Following up...
I've debated JD many times on rahe - and if Deja.com still worked, I'd suggest that you check out those goings on...
In brief - JD does some nice work, but fall short of being able to make a *definitive* statement as to what is audible and what is not. I believe that he is wrong in his conclusions only because his tests do not reach far enough.
See below.
John Dunlavy On "Cable Nonsense"
2. John's philosophy of cable design is to insure that "loudspeaker cables possess
all the electrical properties required to ensure that no audible degradation of
complex musical waveforms and transients can occur between the output
terminals of a power-amp, and the input terminals of a loudspeaker." (from his
website)
Yeah, sure. The problem is the term "audible degredation" - this is problematic. The limiting factor(s) in determining audible degredation are 1) YOUR EARS/BRAIN & 2)THE ULTIMATE QUALITY OF THE SYSTEM. Unfortunately, both represent real and measurable compromises (even JD's speakers). So, just because the "literature" has not yet published tests that have found "differences" it does not mean that they are not there. The published tests are flawed IMHO, which is why the results are what they are.
3. John claims he can always 100% of the time hear the difference in sighted test
and never in blind test.
Then he is not hearing the difference at all. Few if any men of the age that John is CAN hear particularly well. :- (
That's bad for all of us. When I was 16 I could easilly hear the TV horizontal freq AND ultrasonic motion detectors (some of them). No more.
4. He also claims that no one has ever figured out which cable was playing when
he performed his blind test in his office.
He probably only used musical selections to do the tests - this makes zeroing in on the "differences" VERY difficult in the SHORT TERM. The real test of any system is simply stated making it *easier* for your brain to 'figure out' what it is hearing (less internal processing to extract and decode). You CAN listen to a 2" TV speaker and still hear music and recognize voices, right? There are other test signals that are much better suited to spotting instantaneous differences, oddly these "objectivists" and "scientists" never seem to manage to find or use them (as far as I am aware).
5. Never has he said "no one can hear a difference"
But manages to bash anyone who thinks they do?
_-_-bearlabs (bearlabsUSA.com) |
So You Want To Argue.
Dunlavy's argument is that people can not distinguish between cables better than
chance IN BLIND TESTS. His support is the result of many tests conducted at his
facilities.
Right, and this is the problem - YOU CAN NOT GENERALIZE from one test situation and then declare that "A" is the truth based upon these results. This is what John *tends* to imply by his posts and writings - when questioned closely he either avoids the issue(s) or has to agree.
If you want to argue against Dunlavy, prove that people can sucsesfuly
distinguish between cables IN A BLIND TEST, or that his sample was not
statisticaly strong enough to make the generalization.
I do not have to do EITHER to question his CONCLUSIONS. It is sufficient to say that his test methodology is flawed and so the results that stem from it are equally flawed. They are statistically valid, but this comes from FLAWED tests. PERIOD.
Mentioning that you can discern differences under other circumstances is
TOTALLY IRRELEVANT. That is what he claims is the placebo effect which he is
so frustrated by.
I don't want to slam JD or his products, but let me say again that the LIMITING FACTORs in such tests are A) the listener's hearing ability, B) the system. I would suggest that at least in the tests published so far that item (B) is extremely questionable. Let me add that (C) the source material is also a limiting factor. NO DIFFERENCES CAN BE HEARD if any ONE or ALL of the limits are reached.
And even if under his circumstances YOU can, that does not refute the claim, you
would also need a reasonably statisticaly valid sample of people to perform the
same way.
Statistics are only as good as the TEST that underlies the statistics. There's little doubt that one can design tests that will give the appearance of being statistically valid for almost any result.
Think about the problem more fully.
_-_-bear (http://bearlabsUSA.com)
|
Most cable manufacturers are not wire manufacturers at all. It requires a significant capital investment to set up and operate a wire factory. The barrier for YOU to become a wire manufacturer is mostly the need to run the damn thing 24 hrs a day to make a profit and cover your overheads! So, unless you can sell MILLIONS upon MILLIONS of feet of wire, you're not going to be in any sort of wire business. MANY of the domestic wire operations have folded in the last several decades because it's tough to sell that much wire.
So, it is necessary for audio cable manufacturers to sub contract out their needs to one or more existing factories - their runs are insignificant compared to a typical industrial order - THUS, the prices are higher.
I have my pure Silver stranded wire drawn, stranded, and jacketed with PTFE ("teflon") for me by a small specialty manufacturer, I assemble it into what becomes Silver Lightning Interconnects. It is very expensive per foot because of the small quantities involved. By the time I'm done with the labor involved in the assembly and add in the costs of overhead and the materials, believe me there isn't a ton of profit. You can't look only at the cost of materials when you think about these cables.
