As bear said...
This is always true for everyone including bear I hope some day he will accept that.
I've not looked at this thread for awhile. Reading what's been written since then, I'm curious as to what happened to the test that bear was going to do with opamps in series. Has the test been done and posted somewhere else?
Jay
Jay
It's been a lot of words, not much action. (Myself most guilty of contributing to that direction)
It turned out Sy had done a test like that and published it in Linear Audio. It is the "Testing One, Two, Three" article at https://linearaudio.net/volumes/783
Bear may have decided he didn't need to do something similar.
Thank you for the answer Daniel. A couple of weeks ago seemed like a pretty simple proposition. Bear would or would not be able to hear the difference between a straight wire and some opamps in series. Standard testing conditions. What's so difficult?
Jay
Jay
Mark
We seemed to have cross posted. The discussion of testing continued after the mention of the article by Sy in Linear Audio. Bear wanted the discussion taken off thread. I was curious if it had or I guess not been done. Is there a problem?
Jay
We seemed to have cross posted. The discussion of testing continued after the mention of the article by Sy in Linear Audio. Bear wanted the discussion taken off thread. I was curious if it had or I guess not been done. Is there a problem?
Jay
I don't know if there is a problem. But, I don't think it would be controversial to suggest that if you put enough opamps in series, you will start to hear more stuff. I have a one channel, three band parametric EQ that runs the signal through 14 opamps. So, I think it turned to discussion of a single stage line level amplifier, maybe with gain =10. It wouldn't be too surprising if that produced inaudible distortion, in fact, it might be more surprising if distortion were audible. Then again, it depends on the circuit, layout, power supply, load, etc., not just what's on the schematic, when thinking about all the things that could influence the outcome.
Mark
Interesting to hear about your 3 band parametric EQ. My memory, could be wrong, was that it was to be a series of opamps with a gain of one. The opamps in the article by Sy had a gain of one I think. Simple enough. Either you hear a difference for a number of opamps in series or not. I'm sure it would take many opamps for me to hear anything. I was curious about the results that bear got.
Jay
Interesting to hear about your 3 band parametric EQ. My memory, could be wrong, was that it was to be a series of opamps with a gain of one. The opamps in the article by Sy had a gain of one I think. Simple enough. Either you hear a difference for a number of opamps in series or not. I'm sure it would take many opamps for me to hear anything. I was curious about the results that bear got.
Jay
Both those statements are, unfortunately, all too true for most of us, most of the time.
But, thankfully, that doesn't mean they're true for all of us, all of the time.
In our better and smarter moments, at least some of us can be smart enough, and educated enough, to use the scientific method.
The whole point of the scientific method is to learn that what *I* think isn't necessarily right; rather, reality is whatever is revealed by experimental data, whether or not those results agree with my preconceived notions. It is hard on the individual ego, but extremely effective in terms of finding out what makes the world around us tick.
And so, centuries ago, thousands of people found out that the earth wasn't flat, even though they had always believed it was, and even though the evidence of their own senses told them it was flat. It turns out that our senses are unreliable and limited, and when we work a little harder to obtain data, there are a lot of different experiments, all of which lead to the conclusion that the earth was more or less spherical, and not flat after all.
But the lesson that we can't trust the evidence of our senses is easily forgotten. And so, some of us now believe that we can hear things that we can't measure, even though there is a huge body of evidence that shows that the opposite is true, and it is our hearing that is flawed, limited, inaccurate, and untrustworthy.
I once measured 18% third harmonic distortion from a subwoofer. Nobody who listened to it - including many educated ears belonging to audio professionals - made any comments about hearing that distortion. Basically, nobody could hear it (I was no exception). And yet, we have arguments here that an opamp with 0.0005% distortion over the entire audio bandwidth sounds audibly flawed to the golden-eared.
Ah, but what if all measurement methods are mysteriously flawed in some unspecified way! It's my golden ears that are correct! Your amp is misbehaving when playing Bach on my super-duper ultra-slew-rate turntable, even though it passes all your audio measurement tests perfectly!
There is even an experimental answer to this ultimate subjectivist gotcha. Many decades ago, Peter Walker used to use a pretty simple set-up to see if his amplifiers were somehow misbehaving, while playing actual music, and driving actual loudspeakers.
Basically, he subtracted the output voltage of his amplifiers (appropriately divided down with a pot) from the input signal, and looked at the difference. If the amp behaved properly, then that difference signal would be inaudibly small. If the amp misbehaved in some way - mysterious or otherwise - while playing Bach or what-have-you, then that difference signal would immediately become huge, and that would instantly reveal that there was a problem.
