@SY:
Hey, a sample of one is a big step up from zero!
Just need some more like that.
That's exactly what we don't need. A zillion solo listeners writing in saying that it sounds better looks impressive, but is MEANINGLESS and only leads to potentially false conclusions.
We need people to STOP writing in with 'evidence' like this.
w
Doh!
Yet it isn't. 50% chance, single trial with coaching from the cable guy, is not exactly persuasive.
I expected that but show where there was coaching from the cable guy - or stop making it up!
Before I embark on a controlled Listening test could we get some agreement on what qualifies so as this is not a waste of time? What group size, how many cables, double blind, level matching, etc.
Last edited:
Yet it isn't. 50% chance, single trial with coaching from the cable guy, is not exactly persuasive.
Nah then SY, that comes too close to calling me (and by inference, my forum friend) liars, which I seriously object to - if you cannot accept the following assurances then I am wasting my time and you do not want to be open minded: -
1) there was absolutely and categorically NO 'coaching' from the cable guy - see my post above where I offer the possibility of confirming this with the guy concerned.
2) I would have no problem with your 50/50 chance between two cables if the difference was close, but it wasn't - it was so distinct as to be a 'no brainer'.
If you can't accept these assurances unreservedly and at face value then it negates the whole forum concept of sharing our experience - your choice, come out and smell the coffee or stay blinkered in your own little mindset - I have nothing to gain or lose either way.
Dave.
And exactly what relevance has all that to the subject under discussion? I have expressed no doubts about your audiophile capabilities or knowledge, only about your rigid closed mind set. And since when has quantity counted more than quality? You hang on in there with your beliefs and don't let anyone shake your convictions.
Dave.
Agreed.
I was just pointing out that of the zillion, this was the first blind one, FWIW.
Also agreed. That will be a problem with any listening tests, no matter how well controlled. A combination of listening tests and measurements would be more useful.
No, what I'm saying is that you don't have much of an idea about how a controlled listening test is structured or interpreted. You also are unfamiliar with the various ways that the controls can be inadvertently broken.
On the plus side, you reached out to Jan for help in doing one and (at least for the moment) appear to be willing to give it a try. I highly commend you for this.
Hi Jan,
I'm not being thick, I genuinely am new to this so I have no real idea of what a contolled subjective listening test is, let alone how to conduct one
Test 1 consisted of a listening session consisting of 3 tracks, one orchestral, one female vocal and one electronic, with the existing setup to give me a current bench mark, test 2 involved repacing the coax cable linking HiFace and DAC with one of the trial cables and listening to the same three tracks and test 3 involved replacing the first trial cable with the second one and listening to the same three tracks. My first impression placed my original cable between the two trial cables in terms of the SQ I was hearing and the best cable was so obviously better to my ears with my set up that within 45 minutes of starting the test I registered with the guy who made the cable what my initial opinion was and why. Other than saying "That's interesting" he made no further comment. I continued to repeat tests 2 and 3 with different tracks for a couple of hours that evening and again at intervals the next day. The result was that my I became even more convinced that my original opinion was validated, the difference was so obvious it was a 'no brainer' as they say. I reported my confirmed opinion to the friend who made the cables and he confirmed that the cable I preferred, tagged blue, was the 10db attenuated one.
Subject to me confirming his permission I am quite happy to put anyone in direct communication with the guy who made the cables - he is/was a Moderator on another very popular audio forum so his word may carry more weight than mine. Neither of us have any connection with or have ever met John Kenny - the only reason that I did the listening tests is that I am the only one with a HiFace of any description and I asked my more knowledgeable friend for his opinion on John's suggestion that such attenuation wouls improve the SQ from a HiFace/DAC combination. My friend was initially sceptical but open minded enough to volunteer to make the test cables so that I could undertake a 'blind' listenig test for myself. The rest is history as they say - we have absolutely no axe to grind in reporting my findings.
Jan, if you care to PM me with an outline of what you would consider to be a controlled subjective listenig test (if the above doesn't qualify), if it is within my capability I'm willing to have a go.
Over to you.
Dave.
Dave, your test is one of the better tests I've read about, obviously you instinctively understand the issues involved.
In my opinion, the most important part missing is that it wasn't double blind. You should listen without knowing which cable is connected, and if the cable is changed, you should not even know whether the change actually was a change or whether the previous cable is still connected - a fake change, so to say.
Also, very important, the guy who does the change (or not) should also have no idea what is what. It has been shown in medical testing that medicin or placebo effects are even depending on the administer knowing (or not) which one is the placebo; so powerful can body language be).