I agree, some are junk, snake oil, and WAY overpriced. But not all. |
No way to quote an earlier post...
Unfortunately there's no way to *quote* a previous post
in order to respond - like you can do on a usenet newsgroup by hittin the "reply" button... Grungle, I was responding to
your post and copied parts of it in order to make point by point responses... sorry if it caused confusion...
plus it looks like the software quotes the first few words of your post to make a title...thus my first line above!
:- ) |
JHunter wrote...
-->
Bear is up to his usual "debating" tactics. If decades+ of established acoustics
methodology and all the evidence (from scientifically valid testing) goes against
you, then just claim that the testers don't have a system with sufficient
resolution. Could you give us an example of a system that you feel does have
sufficient resolution? <--
I'm not sure what you mean, since I am relatively new here on Audiogon...
But, the published tests quote the systems that are used for the "testing." so it is fairly simple and easy to determine what level of resolution they are capable of.
As I have stated earlier, the limiting factors are: A) your hearing, B) the system and C) the source.
I have also said that it is far easier to hear very subtle
changes on an instantaneous basis when listening to PINK NOISE, as compared to *any* musical source. None of these tests involved any Pink noise.
HOWEVER, what counts in terms of long term listening is how much 'internal brain processing' is required for your concious to figure out what it is hearing! That is the difference between systems - nothing more.
It does NOT require a high-end system for you to recognize speech - a telephone is good enough. You can listen to a tiny 2" TV speaker and understand what is going on. Right?
So, the point that I make is that thus far the systems used for these tests are at minimum *questionable* in terms of ultimate quality and resolution, AND the source material is also questionable. SO, the conclusions drawn are valid ONLY for the specific TESTING that was done, nothing more.
JUST to exagerrate for clarity, IF the cable 'tests' were done with 2" TV speakers, it is very unlikely that anyone could possibly hear any differences, right?
This is clear.
As far as a "system" that I think has sufficient resolution, there are all sorts of candidates that I think would likely do the job. But, there is little point in quoting a list of components, as that is NOT the point at all. Needless to say, IMHO, none of these candidates were utilized in said published tests.
The point is to truly understand the limitations of published tests, and what they mean in reality. |
... internal processing...
By internal processing I do not mean that familiarity leads to recognition - although there is an element of "ear training" that plays a role in how we all hear. The question is not familliarity as much as how much internal processing you are actually doing to have your mind say "ahhh I know what that is/means."
An example of this *sort* of thing is when you randomly tune your car radio to the middle of a song, one that you hear all the time, and it takes a few seconds to figure out which song it is, as you have tuned into a spot that is perhaps in the middle of a musical phrase, so you don't have the benefit of a "start" point (musically or lyrically) to guide you.
What I am speaking of is similar, but more subtle in that the clues that tell you where/what on a hi-fi system, and really give you the detailed tonal and spatial information are much more minute.
Perhaps one of the reasons that you can "hear" more on your own system is that you have given your brain an "algorithm" with which to quickly process the raw sound into intellegable information.
So, the test then is to be able to walk in "cold" (no pre-programming of your brain, just "natural" sound algorithms) and be able to *instantly* recognize all the sonic clues! Of course, this is not a constant, nor is it (as far as I know) measurable (yet), but if you think about what happens when you are out in the world everyday, or at a non amplified live performance, you don't have to strain or work at all, it all just *IS* <-- kind of a zen thing.
Hope that explains...
_-_-bear |
JHunter asked...
- are all the systems used for published DBTs inadequate?
Yes, IMHO, based upon those articles that I have read (which may or may not be all of them) the systems that were used are/were wholly inadquate on a number of levels.
There are also questions about other factors.
No, the tests are not worthless, but they are only valid for the test conditions - I object to these tests being used to draw wide ranging genralized conclusions.
I do not object to the DBT/ABX methodology, as far as it goes. It is certainly a methodology that does *test* something and do it reliably. What it is testing is still a big question! :- )
I personally have not tried Pink Noise in a DBT test. I am certainly willing to do so, and if someone out there wants to ante up a few $$ and some time and effort it can be made to happen... Since I'm short on both time and $$, I can't be an audio philanthropist for this one. If someone reasonably local to the NY/Phila/Boston axis wants to do such tests, I'd be willing to try to participate and lend whatever expertise I might be able to bring to bear.
