We're talking < 100 nano meter thickness to make a difference, the best data I found was at 69nm for copper thin film you were already at a mean free path of 29nm vs. 39nm for bulk. So I say no difference 32 vs 8 wire (32g = ~2e6 circular nm) and jn is the definitive reference on skin depth though I don't think it can factor much here.
Last edited:
I will not go so far as to call it ridiculous, again I have not the full explanation, but I can say that it is easy to demonstrate convincingly.
components sound signature only affects the audio technology, these things have no meaning in other elektronic.
This separate audio elektronic from other electronics, and it is a relatively unexplored area.
And so shrouded in in mystery, snake-oil, phousdo explanations etc.
All our electronics is located in parallel with respect to AC signals. in one way or another.
Can you recognize the that a bleeder resistor can change the overall sound we hear, or that a decoupling of a large electrolyte with a small film capacitor also can affect the overall sound, so we have come some way towards an explanation.
Last edited:
I should have written:
"components sound signature only affects the audio technology, these things have no meaning in other elektronic."
I'm not a complete idiot, so naturally I work sighted, no need to make the already difficult process even more difficult.
Sound difference in itself is forgettable, it is about sound difference gives greater credibility that means something.
Last edited:
True if you desire positive results, regardless.
Scientific tests can have err, inconvenient results.
Yet so many audiophiles seem to be obessed with presumed sonic differences. There seems to be a presumption, any difference is an improvement. This is due to a lack of stable references.
Perceived differences are always the same as actual differences.
Try stick your ear to a lamp dimmer.
(I can hear a triac dimmer buzz three floors up. But I also favor the practical comfort of adjusting light intensity. Reason for having spent quite a bit of time, energy, and funds on examining dimmers up close. Just the living room of my home has 6 in-wall mounted, plus a couple of liberos)
Maybe you should try (at least once in a while) to understand a conceptual idea before posting something.
_Practical_ _relevance_ means it is relevant for the usual normal listening; if a listener can not remember a difference after a couple of seconds it will be irrelevant.
If people like to try seriously your test proposal (and maybe do some training you´ll be quite surprised by some results)
Btw, feel free to cite some studies in which the relationship between differences and the ability to transfer to long term storage is examined.
It _is_ a matter of "might" just because some participants will loose their sense for the whole impressions while switching to fast from sample to sample.
And as said above it is a matter of practical relevance.
If you don´t follow some misconceptions about memory storage there are no odd circumstances.
As said before, we started with controlled listening tests back in ~84 after reading some articles from Dan Shanefield.
If your are listening for a specific "sound signature" a delay of 30s to 1 minute is in no way excessive and did not prevent me from detecting a difference.
At that time we started right away without any switching unit and had therefore to include a fixed time span between trials in which any manual changing had to be finished. Fixed time span to avoid the invention of any clue (true or misguided) if any change could have happen.
Having read most of Levitin´s work (and most of the material from the reference list as well) i am still confident that no study is included that examined the degree of a difference between two sound events and and the ability to transfer it to long term memory.
You might reread the related chapter of the book; we are all prone to confirmation bias and as far as i remember Levitin did explain some of the various models of memory (one of it was surely the event recorder theory) and the need for a unifying theory.
As said a couple of times before, i searched for factual evidence for the recommendations that are covered by BS.1116 and found some in most cases but not for this somewhat miraculous time span and the related recommended length of sound samples.
If you compare the recommended length (10 - 25s) with the information from Cowan you´ll note that it is already to long as the content in short term storage begins to fade away within 15 - 30s.
A valid assumption otoh is that the ability to categorize auditory input while transferring it to long term storage depends on learned skills of humans.
This led to a proposal by Precoda/Meng to use even shorter samples (5s) in a continuos loop to (hopefully) force the listener to use only short term memory during the evaluation.
Precoda/Meng; Subjective audio testing methodology and human performance factors; AES Convention 103, 1997, paper no. 4585
That these short snippets will led to greater difficulty to get an impression of the musical work (and related sound features as well) should be selfexplanatory.
Last edited:
Yeah, I was thinking tens of atoms, given this is interfacial layer stuff. 100 nm is already moving towards bulk.
I figured (perhaps incorrectly) that everyone discussing this issue knew we were talking absolutely infinitesimal magnitudes. A much larger very very small number is still a very very small number.
Last edited:
Gratuitous insult noted.
So called practical relevance would be a concept that you have gratuitously introduced. It's an inherently weak concept as it can mean anything you want, which I think is the whole point. The goal appears to constrain and contrive the situation until a certain viewpoint is supported. This is certain means for supporting highly flawed viewpoints. Schoepenhauer listed it over a century ago in his 38 Strategems, for example.
In the general case the listener has no such limitations. A/B switches capable of quick switching (subsecond) have been fixtures in audio salons for over 50 years. I was there at the time... I built a few of them...
The question we are trying to address is whether there is any circumstance that audiophiles may find themselves in, in the real world, in which the difference is audible.
Obviously, my cited references were in vain.
Listening is a form of athletics and just as surely as observation of the running of a 30 second mile is some time in the future if ever, so are vast exceptions to the limitations of listeners that many careful scientists have observed and reported. Only audiopiles can exceed those!
Been there, done that.
In the relevant tests I have done the switching was done by the listener in accordance with his experience-based wishes. He was not hurried by anything but a desire for the most sensitive results possible. In short, your appeant fears about too-fast switching are irrelevant.
