Yes, your comment seemed to be confused about the context by which Ed was making his.
I grew up around these types. Crazy huh! Later I came to find that they don't care about really good playback because they do so much live playing. They get their musical fix at the source.
In another case of how Stoic hardcore physics types are
Billion-tonne IceCube: Sterile neutrinos do not exist ? The Register
20 years in preparation. A year spent measuring light flashes in a billion tonne block of ice. Crunch the data...
...And a null. Which actually makes them happy. As Randall Monroe says 'science, it works bitches'
Billion-tonne IceCube: Sterile neutrinos do not exist ? The Register
20 years in preparation. A year spent measuring light flashes in a billion tonne block of ice. Crunch the data...
...And a null. Which actually makes them happy. As Randall Monroe says 'science, it works bitches'
I've found it's often composers of "classical" music that are quite undemanding of audio quality, as long as they can hear what they have written. This led to the popularity of multi-miked recordings of orchestral pieces, with sometimes as many as a microphone for each stand, and the practice of the recording engineer armed with the marked-up score goosing up the levels for a given passage. Balance problems in my orchestration? What balance problems?? Listen---there's the clarinet solo, now the string section to the fore.
When Keith Johnson and Stan Ricker traveled to the USSR with the Sheffield folks to make some wonderful recordings of Russian and American pieces with an American and a Russian conductor, The Moscow Sessions, they encountered a setup with a huge number of microphones and immense multichannel consoles. When Keith used just a small handful of microphones and minimal mixing, many of the musicians were very concerned that they would not be heard anymore.
When they were ushered into the recording booth and heard the results, they were reassured. Those recordings hold up well today.
Why would you ask me? 😱
Ask SY, ask Wurcer, ask one of the other "experts"!
I have no way to predict... I'd have to actually listen to it myself. But that's just my personal delusions, we're after solid, hard science and information. Surely this has been done, or else there is solid technical and engineering reasons to say one way or the other...right?
Actually having spent time with some of "them" of various primary "genre", the answer is somewhat different according to what I have been told.
What happens is that professional musicians literally listen very differently. They are not listening to the performance the same way that a typical "audiophile" does - apparently the performance takes place "inside" their heads, not so much in the soundfield.
But it works out to be the same thing in practice, they're usually not terribly concerned about "hi-fi" in the same way.
Exactly.
How much of a difference would it need to become audible??
I say according to the problem as presented, the "THD" remains the same, that the difference(s) ought to be inaudible by the criteria of measurement(s), unless you now allow that minute alterations are audible.
I have no position on the answer, I only seek it.
_-_-
RNM concurs that this small of a change would (likely) be audible.
Just noting...
So, I go back to the notion of which two amplifiers will sound indistinguishable?? Two or more that is... And I made the question easy, I specified amplifiers that are different!! (not differing by parts selection, etc...)
So, let's consider another hypothetical case...
...given amplifier model "Boondoggle", we find that this particular amp has the lowest distortion at a particular bias point (found of course by measurement of THD vs. bias).
Now, let's say that the distortion figure works out to be say 0.003%. If we raise the bias higher the distortion can only increase, not decrease since we have found the lowest distortion bias point. Similarly, if the bias is decreased the distortion will also go up, not down. There will be two symmetric points on either side of the distortion minimum where the bias will be higher and equal when measured.
Will those two bias points when listened to sound the same??
or
Not sound the same??
OR will all three bias points sound identical.
At what point in the slide of the bias might an audible difference be heard?
And what sort of % change or crossing of a threshold will cause a difference to be heard?
Seems like a logical enough question...
...given amplifier model "Boondoggle", we find that this particular amp has the lowest distortion at a particular bias point (found of course by measurement of THD vs. bias).
Now, let's say that the distortion figure works out to be say 0.003%. If we raise the bias higher the distortion can only increase, not decrease since we have found the lowest distortion bias point. Similarly, if the bias is decreased the distortion will also go up, not down. There will be two symmetric points on either side of the distortion minimum where the bias will be higher and equal when measured.
Will those two bias points when listened to sound the same??
or
Not sound the same??
OR will all three bias points sound identical.
At what point in the slide of the bias might an audible difference be heard?
And what sort of % change or crossing of a threshold will cause a difference to be heard?
Seems like a logical enough question...
The notion that matching of components is key to performance enhancement is misguided, if taken at face value. If qualified by a particular topology then it can have bearing, such as the distortion associated with a differential pair, or the matching of complementary devices like P channel and N channel JFETs for the magnitude and current dependence of transconductance.
The more general notion of having a forward path with a lot of distortion versus one with a little, around which global feedback is applied, is often used as something of a strawman by the anti-feedback brigade. The notion is Who cares about open loop linearity if we can just beat the hell out of things with feedback?