Walker did find problems with some amplifier designs on the market at the time. But he also found that the best amps of that period were already audibly perfect; the error signal was buried in the noise floor of the amp, and far below the threshold of audibility. Quite literally, it was impossible for any human being to hear any flaws caused by the amp; in some cases, the distortion artifacts were even lower than the "shot noise" from air molecules in the room bouncing off the listeners eardrums, the irreducible thermal noise of the air in the room itself!
That test method should have become the ultimate gold standard for settling arguments about mysteriously misbehaving audio electronics (whether opamps, or otherwise). Instead, it remains almost completely unused, and has been almost completely forgotten. (Rod Elliot does have a description of a more modern version of this test, somewhere on his Elliot Sound Products website.)
The most logical conclusion is that many - (maybe most?) - of us do not actually want to find out the truth. We'd rather continue to believe whatever it is that we already believe.
Do I use this test on every audio device I own or build? Nope. I trust the huge body of experimental evidence that says that the usual battery of routine audio tests - particularly frequency response, distortion spectra, and intermodulation measurements - are sufficient to reveal any audible problems with my electronics, and are certainly far more trustworthy than any ears on the planet. And I also know that my speakers probably produce at least a hundred times more distortion than the electronics, making it utterly ridiculous to focus on the latter, rather than the former.
This thread belongs in Salem, with it's long history of witch-hunts. How do you prove that an op amp isn't a witch, sorry, isn't flawed? Heat it up to 300 degrees Celsius for ten minutes, and if it doesn't burst into flames, then it was a good one? 😀
-Gnobuddy
Hi Scott, not so fast.
Just last week I compared an elcheapo moulded ends Stereo 3.5 to Stereo 3.5 cable that came in a Logitech setup.
When used as interconnect from portable phone player to my hifi, this cable assembly causes the stereo to sound different according to direction.
I have observed directional wire and directional cables effects for a very long time.
I put this question to an old nuclear physicist who blandly stated "Yes of course, expect it".
The conversation turned so I did not get the explanation as to why, so as yet I don't have proper explanation, sorry about that, but I do have ideas.
So I accept this as fact, as this conversation reinforces my findings and associated comments on wire directionality in the past, including prior to the above conversation.
I have related at length my experience with reversed hardwired turntable interconnects causing sideways shift of stereo image that was not correctable with the balance control, and other examples also.
If you think those accounts fully through you are on your way to working out the subjective/system effects of wire inherent directionality, and hearing it, then recognising it, and then being stuck with it...
The physics reason is of course at a deeper level, I'm working on it.
Dan.
Last edited:
Yeah, I have that since long ago.
Rotten sounding recordings sound exactly that, rotten for the production faults but no 'reactive' nasties that don't belong are added, rendering those 'rotten' recordings tolerable, listenable or even enjoyable despite the 'errors' in the production processes.
This is a big distinction between good gear and lesser gear...really good source material sounds really good on just about anything, HOWEVER bad recordings cause a bad system to 'misbehave' often intolerably.
Signal embedded noise is the driver in all of this, good systems ignore it, bad systems 'chuck a hissy fit' and objectionably audibly so.
My experiments cure this signal embedded noise dependency out of the equation.
Dan.
Easy to agree with you at the extremes. Hearing 0.0005 distortion is unlikely. Especially a little flattening on the top of sine wave.
The questions I would like to examine are more along the lines of, (1) if someone uses some kind of ear training to improve distortion detection, how much improvement is possible? (2) exactly how sensitive to hearing distortion are the very most sensitive humans?, (3) what distortions are experientially most objectionable, and least objectionable? and (4) which are the most audible, and least audible?
It's not clear to me we have unassailable answers to all of those questions.
In medical research, and hearing acuity is medical research, replication of studies is very important. Not just running different, but similar studies, but actual replication. Experience has shown that effect sizes are often diminished when attempts to replicate are made. Part of the reason for that is publication bias affecting the original studies. There are other reasons as well, many not well understood.
Last edited:
Well, that's at least a starting point. 😀
Tracing distortion from vinyl record playback has been shown to exceed 10% THD at the inner grooves in many cases; yet, nobody seems too bothered about that, including so many supposedly golden-eared audiophiles. Rather curious, no?
I think all of those are perfectly valid questions. I also think answers to all the important ones have already been established in the huge body of research and listening tests that have already been done over the last fifty to ninety (!!) years. Remember, Bell Telephone Laboratories was already very interested in audio quality (for telephones) in 1925, ninety-one years ago.
