I don't know what OP heard. IIRC he asked if it mattered what order the resistors were in relative to signal flow.
What I do give credence to are the claims of John Curl regarding resistor sound. For one thing, I bothered to check for myself by putting some different-construction feedback resistors, all of the same value, into a discrete voltage regulator for dac AVCC (which has zero PSRR, so it makes for a sensitive test). My conclusion was that resistors can affect sound in some cases. In addition I talked to some other people about it who would be in a position to know. They provided additional confirmation that resistors can sometimes affect perceived sound. May I ask what you did to check, if anything?
What I do give credence to are the claims of John Curl regarding resistor sound. For one thing, I bothered to check for myself by putting some different-construction feedback resistors, all of the same value, into a discrete voltage regulator for dac AVCC (which has zero PSRR, so it makes for a sensitive test). My conclusion was that resistors can affect sound in some cases. In addition I talked to some other people about it who would be in a position to know. They provided additional confirmation that resistors can sometimes affect perceived sound. May I ask what you did to check, if anything?
Last edited:
You are barking up the wrong tree. I have not made any sweeping claims of inaudibility of resistors (https://www.diyaudio.com/community/threads/placement-of-resistors-in-signal-path.384534/post-6984779). However there is very little (if any) support in resistor characteristics for OP's claims.
The proper word is "perceive" because whether they actually heard it or not needs to be evaluated to confirm which basically none of them bother to go through, including you.
What I do give credence to are the supporting evidence. You should try it sometime.
I've not seen any evidence that they did. Or which DACs there were comparing with. As such I still reckon that is a marketing ploy. If they had published details of the DACs under test and the music used then it could be reproduced. But they didn't so we can't.
That wasn't the point of the blog post according to my reading.
Bill, you can pick and choose your arguments as you wish. I wouldn't have been surprised if you had said something along the lines of, "Extraordinary claims require extraordinary proof, and a never before described violation of masking theory is clearly an extraordinary claim."
Well you seem to. I just view the ESS claim as marketing until they put up something that allows someone else to replicate the experiment. Which they won't, but you still hang on it it like a terrier to a bone even though you have other DACs you think are superior and don't have this secret sauce in them.
But it's more complex than just 43Hz modulating the carrier. If not, there wouldn't be so many side bands.
Correct. You got it. Just modulate the carrier with 43 Hz. Nothing fancy.
I thing you should produce the file first, then we'll talk.
Relating back to the thread topic....resistors in the signal path....
It would be helpful if someone who has tested and "auditioned" resistors and determined a difference would post:
1. Exact examples. Resistor types, models, values, measured and audible results
2. The exact test circuit
A. The signal source and source device
B.The driving circuit
C.The position of the DUT, its type and value
D. The means used to sample the result
E. The system used make measurements (exact FFT parameters, etc.)
F. The exact system used to audition the result
G. A description of the acoustic environment the audition system was used it
And that's just a start.
The goal here should not just be individual isolated subjective claims, but actual data with a test that is repeatable with reproducible results. Data can of course include compiled subjective observations, but a single observer's observations do not provide enough data for reproducibility.
It would be helpful if someone who has tested and "auditioned" resistors and determined a difference would post:
1. Exact examples. Resistor types, models, values, measured and audible results
2. The exact test circuit
A. The signal source and source device
B.The driving circuit
C.The position of the DUT, its type and value
D. The means used to sample the result
E. The system used make measurements (exact FFT parameters, etc.)
F. The exact system used to audition the result
G. A description of the acoustic environment the audition system was used it
And that's just a start.
The goal here should not just be individual isolated subjective claims, but actual data with a test that is repeatable with reproducible results. Data can of course include compiled subjective observations, but a single observer's observations do not provide enough data for reproducibility.
........
I have been training to fly by flapping my arms for decades now.
To hear what´s being talked 100 meters away, in a busy street.
To swim underwater for 200 meters just holding my breath.
To kick a ball over the Empire State.
WHO are YOU to tell me I can´t?
Have YOU ever tried it?
If not, don´t comment, you are not qualified.
😉
In 1960, the average USA family had 2.3 children. Currently the USA average family is only 1.9 children. But in all the people I have met, I have NEVER EVER seen a real average family. They only seem to come in whole numbers somehow.
Statistics do not apply to individuals.
Here is the problem. You want to test a single resistor. You want to specify the test circuit, all the circuits around it, the signal level and nature, and so on and so on. What you cannot do is decide that the same results will obtain in other circuits, other placements, other conditions. Component characteristics do not necessarily scale as parameters change. If I put a 10v signal across the resistor and find it makes noise at -70db, that does not mean I can assume a 1 volt signal will result in a -70db level,
If I find a certain resistor adds "warmth" - whatever that may be - when between the collector of one transistor and the base of the next, I cannot assume it will therefore add warmth if I place it in an emitter follower. A resistor behaves in some manner in a voltage divider in the signal path, but the same resistor used to set gain in the feedback loop of an op amp may not act the same.