One can construct topologies that behave badly, particularly under transient conditions, and these can exhibit ill behavior, but although instructive these are faulty designs and can be avoided.
The more general notion of having a forward path with a lot of distortion versus one with a little, around which global feedback is applied, is often used as something of a strawman by the anti-feedback brigade. The notion is Who cares about open loop linearity if we can just beat the hell out of things with feedback?
One can construct topologies that behave badly, particularly under transient conditions, and these can exhibit ill behavior, but although instructive these are faulty designs and can be avoided.
Brad, the question at hand is not the effect of matching or not... nor the gfb vs. zfb issues.
The underlying question is essentially, boiled down, is it possible for two amps (in this case according to Kindhornman's post) to sound different while measuring the same THD.
The feedback amount is somewhat less than the main point... one might be able to build an amp with highly paralleled output devices, vs one without same, keep the same topology and get reasonably close Zo and two different dB of feedback, I'd guess... the question is less about the technical reasons, and more of what IS or is not going to ever be audible... imho.
The underlying question is essentially, boiled down, is it possible for two amps (in this case according to Kindhornman's post) to sound different while measuring the same THD.
The feedback amount is somewhat less than the main point... one might be able to build an amp with highly paralleled output devices, vs one without same, keep the same topology and get reasonably close Zo and two different dB of feedback, I'd guess... the question is less about the technical reasons, and more of what IS or is not going to ever be audible... imho.
Last edited:
WRT miking for acoustic music, "point" recordings made with a stereo mike pair close together just make sense.
Even from amateur youtube videos made with a cell phone or camera I hear very enjoyable presentations.
This looks like an interesting mike:
https://www.neumann.com/?lang=en&id=current_microphones&cid=ku100_description
Even from amateur youtube videos made with a cell phone or camera I hear very enjoyable presentations.
This looks like an interesting mike:
https://www.neumann.com/?lang=en&id=current_microphones&cid=ku100_description
Bob Carver did this years ago at Stereophile- matching his amp to a "reference" amp and supposedly they were indistinguishable. I'm sure someone here as the details. The core was matching the frequency response under load with the same source. As I remember (too lazy to dig it up) the reference was a tube amp and his was solid state. I think the long term result was a Carver tube amp- if that's what customers, want build it. The core of this is matching the response in the full system, including the source impedance.
Yes I remember the Carver Challenge well. He did a nulling test on Haflerian lines and fiddled with components until the residual difference signal was negligible. The comparison amp I'm not sure of, and it was supposed to be a closely-guarded secret, but was probably a Conrad-Johnson (it might come up in a search).
This led to people asking, if he could do that, why didn't he make his own amps sound like that. Not too long after he did his Silver Seven series, of which he claimed he didn't expect to sell many. Actually the demand was said to be far in excess of expectations.
Demian,
Correct me if I get this wrong. If you set two amplifiers as Carver did so that the measured results of FR, input impedance, output impedance, harmonic structure and such you can in effect make two very different amplifiers sound the same.
Now what I think Bear is asking for is a standard, or limits that make it possible, whereby all amplifiers would sound the same or two commercial units that do indeed sound the same.
How I would answer that is that there are no true audio standards that says that commercial products have to sound a particular way. There are no requirements for output impedance for the source, fixed input impedance for amplifiers nor fix output impedance for the output. the harmonic structure or ratio of 1st to 2nd to 3rd etc is not fixed and therefor even a so called electronically correctly functioning amplifier or other component has no fixed target to be built to.
So to think that all components do in fact sound the same or are interchangeable seems to be a flawed idea or reference. That is not to say one is superior to another, just that in the end they can be slightly or very different in the end. We have no real standards besides something like 20-20Kz FR, flat phase response and let's say no THD above a certain amount. There can be other parameters that are specified but those do not determine that there is a standard sound quality or end product.
We aren't building medical devices that all must produce the same test results, nor telescopes that can't have aberrations that skew the optics of an orbiting telescope. We are talking about sound here, there are no perfect results, there are no controlled specification beyond a basic level and there are no rules as to allowed topology. There are no perfect amplifiers anymore than there are perfect apples, what would be required to have the perfect apple, you can't define that.
If you could say that the audio signal had to match the original live output of a real instrument and could actually do that we wouldn't be having any of these discussion but we all, or at least I think most of us understand there is no perfect reproduction system, none, not from JC or Wilson or anybody else. There is no perfect way to make a recording, there is just as good as it comes in many different colors. Choose you preference and understand that is what it is, preference.
Correct me if I get this wrong. If you set two amplifiers as Carver did so that the measured results of FR, input impedance, output impedance, harmonic structure and such you can in effect make two very different amplifiers sound the same.
Now what I think Bear is asking for is a standard, or limits that make it possible, whereby all amplifiers would sound the same or two commercial units that do indeed sound the same.