In short, the idea is to exclude *everything* that doesn't concern the sound itself.
It is not necessary to find out which is better or worse; it is enough to be able to reliably and consistently be able to distinguish between the two under controlled double blind conditions.
Then there is the issue of number of trials. Standard statistical theory says that in principle you can never be 100% absolutely sure one way or another. The idea is to do enough trials that the confidence in the outcome is say 95%. You want 99% confidence, you need more trials.
Look at it this way: if you do a single trial, and you have it right, that's close to useless. If I say that I will flip a coin and that I have such keen sight that I can tell the outcome before it stops moving, and I do, do you believe I can correctly 'see' any flip outcome? Don't think so. Now I do 50 trials, and each time I tell the outcome and each time I get it right, do you now believe me? Probably; but it still is no 100% guarantee that I get it right the 51st time.
jan didden
Last edited:
So, 'overwhelming majority' or 'consensus of opinion' bear no weight in your little world then? I'm not saying that either of the above is in my favour, just that there is no such thing as the perfect SQ, it's all down to the perception of the listener, so all we are left with is individual opinions at the end of the day. If you don't accept that then there is little wonder that you are so fixed in your beliefs.
I think 'doh!!
Dave.
No Dave, neither you nor 'the guy involved' can be sure whether there was no coaching or not. You are not accused of lying at all. Because the guy knew what is what, that's enough to invalidate the blind requirement. See my post before.
If this sounds unnecessarily harsh, remember that the scientific method (because that is what we are talking about) has been finely honed to try to be as sure as possible that a result is or is not to be trusted. Audio differences are pretty trivial as far as importance to society goes, but the same method is used, for instance, to determine whether the earth is getting warmer. Or whether cold fusion is possible.
jan didden
Last edited:
Yes, so let Waki explain how if, in a blind test, a preference is shown for the cables with attenuators that the argument won't suddenly become "oh this could just as easily be extra jitter" - makes the cry for controlled listening tests pretty meaningless & transparent for what it is - a ruse!
I don't understand the defensiveness Dave. No-one is calling you a liar - except you.
The "test" you did was in layman's terms blind, but in reality it wasn't, not even by half.
First there is the issue that you EXPECT to hear a difference - you know that each cable is different, you just don't know which one. So you listen for difference. In preference to listening for similarity. Or even just listening.
This is subconscious - you probably did not do this on purpose or with any intent to be biased from the start, but that is the effect.
You then changed the cables yourself. You still didn't know which was the altered one, BUT you did know which colour cable was in at each change. That leaves you with a 50/50 chance of choosing to "test" the altered cable first up, ascribing a preference to it, and then continuing that bias again probably subconsciously.
There is a bunch of other stuff around your interaction with the cable guy - again, nothing said but subconscious signals you have probably picked up on.
You didn't lie - you fooled yourself.
So now accuse Dave of not having much of an idea - he has already stated his credentials in all of this. You said the cable guy was coaching, please present your evidence or withdraw it.
@Jan, the cable guy wasn't there changing the cable - you are making assumptions, I'm afraid - he handed over the cables & had no contact during the listening. What do you see wrong with this?
See my previous posts. The idea is that you should have no clue - as in being ignorant - as to which cable is playing at each trial, nor should the person doing the trial have any idea what the difference between each cable is. There should be absolutely no contact in any way between the listener and the test administrator. Obviously I don't know all details of your test so it is possible I made a wrong assumption. I shouldn't.
jan didden
There is ample research into the problem of un-intended coaching by the person providing the test - in the case the cable guy. It doesn't need evidence in the particular - the description of the test as it was applied means the coaching can be taken as read.
Wrong John - read the post again.
Well, if you find the call for a well-controlled test a ruse, what can be do?
Lets do one thing at the time and first establish whether there is a reliable, repeatable, audible difference between the two cases. That would be a huge step forward one way or another.
jan didden
Ok, so if the controlled listening tests show a repeatable difference between the two cases then SY's measurements are seen as not capable of revealing the difference, right?
We will then need somebody to test them who is capable of resolving low levels of jitter of the order of ps that Joe K mentioned. Have we got anybody that can do that or will this just end up as "it could be the result of more jitter" so it's rejected as of use!
You can see that without this jitter difference measurement, the listening tests are less than useful to prove the topic of this thread!
Wrong. My comparisons were between your $500 box (with attenuators) and a $20 Chinese box (without).
- Status
- Not open for further replies.