Again, the differences that at least I am talking about are ones of *clarity* in the sense that information is easier to recognize. Since you *can* already recognize the information with what I will call "less than optimal" equipment/cables, this is a difficult thing to test - since the difference is one of *effort* on the part of the listener's brain!
I think this is really where the problem lies...
_-_-bear (bearlabsUSA.com) |
Bmpnyc wrote: "Lamp cord is not acoustically transparent. The quality of the copper must have an effect on signal transfer.The cable companies that offer the purest materials
seem to garner the best reviews. Harmonic Technology's single crystal wire comes to mind."
The largest effects upon subjective results with speaker cable is not the conductor. It is two things: the insulation and the geometry.
Lamp cord copper would work reasonbly well if the insulation was changed and the geometry was changed.
If you doubt that the geometry has a big effect, then try this and report back. Get some runs of lamp cord. Run one set like lamp cord. Take set #2 and separate the two condutors completely, by inches. Listen.
I think you'll hear a difference.
Having said that, I do think that for reasons that no one knows that pure Silver and copper do sound different, and SPC sounds different again... although again, this is more subtle than the difference in insulation and geometry.
_-_-bear (bearlabsUSA.com) |
Geez...
It's hard to follow threads on a non-threaded heirachy site!
Anyhow JHunter suggested the pink noise method for testing cable "differences." I certainly do like to use pink noise for discerning very small effects. You can use pink noise when tweaking a crossover, change a *very small* resistor value and *hear* a change. You can forget about hearing the change using any sort of music.
So, does it matter if you can't hear the change when using music? You bet! The resulting sound when the improvments have been made are almost always subjectively better sounding and most often objectively test better with test equipment.
this is a follow up to Jhunters post back 5-31-01
sorry for the delay! :- )
_-_-bear |
Frap's post is interesting and points out a real issue.
There are two intersecting factors in his turntable story. The first is that since phono has the lowest signal levels that you are likely to find in any system, they seem to be most sensitive of all. People who use my Silver Lightning product between the turntable and the phono pre often report rather significant improvements in almost every aspect. I take it to mean that the low level signals are most likely to be effected by *anything.* The second factor is that most people have really *never* heard, or had an opportunity to hear the effect of a single, simple change (assuming it had an effect at all)! So, it is obvious that they will be very skeptical to say the least.
I had a really wierd episode with sonic effects of materials at one time, which made me really scratch my head and wonder if I wasn't hallucinating. To make it brief, I thought my big Symphony No.1 amp, one just built and being listened to for the first time, was distorting badly in both channels. I was really worried that I had made some stage on the PC board non-linear by inserting the identical wrong parts on two boards! It really sounded odd, the highs were "lifted" above the speakers, and they sounded oddly strident as well. Wrong as far as I was concerned. I am sorry to report that the problem was cured by changing the binding posts. The binding posts that were "bad" were of another manufacturer and were the standard 30 amp 5-way style, and *identical* in every visual respect to the "good" ones!
No, I did not (this is now 12 years ago) save the bad ones, and run comprehensive tests to see what the technical differences were - I wish I had. But, since the problem went away immediately, there was nothing much to think about.
If you never had this sort of experience, you'd think I was completely wacky in the head. Indeed, the experience is/was wacky, and I can not explain it. It was certainly not a mass delusion or some sort of sighted bias at work. The last thing that I wanted was to have completed an amp at 2:00 in the morning, and have to trouble shoot it because it sounded like dog meat.
So, there are those who have had these sorts of things happen, and those who have not. Some of the have nots simply don't have gear that would permit them to discern what was going on (like if your amp has those brand binding posts on it all the time! :- ) ) or simply have just never been in a situation where it simply HAPPENS. They, with good reason, tend to be very skeptical about these reports. I understand.
_-_-bear |
I presume the meter was FIXED on to a stand or tripod, and that the equipment was not changed in setting or other value. I think you said that you ran several iterations of the test to verify results - which would tend to verify them.
One also has to be certain of YOUR physical position at the time you take the measurements, it should be the same (of course, ideally, you wouldn't have to be close to the mic at all.
I'd like to know what the physical geometry of the cables are, both like zip cord? And what speakers and amp are you using? Do you know the impedance and phase plots for it?
The next relatively easy test to make is a gross inductance and gross capacitance check using a DVM that has these functions. Of course slicker methods would be nice, but at least you'll have a first step.
Anyhow the results you find are indicative of either major differences in C & L between these two cables, and perhaps a major interaction with the crossover components in the speakers.
_-_-bear
bearlabs.com |