For example, Floyd O'Toole mentions in one of his talks that trained audio professionals could detect some loudspeaker deficiencies at a statistically significant rate, while members of the general public couldn't. IIRC, the difference wasn't dramatic, though.
There's a possibly useful rule of thumb, here, that gives us an idea how big the standard deviation is when it comes to the statistical distribution of human abilities. I note, for instance, that the fastest (human) runners are typically about twice as fast as an average fit person; very few people can run a ten-second 100-metre sprint, but tons of people can do it in twenty seconds, and even the physically unfit could waddle their way to the finish line in thirty seconds.
My point is really that the physical abilities of even the rarest of the rarest of exceptional human beings is usually not even one order of magnitude removed from the median human being. Nobody can run ten times faster than a normal person of normal fitness, for instance.
You can extend that reasoning to senses like hearing and vision. Nobody can see ten times further than a normal person. Under the same ambient noise conditions, nobody can hear a ten times quieter sound than a person with normal hearing.
It is therefore highly plausible that nobody can hear ten times lower distortion than an average person.
Today, most of our electronics is several orders of magnitude better than anything that has ever been proven to be detectable at a statistically significant level in a properly controlled double-blind listening test. For instance, plane-jane NE5532's in a bread-board will get you to 0.002% THD.
That's equivalent to two hundred thousandths of a watt of distortion riding on one watt of signal power; fifty decibels lower, utterly inaudible, under any circumstances, by anyone. You can hear 20 microwatts of power fed into a sensitive speaker, yes; but not with a million microwatts of signal blasting out at the same time!
Meantime, our loudspeakers continue to produce several percent distortion - at least a thousand times more than that NE5532. In the same audio chain.
So why are we even wasting our time considering the supposed failings of the 5532? 😕
-Gnobuddy
This was one of the things that drove me nuts with LP's. I did check with friends at the time if they also often found the last songs on a side to sound worse than the rest, but this sensitivity seemed to be rather personal.
Once sensitized to a kind of distortion, it becomes impossible to not hear it when it occurs.
I did, frequently. In my case though almost certainly not tracing distortion but build-up of fluff around the stylus.
Hi Scott, not so fast.
Just last week I compared an elcheapo moulded ends Stereo 3.5 to Stereo 3.5 cable that came in a Logitech setup.
When used as interconnect from portable phone player to my hifi, this cable assembly causes the stereo to sound different according to direction.
I have observed directional wire and directional cables effects for a very long time.
I put this question to an old nuclear physicist who blandly stated "Yes of course, expect it".
The conversation turned so I did not get the explanation as to why, so as yet I don't have proper explanation, sorry about that, but I do have ideas.
So I accept this as fact, as this conversation reinforces my findings and associated comments on wire directionality in the past, including prior to the above conversation.
I have related at length my experience with reversed hardwired turntable interconnects causing sideways shift of stereo image that was not correctable with the balance control, and other examples also.
If you think those accounts fully through you are on your way to working out the subjective/system effects of wire inherent directionality, and hearing it, then recognising it, and then being stuck with it...
The physics reason is of course at a deeper level, I'm working on it.
Dan.
Some proof would be nice... Some hearsay from a nuclear physicist is not proof, is he an expert on signal propagation in cables... |it is also one up on the Axle Rose in the Kitchen...
Cable directivity is pure BullS*** in nearly 10 years now on this site I have not seen or heard one plausible explanation and 30+ years in getting signals from a to b I have never seen or heard cable (or PCB copper) directivity discussed (except on audiophile sites), I wonder why no one else has noticed it...
Sorry proof not anecdotes needed...
People like vinyl because of it's flaws, most vinyl heads either don't know this or just wont admit it. A transparent system aimed at getting the best out of digital often makes vinyl sound terrible, the surface noise and distortion become so obvious that it can ruin the experience. This is why there are still so many vinyl and valve fans. I used to hear old timers commenting on the nice tone a hifi had, they at least knew the sound was being modified even if they were clueless as to how any of it actually worked. Vacuphile is right, once you start noticing distortion it becomes impossible to ignore. Gnobuddy's post is also bang on the money.
I agree, they should be banned, all stocks destroyed and any engineer using them burn at the stake, before you know it they will be using the little buggers to discover gravitational waves... In fact I would go further, we should destroy all the looms and burn down the mills, pass me Enochs' hammer and lets get smashing...
- Status
- Not open for further replies.