How I would answer that is that there are no true audio standards that says that commercial products have to sound a particular way. There are no requirements for output impedance for the source, fixed input impedance for amplifiers nor fix output impedance for the output. the harmonic structure or ratio of 1st to 2nd to 3rd etc is not fixed and therefor even a so called electronically correctly functioning amplifier or other component has no fixed target to be built to.
So to think that all components do in fact sound the same or are interchangeable seems to be a flawed idea or reference. That is not to say one is superior to another, just that in the end they can be slightly or very different in the end. We have no real standards besides something like 20-20Kz FR, flat phase response and let's say no THD above a certain amount. There can be other parameters that are specified but those do not determine that there is a standard sound quality or end product.
We aren't building medical devices that all must produce the same test results, nor telescopes that can't have aberrations that skew the optics of an orbiting telescope. We are talking about sound here, there are no perfect results, there are no controlled specification beyond a basic level and there are no rules as to allowed topology. There are no perfect amplifiers anymore than there are perfect apples, what would be required to have the perfect apple, you can't define that.
If you could say that the audio signal had to match the original live output of a real instrument and could actually do that we wouldn't be having any of these discussion but we all, or at least I think most of us understand there is no perfect reproduction system, none, not from JC or Wilson or anybody else. There is no perfect way to make a recording, there is just as good as it comes in many different colors. Choose you preference and understand that is what it is, preference.
Of course this is interesting... I remember it being discussed at the time.
However, if you go back father in time, allegedly people were "fooled" before, the famous Audio Research display at Grand Central in NYC, and of course Edison...
So here's an angle. IF you use DSP to match the frequency response, do you need to match the source impedance, and why? (I'm presuming Carver did not use EQ, but dorked with the feedback network in the amp??)
Can I use a series resistance for the output Z?
And would it fly today with hi-res sources and cleaner everything else??
(hey, everyone! still waiting for that list of amps and the specs...) 😀
_-_-
On the flipside, if two amplifiers are, in essence, VERY close to each other in all those performance metrics, then it's a safe assumption to make that they're going to sound very similar. Even if the means to get there are wildly different.
They may sound quite different when they go sideways, but that would also show up in an appropriately designed synthetic test.
As such, the example of Tomchr super low distortion composite vs Bob Cordell's super-low distortion designs vs whatever Douglas Self's most thorough blameless amp are (if it's even possible to gain match them) going to have a lot more in common with each other under light-to moderate loads (not clipping or getting close) than they're not. Someone who claimed huge differences in sound between the amps would first need to be validated via system interaction (namely input impedance as all these high FB designs have stupid-low output impedances), or RFI incursion.
They may sound quite different when they go sideways, but that would also show up in an appropriately designed synthetic test.
As such, the example of Tomchr super low distortion composite vs Bob Cordell's super-low distortion designs vs whatever Douglas Self's most thorough blameless amp are (if it's even possible to gain match them) going to have a lot more in common with each other under light-to moderate loads (not clipping or getting close) than they're not. Someone who claimed huge differences in sound between the amps would first need to be validated via system interaction (namely input impedance as all these high FB designs have stupid-low output impedances), or RFI incursion.
There is an egg before the chicken. They have the ability have the performance in their heads and fill in the missing sonic "data" (as Sy would call it) because of the constant "you are there" memory refresh that precedes.
I haven't just spent some time the circles of professional musicians, I was immersed in them from the age of 6 to the age of 22.
My dad;
https://www.discogs.com/artist/519314-Bob-Morin
3 of his best friends who I have known since that tender age of 6;
https://en.wikipedia.org/wiki/Ernie_Watts
Peter Robinson | Credits | AllMusic
https://www.discogs.com/artist/519315-Wolfgang-Melz
You should have heard how loud it was when they came over to rehearse. It was hard to concentrate on setting up Hot Wheels track when they were making so much racket 😉
Not so simple, IIRC two hard core experienced listeners admitted in the end with no peeking they could not tell the difference between two very different amplifiers. Why should we accept your constant toss off comments about NEVER hearing any components sounding the same when you discount any stories of the opposite happening. If I'm correct you asked John to clarify his comment and he didn't, if he does not peek some things sounds the same the differences disappear (they are not from the same reality).
You did allow for an exception, if the amplifiers were the same design and BOM from the same manufacturer, IMO this is just the point where someone walks in and puts a Shakti stone on one and now gee they sound different (sighted of course).
This distortion stuff as any kind of individual metric is stupid, going several orders of magnitude below any estimates derived in reasonable studies (save for Ed's numerology 🙂) is easily achieved. Please why don't you perform your feats of perception for an audience, it would be SO easy.
Last edited:
- Status
- Not open for further replies.