Statistics do not apply to individuals.
Here is the problem. You want to test a single resistor. You want to specify the test circuit, all the circuits around it, the signal level and nature, and so on and so on. What you cannot do is decide that the same results will obtain in other circuits, other placements, other conditions. Component characteristics do not necessarily scale as parameters change. If I put a 10v signal across the resistor and find it makes noise at -70db, that does not mean I can assume a 1 volt signal will result in a -70db level,
If I find a certain resistor adds "warmth" - whatever that may be - when between the collector of one transistor and the base of the next, I cannot assume it will therefore add warmth if I place it in an emitter follower. A resistor behaves in some manner in a voltage divider in the signal path, but the same resistor used to set gain in the feedback loop of an op amp may not act the same.
Oh boy. Missing the point entirely. An average parameter may not apply exactly to an individual, but it will be "close" to describing most individuals. The more individuals you sample, the more you would see that the statistics do work.
Whole numbers are a choice in presentation.
You have to start somewhere. If someone claims a particular passive device does something, and this shows up at least partially in objective testing, then that test should be reproducible. Sure, you can create an infinite number of test conditions, that's not the point yet. It's important that one single type of test be reproducible. We don't even have that so far.
With regard to testing a single resistor, again, you have to start with that, then test a LOT of single resistors. See, one major problem I have with this whole resistor thing is that in the signal chain there are already thousands of them, of all types, doing, in theory, all sorts of changes. So what happens if you add one more? Or 10. Or 100? How can those change what's already been changed? If there's an answer, it also should be reproducible. And that also goes for controlled subjective testing.
If a resistor has a certain nonlinear behavior, where it is used would only change the magnitude, or vector of change. Again, that's not even important yet. We still have zero to go on. We have a couple of people with single subjective opinions, and that's about it. That's not proof, no matter who they are.
@JMFahey, What do want then? An exact black and white number below which no human can detect an imperfection? A rough number such as .01%?
@jaddie, (1) True that lrisbo waveform is not perfectly mathematically a simple 43Hz modulation. Nor is a Duracell 1.5v battery exactly an idealized 1.5 voltage source. We use engineering approximations all the time to name this or that. Why does lrisbo using of 'modulation' require mathematical perfection in this case?
...and (2) True that we don't have 'proof' about resistor sound, even though you probably know 'proof' only exists in mathematics, not science. Nonetheless using the word proof loosely for a moment, we don't have proof that lrisbo files sound different. All we have are a few subjective opinions which may be considered suspect due to the violation of masking theory.
Moreover, IIRC correctly in an earlier post of yours, before lrisbo files were posted, you wanted to see the same level of 'proof' for his claim of phase rotation audibility as you now want to see for resistor sound. Change your mind? If so, due to your own subjective listening experience?
So I would rather have you listen to some resistors in a simple circuit so you can hear for yourself just as you did with lrisbo files. Less years of research required of me that way.
@jaddie, (1) True that lrisbo waveform is not perfectly mathematically a simple 43Hz modulation. Nor is a Duracell 1.5v battery exactly an idealized 1.5 voltage source. We use engineering approximations all the time to name this or that. Why does lrisbo using of 'modulation' require mathematical perfection in this case?
...and (2) True that we don't have 'proof' about resistor sound, even though you probably know 'proof' only exists in mathematics, not science. Nonetheless using the word proof loosely for a moment, we don't have proof that lrisbo files sound different. All we have are a few subjective opinions which may be considered suspect due to the violation of masking theory.
Moreover, IIRC correctly in an earlier post of yours, before lrisbo files were posted, you wanted to see the same level of 'proof' for his claim of phase rotation audibility as you now want to see for resistor sound. Change your mind? If so, due to your own subjective listening experience?
So I would rather have you listen to some resistors in a simple circuit so you can hear for yourself just as you did with lrisbo files. Less years of research required of me that way.
Last edited:
If he does what I asked we might take the discussion further.
Not true. We can prove the speed of sound in air at any specific temperature. We can prove that certain distortions are clearly audible to everyone. And it doesn't take all that much math.
Sure we do! They are identifiable, blind.
We have that too.
His files do not prove that "phase rotation" is audible. No, I have not changed my mind at all, I still require proof.
Sure, but I have no idea what circuit that is! So far, nobody does. So far Irisbo's files prove only that Irisbo files are audibly different with similar (yet not idetical) spectra.
Not having the slightest Scientific proof for wild super human claims, next best is to disqualify Science itself.
Not until we have proper ABX DBT.
EDIT: Regarding your comment in your last post above on audibility of phase rotation, you just claimed lrisbo files sound different. Did you forget they were produced by phase rotation? Therefore isn't is logically true that phase rotation is not always inaudible?
Last edited: