Nor is there any sort of decent plausibility for it to come from the electronics side given it's a mechanical issue. Where's the energy storage going to come from? An audio-bandwidth CSD plot of an amplifier had better be delightfully boring.
Oh how it'd be great to have some more precise language with, oh, at least correlations to data! I'm not even asking as much as causality. Until then it's all really just haphazard tweaking.
Agreed, a lot of it comes from transducers. But, to bear, amplifiers produce something that results in a similar experience in his conscious awareness. He describes it in part by saying that all cymbals sound the same, and sometimes swapping amplifiers instead of speakers can make it better or worse. Other people may describe the change in cymbal sound due to amp swapping in different subjective terms. But, I have heard what he is talking about, and when I swapped somebody else's amp to see what they thought, they unambiguously heard a significant improvement in cymbal articulation. Nothing surprising about that. We just should understand that to bear, swapping the amp changes what he experiences as a change in a sibilance-like timbre in cymbals.
Last edited:
Okay, so if we wish to move this forward with some rigor, surely there's a way to assess this in a quantifiable manner? What shows up in their comparison? Again, not trying to be antagonistic, just what can we pin down and look at in an controllable fashion. (That which gets measured gets improved and all)
"Subjective listening tests" is much more vague than what I have in mind. That's not what I was thinking of at all.
You state what you're doing and ask listeners a specific question: "Each passage will be played twice. One of these is an actual violin performance, and the other is a reproduction of a violin performance. Please give your best guess as to which is which." Go through maybe ten musical passages, each played twice, one reproduced and one live, and the reproduced-and-live order randomly selected each time.
From this you can collect the answers and get numbers (actual measurements!). If a listener consistently correctly identifies the live performance, or consistently INCORRECTLY identifies it (says the live performance is a reproduction and vice versa), then we can state the listener hears a difference between the two.
From this we can generate statements such as this: "When system A was compared to a live performance, four out of 20 listeners could not reliably (using the threshold of p < .05 and other commonly used statistical methods) distinguish between the system and the live performance. With system B, 14 out of 20 could not tell which was which."
So here we have a "scientific fact" that in this listening session system B fools more people than system A. It's not a subjective listening test.
And of course from this you can look for correlations with technical measurements of systems A and B.
If no rigor is applied in factoring out the possible cable/speaker/amplifier system issues it's just an anecdotal folklore generator. If you dig down to what is necessary to truly isolate a single aspect of something it becomes a tedious and boring time sink (virtually no one does it). That is (real story), "The amplifier was oscillating at 60MHz", "so what, no one can hear 60MHz.".
Sometimes in medicine and other sciences, all we have for detectors is humans. Makes it tough, but we just have to deal with it.
In bear's case, let me propose a thought experiment that I would not carry out due to ethical concerns. Suppose we have two amplifiers with very low distortion numbers, bear claims to hear a difference, but we don't think so. So, what we do is hook him up to something equivalent to lie detector to measure physiological responses. Every time we play through amplifier A, we give him a little electric shock, so we train his System 1 to respond in a way we can measure with our instruments. Then, once so trained, we play amplifer A and omit the shock. Does bear's physiology show a fast stress response 100 ms before he even has time to know what he is hearing? If so, then his System 1 has learned to recognize something we didn't think was possible.
I don't particularly care if bear's feats of hearing acuity can be tested or not. Say there are 20 or 30 op-amps from several makers that should be excellent for a line stage. If bear wants to claim he can line all 30 up in order of preference and have it come out the same every day of every week, so what. I personally have a large cloud of uncertainty where on a select few I can't tell without looking or simply the differences are so small they are a don't care to me and they come and go. When I've mentioned any listening experience I've had my associated equipment or source dissed that's not going to change.
gpapag, be more astute than what you posted? you're a moderator, so you ought to be able to see beyond the surface, I would hope. next time, send me a PM should you have a troll-like question such as that?
As far as this issue, WRT to power amps, I did suggest testing WITH a load that is other than a resistive one? Perhaps then we'd see some sharper divergence between apparently well spec'd amps? Just another silly bear idea.
Scott, your experience is fine. But then simply say that is what it is. Generalizing that therefore this must be what is becomes where the problem lies.
I notice the difference between very fine woven Egyptian cotton sheets and the cheaper asian made cotton sheets. (for example - I don't care, but that's another story)
As far as this issue, WRT to power amps, I did suggest testing WITH a load that is other than a resistive one? Perhaps then we'd see some sharper divergence between apparently well spec'd amps? Just another silly bear idea.