Which has already been dealt with critically and negatively.
Good advice that you should take!
You were a bit late on the scene as his landmark articles were published in Stereo Review and High Fidelity (one under a pseudonym) 1976 and 1980.
They were anticipated by some things he wrote for the BAS speaker. At any rate, we did our first DBT on or about May 7. 1977. ABX Home Page
That would have to be a pretty gross difference, to be reliably detected after such a long time. Of course we don't know what kind of statistical test was used, whether the test was truly a DBT, etc. etc.
Good enough if you are satisfied with far less sensitive results than are possible.
Just goes to show that people read what they want to read. We seem to have come up with contradictory results from the same reference. You're not alone - one of the Stereophile authors seems to think that Levitin supports his viewpoint.
That is a two edged sword at best for you. Your confirmation bias oddly interprets Levitin's rather clear meaning.
Its that confirmation bias thing, again, I fear.
You've neatly ignored the influence of working memory.
Of course you've got a position to justifty.
Let's look at it this way. I read the accepted literature to say that listener sensitivity to small differences fades after a second or less. I build my tests with that in mind. I allow the listeners to do whatever they can do within the context of the overall test controls do get the best results they can. I find that virtually every listener finds that quick switching under his control allows him to obtain the best results and in a way that he is most comfortable with.
Nobody forces ABX listeners to use only short term memory or working memory against their will. We have on occasion looped samples, both long and short.
Again, your discussion of the various kinds of memory is incomplete in a self-serving way.
No text quoted from the reference, so what you mean by this cite is impossible to discern by other than paranormal means.
However I have the paper before me, and quote this:
"
However, information stored in echoic memory decays with time. An early experiment to determine this decay time was reported by Guttman and Julesz [2]. Their subjects heard repeatedly looped segments of random noise and were asked which lengths of segments produced an impression of periodicity. There was clear periodicity when the repeating segment was up to 250 msec long, and some periodicity with segments up to 1sec long, suggesting that a precategorical acoustic store contains not more than about 1 sec of auditory stimulus. A longer estimate was given by Glucksberg and Cowan [3], whose subjects repeated aloud speech they heard in one ear, while ignoring irrelevant speech with occasional digits in the other ear. From time to time the subjects were interrupted to report any digits that had been heard in the ignored ear; minimum performance was reached when the delay between the digit and the interruption was about 5 sec. In another study, by Darwin, Turvey, and Crowder [4], subjects simultaneously heard a sequence of three letters or digits played through the left headphone, another sequence through the right headphone, and yet another through both, which created the impression of sound sources in three separate locations. If subjects heard the items and were then immediately told to report the items from a given one of the sound sources, their percent correct rate was about 55%. If subjects were asked to report all the items, about 47% were correctly recalled. The actual number of items reported correctly from all three sources was higher than the number reported correctly from only one source, and thus the lower percent correct for all three sources was interpreted to mean that the limiting factor was the memory decay that occurred during the report itself, rather than the memory capacity. Decay time for echoic memory was therefore estimated by delaying subjects' report of items from only one sound source. Correct reporting levels for a single source equalled correct levels for all three sources when the delay was 4 seconds, suggesting this as a decay time.
"
OK, 4 seconds... ;-)
So my suggested experiment was too complicated for you? OK, fine, but others can try it and see for themselves.
You're right, the lack of a reference is a widespread problem.
Referancen, should in my view be live music with acoustic instruments.
Over time you find out whether you get closer to referancen with sighted test, and I do, it works quite simply.
I'm sure I had not achieved the sound I currently holds if I had used a double blind test
I have not tried to hear diffrences between R types...except that cc is noticeable over film. I select by distortion, TC, VC and noise. BUT, variable controls is another matter... differences are some times noticeable and we have other issues.... contact distortion, oxidation and plastic film and wear affects, etc.
I assumed we could not hear resistor elements and had been listening without a pot for level control on my headphone amplifier...... but when I added an input level control... there was an UN expected change in the sound. (using NOS Alps). So expecting one thing and getting another seems to imply that bias doesn't always dominate what you hear.
THx-RNMarsh
I assumed we could not hear resistor elements and had been listening without a pot for level control on my headphone amplifier...... but when I added an input level control... there was an UN expected change in the sound. (using NOS Alps). So expecting one thing and getting another seems to imply that bias doesn't always dominate what you hear.
THx-RNMarsh
Last edited:
The problem is that this reference is not just one thing. Music in a room both measures and sounds pretty dramatically different depending on the location in the room. The science of room acoustics quantifies such things.
Given the rather gross flaw found above, well it might be simple but it is not very reliable.
Very many people who have said such things have been proven to be very wrong. Every circumstance can be different, but its IME a very risky assertion unless the differences are measurably large - for example on the order of several dB in the range where the ear is most sensitive (around 2-4 KHz).
Last edited:
I think you do these things more complicated than it needs to be, there is no doubt when the sound is approaching live music, then will no one be in doubt.
when the sound is quite a distance from the live music, there is also no doubt.
But personal Preferences, space, appearance, design, mood, etc. will mean a disproportionate amount in the evaluation.
And this is where you have to put your direction and having your referance in place.
And double blindtest ABX, etc. will only confuse and make the assessment more uncertain.
And obviously used a top-class system in a good acoustics for evaluation, I use self LEDE
Ps I end the conversation here, my English is not good enough to take this discussion.
Last edited:
- Status
- Not open for further replies.