS/Dolby Digital for example), if there is any unrecoverable error you will lose the entire packet of data, in which case the decoder will output complete silence for a short period.

Lesson: if you're buying a digital cable, go coax and get a cheap RCA video cable. The chance of the sound being one whit worse than that of the most expensive optical cable you can buy is nil (unless the cable is defective or you have some pretty shocking interference, in which case you will KNOW). That's the whole point of digital.

Same thing goes for CDs versus CD-Rs. We are talking about digital technology with error correction. Unless your CD player has problems physically reading the CD-R media or your burner is highly defective (which could cause the occasional error, again NOT manifested as a general difference in quality but as occasional garbage), there is no possible difference.

 

[Phil A]

On the transmisson of a digital signal, the actual transmission is not '1s' and '0s,' but analog. What is transmitted is the signal and the clock/timing. In any test, besides the fact that some equipment may be more sensitive to others, the source material is important as well to discern what differences, if any, there are.

 

[Jim A. Banville]

I know of several people who can hear more deeply into tapes from the recording studio.

They claim to be able to hear more deeply...yada..yada...

 

[Lee Scoggins]

Ian,

If only life were that simple and bits were bits...

Alas, those of us in the high end recording studio know that digital cables can make a difference. All sorts of things impact a digital signal including jitter which is time-based distortion, how the connections work well together, etc.

Yes, I have heard differences with digital cable in A/B tests on high resolution systems. It does seem odd but it is true. One would be forgiven for thinking that 0's and 1's are all that matter.

Lee

 

[Ian Montgomerie]

I'm sorry, it doesn't really matter what you think you hear, there is no possible difference between digital cables unless one of them is so bad that it produces data errors that cannot be corrected. The fact that what's "really" being transmitted is an analog signal doesn't matter - the whole point of digital is to set it up so that a wide class of analog signals is interpreted as the same information. Digital data transmission errors do not manifest themselves as subtle differences in quality, they manifest themselves as point errors (in PCM) or corruption of entire packets (in encoded audio).

In DTS and Dolby, for example, it is quite impossible to have an error that results in anything but one or more packets' worth of silence, which is quite noticable as such. Unless whoever designed your decoder had no idea what they were doing (unlikely that they would pass certification in that case).

In uncompressed PCM (like digital output when playing a CD), it is possible to have bit errors which affect only a single sample, but the random characteristic of this error will mean it expresses itself as noise on the signal, with occasional transients as a high-amplitude spike appears in the middle of a quiet passage, and so on. If there are no bit errors, there is no difference (and your average el-cheapo cable will usually not have bit errors without major interference, cruel and unusual punishment to the cable, etc).

Oh, obligatory background: I write system software for LSI Logic Broadband Entertainment Division (formerly C-Cube Microsystems), specifically for the Ziva5 DVD processor, and I spent most of my time last year working on the audio datapath.

 

[Kiet_H]

Lee Wrote:

1. You used non-audiophiles in your test. It has been shown in respected science journal that people can improve their ability to listen with practice. As a result, we get "golden eared" audio reviewers who can more clearly delineate differences in tempo, tone, texture, soundstage, midrange purity, etc. The composition of your group probably influenced the results by mixing listening abilities.

2. While the system in question is very good, it is likely not high resolution enough to properly accentuate the cable differences.

3. As Bob pointed out, level matching is critical.

In any event, your test provided further evidence that expensive cables can make a difference. On my home system, even a power cord change can be dramatic. Some friends of mine will choose certain speaker wire and interconnects to balance a "brightness or dullness" in the rest of their system. It seems wild that cable can have such an influence but it does.

Lee



Lee,

The test was performed at my home and I will attempt to answer your questions below.

1: Although I do not have dog ears, I and a friend who was present for the test do consider myelf audiophiles and yes I have trained my ears for many years to more clearly delineate differences in tempo, tone, texture, soundstage, midrange purity, etc. Yet I heard no difference.

2: Yes I agree the system that is used for the test is no where near high-end. It consisted of:

NAD c541 cdplayer

NAD 304 Int Amp

Pair of Energy C2's.

Reference interconnect cables were XLO/ Pro 100 for

Reference interconnect Speaker cables were AudioQuest

Type 6

It is actually my second system and not my main system. My main system consists of

Rotel RB-1080

Parasound P/SP 1500

Parasound P/DD 1550

NAD C541

Energy C6 Speakers

Yes my main system (although still not High-End) should be sufficient "enough to properly accentuate the cable differences. " The reason it was not used is because it was inside my Entertainment center and I did not want to take 4 hours to remove it from the entertainment center and put it back after the test.

Note to Brian.. Looks like we have to redo the test with my Main system.

3: Yes level mathcing is critical, and should produce mroe distinct differences when not matched, but we DID NOT hear any difference what so ever.