Scott, your experience is fine. But then simply say that is what it is. Generalizing that therefore this must be what is becomes where the problem lies.
I notice the difference between very fine woven Egyptian cotton sheets and the cheaper asian made cotton sheets. (for example - I don't care, but that's another story)
Bear, your experience is fine. But then simply say that is what it is. Generalizing that therefore this must be what is becomes where the problem lies.
BTW I do draw the line at directional wire, etc. It isn't, period.
BTW I do draw the line at directional wire, etc. It isn't, period.
Actually it is an issue of damping factor decreasing with increasing frequency in some amplifiers. As a result they do not control the mechanical ringing in the higher frequency drivers. Shows up in the waterfall plots.
What I refer to as sibilance is a resonance causing certain frequencies to present themselves for longer than they should.
Best example was a tour of a major loudspeaker manufacturer's facility. In the anechoic chamber the guide's voice had a lisp not present elsewhere. It was one of the older chambers with the absorption fiberglass constrained by hardware cloth. Nice very narrow band resonance. Hard to detect by 60's generation standard swept frequency tests.
We can train rats!
This also in response to Mark's earlier post on the experiment he did not want to perform on humans. A rat's System 1 at a fundamental level is unlikely to differ much from that of man.
for animal models in auditory research its Gerbils: https://www.google.com/webhp?source...1&espv=2&ie=UTF-8#q=mongolian+gerbil+auditory
No the researchers sometimes get affectionately attracted to the rats, so it is better to use lawyers.
That is true in some cases, amplifier oscillating at peaks for instance another, or how about V/I capabilities. Are we restricted to picking amplifiers with exactly the same current drive capabilities? There's also zero feedback amplifiers that usually have poorer but constant damping factor.
Isn't blind power amplifier rolling a powerful tool in figuring out whats going on?🙄
I'm sorry to hear your equipment was dissed. I don't think disrespectfulness and other types of rudeness are useful for developing productive conversations on the forum. Same thing for dissing people who claim to hear things that most people can't. They may or may not be able to, but dissing them is usually not the best way forward.
Regarding whether or not you care about some of this, it seems perfectly fine if you don't care. I don't care much about new developments in radiotherapy delivery technology either. Had enough of it over the years probably to last me the rest of my days. But, I think some people around here participating in this thread, or maybe just lurking, probably do care about what we have been discussing. So, my comments in response to yours may not be totally directed at you personally, and if that were really needed, private communications might be more appropriate.
For people who are interested, I would just like to see the ball moved down field a little further. Maybe if we can find some evidence of exceptional distortion hearing acuity in a few people that would serve to reduce whatever tendency there may be to ridicule them. Also, if we can show its possible for a few people to hear very low level distortion, maybe that will spur some interest in finding some better ways to measure it. Little baby steps at a time is okay, if it keeps even a little forward momentum going. Over time, some benefits may accrue. I'm sure many in the pro audio community would be interested. Also, maybe eventually some well-to do audiophile will want to fund some research grants to serious people who can make some more progress. But first we have to show it's not impossible. If we can do that, the other steps will follow.
Last edited:
In my playground we use 500MHz oscilloscopes to check the output. Every so often we do see some units do nasty stuff. These units get returned. Some brands this can run to a few percent of the same model.
I did have one new product sample come in a few years back that was quite horrible. Stopped using that brand.
Ringing does show up on normal proof of performance testing.
I have always wondered why OP-amps are used at all. I can see they are justified when there are low power or voltage capabilities.
But in, let's say a pre-amp, with enough power, instead of these high NFB nervous little buggars, why not use simple class A GE or GC stages?
1. Unity gain, bias a mosfet as a sorce follower to around 30mA and be happy.
2. Gain, bias a mosfet as a common source to around 30mA and be happy.
Of course, the "power plant" must have a good regulation.
The 2:nd order harm will be noticable ( some 0.1% ) but due to the lack of NFB the whole thing will be stable and will also roll of in a very controlled manner.
Subjectively, this solution sounds much better than any high priced OP-amp.
But in, let's say a pre-amp, with enough power, instead of these high NFB nervous little buggars, why not use simple class A GE or GC stages?
1. Unity gain, bias a mosfet as a sorce follower to around 30mA and be happy.
2. Gain, bias a mosfet as a common source to around 30mA and be happy.
Of course, the "power plant" must have a good regulation.
The 2:nd order harm will be noticable ( some 0.1% ) but due to the lack of NFB the whole thing will be stable and will also roll of in a very controlled manner.
Subjectively, this solution sounds much better than any high priced OP-amp.
- Status
- Not open for further replies.