There are measurable differences between all resistors, therefore there should also be audible differences.
but I do not agree that the least errors always gives the best sound.
I hate the sound of vishay resistors.
And they have just as much sound as other resistors even though they are technically more perfect.
Last edited:
The article explains the concepts and it is important to understand these. Some numbers are included and if you are interested there is a lot of material in the references.
Understanding prevents the erroneous conclusion that memory of differences between complex stimuli will fade away within ~30s .
If an experimenter is interested in results of practical relevance he will surely question the need for rapid switching between DUTs. (Although it might help wrt efficiency)
Please cite some published articles where the relationship between the degree of a difference and long term storage ability (complex stimuli aka music used) was studied.
Here's a simple experiment that anyone can do which will cut through nonsense like this. Take a track, duplicate it at a different level, somewhere above JND, but not a gross difference, maybe 0.5-0.7dB difference. Do an ABX with rapid switching. You're likely to score significantly.
Now, run the same test, but allow a minute between comparisons. I'll be happy to bet on the outcome of that.
The latter does not necessarily follow from the former. For example, resistor inductance is measurable and differs between resistors, yet is almost completely irrelevant to audio.thor2 said:
Statements like: "it might help wrt efficiency" are IME indicative of a person with no actual practical experience with doing bias controlled listening tests.
It is not a matter of might, it is a matter of as I said, the 500 pound gorilla in the room jumping up and whapping you in the face, repeatedly.
As Sy pointed out, excessive change over delays make doing sensitive listening tests difficult or impossible. I've done the experiment many times - extended the switch over delay to beyond even just a second or two, and watched my statistically significant results disappear in cloud of random guessing.
Where the discussion seems to be going now is that there is an attempt to hypothesize odd circumstances where excessive change over delays don't make doing sensitive listening tests difficult or impossible.
As I've said before, if you do bias controlled listening tests for small audible differences among audio gear, excessive change over delays can be counted to make doing sensitive listening tests difficult or impossible.
Check the bibliography of this book, which itself covers the topic well:
This Is Your Brain on Music: The Science of a Human Obsession
by Daniel J. Levitin
This Is Your Brain on Music
Also you might benefit from studying some documents related to the body of work related to doing sensitive and reliable listening tests:
ITU Recommendation BS 1116
"Methods for the subjective assessment of small impairments in audio systems"
http://www.itu.int/dms_pubrec/itu-r/rec/bs/R-REC-BS.1116-2-201406-S!!PDF-E.pdf
"Since long- and medium-term aural memory is unreliable, the test procedure should rely exclusively
on short-term memory. This is best done if a near-instantaneous switching (see Note 1) method is
used in conjunction with a triple stimulus system as described in Attachment 3. Such switching
demands close time alignment among the stimuli. "
I have to wonder how the instrumentation that is used to measure can work if resistors change analogue signals so much... I do presume you are talking about resistors in analogue circuitry, in digital circuitry they are going to be used as either serial or parallel terminators so will have NO influence on the resulting sound.
I would also be interested in what positions in circuitry resistors change the sound so much, so more specifics on where resistors can change the sound so noticeably....
Hi,
Whilst I agree with some of what you've said in previous postings,
a sentence like the above is totally meaningless in view of the range of resistors they manufacture.
Cheers, 😉
Whilst I agree with some of what you've said in previous postings,
a sentence like the above is totally meaningless in view of the range of resistors they manufacture.
Cheers, 😉
Last edited:
Yes of course, I'm talking about resistors in analog and digital circuits
Interesting is that the resistors / components generally retains its sound signature in the digital domain, sometimes reinforced compared to analog, in my experience
How can a resistor used in digital circuitry change the sound, that's ridiculous..... Read my earlier post.
Last edited:
Of course; the bits get just sooo much nicer. And then when a byte of, say, value 01010101 gets converted to analog, it makes a big difference if the original bits were vishayed or not. Duh.
Jan
Jan
Let's try one more time.
"Not "the" charge, charge", is meaningless. There is no charge independent a particle with the property of charge. In the case of a pure metal charge = electrons (or absence for +).
.
Hence "charges moving per second. With a mean propagation velocity of .8c " is nonsense. An EM wave propagating is not the transport of charge and no particles moving at near c..
Hence also, "that would allow 2600 charges per meter in a cable carrying that current." Is wrong, yes Ed wrong. The metal is charge neutral, there are a lot of free carriers and what goes in comes out and the total = 0, there are no "excess" 2600 "charges" (whatever you mean by that) while the current flows.
DC current in a wire is explained (completely) by the simplest description of drift velocity with j = sigma*e where sigma (conductivity) depends on the mean free path in your particular metal, etc.
Last edited:
Scott,
Actually getting somewhere! If I send a 10,000 Hertz signal through a one meter piece of wire is the "mean free path" the same if the wire is 32 gauge or 8 gauge?
Actually getting somewhere! If I send a 10,000 Hertz signal through a one meter piece of wire is the "mean free path" the same if the wire is 32 gauge or 8 gauge?
Marce, Marce, Marce,
When it is failed open of course! Could you please be more specific in your questions? 🙂
Again another issue of difference, just how good a PC designer can you be when you need to use more than a single layer? 🙂
Yes and no. Surface effects (as in grain boundary and insulator interfacial) will obviously be a much bigger effect at 32 gauge than 8 gauge, ignoring entirely skin effect. Don't have a pulse on how big an effect that is. In the bulk of the metal, they'll have the same mean free path.
This is assuming identical quality of your metal, as defects/impurities will serve as scattering sites (and there goes your mean free path). Also assuming that the insulating sheath doesn't contaminate the outer surface.
Given what NGNPP is I am very very very glad he is not trying to do it on a single layer!
Bingo, you got it! Impurities are scattering sites and surface effect comes into play. So we could ask what the standard deviation is. But more of interest is how to measure the variation in mean free path. How would that show up on our 10,000 hertz signal?
At/near room temperature? If it doesn't affect the bulk resistivity, I'm out. You're going to need some *seriously* specialized equipment to tell.
- Status
- Not open for further replies.