Skip to main content ↑ ↓

Notice

Please note that most of the software linked on this forum is likely to be safe to use. If you are unsure, feel free to ask in the relevant topics, or send a private message to an administrator or moderator. To help curb the problems of false positives, or in the event that you do find actual malware, you can contribute through the article linked here.
Topic: Richard Clark $10,000 Amplifier Challenge (Read 12396 times) previous topic - next topic
0 Members and 1 Guest are viewing this topic.

Richard Clark $10,000 Amplifier Challenge

Reply #25 –
No, the other thread was about flat frequency response for speakers, not for amplifiers. Guys here are saying "no, no, amps must not have non-flat response. I don't see the harm in it.

But the thing about the Richard Clark testing is that it's not something that is reproducible in people's homes. People aren't going to sit there with a resistor fiddling around trying to match two different frequency responses of two different amplifiers.

It's a highly contrived set of test conditions.

Richard Clark $10,000 Amplifier Challenge

Reply #26 –
Well when I first heard about the challenge I was surprised that he even accepted such broken amplifiers. I guess he also wanted to show that you can downgrade a $100 amp to a $5000 "high-end" one with a few dollars worth of passive components. 
Similar to Carver, but far less elaborate.


Quote from: Rich B on
Guys here are saying "no, no, amps must not have non-flat response. I don't see the harm in it.

WHAT? People have pointed out that they'd calls such amps broken.

I could say that the harm is pulling money out of confused people's wallets, reinforcing the designers of said broken amps, helping out an entire industry that is based on misinformation and mediocre but expensive products, and so on. Happy now?


Quote from: Rich B on
But the thing about the Richard Clark testing is that it's not something that is reproducible in people's homes.

It is. Get two proper amps, compare them in their linear operating range, level-matched, blind.


Quote from: Rich B on
People aren't going to sit there with a resistor fiddling around trying to match two different frequency responses of two different amplifiers.

That's only necessary if you have an amp with rolled-off treble in the audible range, high output impedance ... ("broken").
Laziness is not an excuse btw. I could say that doing the statistical analysis of the test results is not something people would do. So what?
Most people wouldn't even get two amplifiers at the same time to compare them, because they trust that they have not been sold a broken product.
"I hear it when I see it."

Richard Clark $10,000 Amplifier Challenge

Reply #27 –
So as far as amplifier "sound" goes, expensive parts don't have any effect on quality below clipping? So if its all frequency response linearity, then better quality caps, transistors, etc etc, is not going to affect the sound?

From the Clark testing, it seems that frequency response is basically the determining factor that results in people hearing differences.

Richard Clark $10,000 Amplifier Challenge

Reply #28 –
You can measure the effect of different parts, but then again you can measure stuff far beyond what is audible to your ears.

Also, I think I've said this before, expensive parts are a complete waste if you have a bad design.
Whenever I see manufacturers advertising what expensive parts, special topology .. they use, I am careful. I don't want to know how much money they spent on resistors, I want to know if the output is clean and powerful and what other features the amp has. That's what the amp is being used for and what I'm paying for.
"I hear it when I see it."

Richard Clark $10,000 Amplifier Challenge

Reply #29 –
Quote from: xnor on
From a practical point of view, you can use a flat amplifier in any system configuration, it is always going to sound the same. But as soon as you start using "broken" amps, that might interact very differently with different loads, you are making your own life harder than it needs to be. You basically start a juggling act that is completely unnecessary.


This must be exactly what is desired by people who want to juggle those components, at vast expense, and then brag about the synergy they have achieved. The synergy they have achieved. Audiophile ego.
The most important audio cables are the ones in the brain

Richard Clark $10,000 Amplifier Challenge

Reply #30 –
Quote from: Rich B on
So as far as amplifier "sound" goes, expensive parts don't have any effect on quality below clipping?
It depends on the design. You can design an electronic circuit so that minute changes in almost any component can audibly change the sound. Or you can design a circuit that is in some sense self correcting and/or comparatively insensitive to moderate changes in components. Expensive components are needed to make the first approach work, but wasteful with the second.

Circuit design is a skill, but there are plenty of good examples available.

None of it is magic. Components are manufactured to known tolerances. Circuit simulation software will correctly tell you what happens to the output when components hit the edge of these tolerances. Including each component's non-ideal properties makes the simulation even more true to life. (e.g. every component has resistance, capacitance and inductance - even if it is supposedly only one, or even none, of those things - where it is at all significant it is good practice to account for this and include it in the circuit simulation to find out what will happen).


Quote
So if its all frequency response linearity, then better quality caps, transistors, etc etc, is not going to affect the sound?
See above.

Quote
From the Clark testing, it seems that frequency response is basically the determining factor that results in people hearing differences.
Essentially yes these days. But with "broken" amplifiers, the resulting audible frequency response is a result of the interaction of their circuitry and the speaker.


What the stereophile article didn't cover openly is that many people prefer damaged sound to pure sound. Some types of distortion are subjectively preferred to a clean signal. Some types of frequency response error are sometimes preferred to an accurate signal. Where they drive themselves insane is in buying equipment that should have a specific simple job, and unconsciously hoping it will do these things too. Let the amplifier amplify. Use an EQ if you don't like the frequency response. Add some distortion if you like it (it's really easy in the analogue domain). What subjectively works is very recording dependent, so it's insane to have it built into the core equipment itself (IME). If people were more honest, there would be more of an open market for boxes that damaged sound (rather than the existing audiophile market for boxes that damage sound while claiming not to  ).

Cheers,
David.

Richard Clark $10,000 Amplifier Challenge

Reply #31 –
Quote from: 2Bdecided on
Quote from: Rich B on
So as far as amplifier "sound" goes, expensive parts don't have any effect on quality below clipping?
It depends on the design. You can design an electronic circuit so that minute changes in almost any component can audibly change the sound. Or you can design a circuit that is in some sense self correcting and/or comparatively insensitive to moderate changes in components. Expensive components are needed to make the first approach work, but wasteful with the second.

Circuit design is a skill, but there are plenty of good examples available.

None of it is magic. Components are manufactured to known tolerances. Circuit simulation software will correctly tell you what happens to the output when components hit the edge of these tolerances. Including each component's non-ideal properties makes the simulation even more true to life. (e.g. every component has resistance, capacitance and inductance - even if it is supposedly only one, or even none, of those things - where it is at all significant it is good practice to account for this and include it in the circuit simulation to find out what will happen).


Quote
So if its all frequency response linearity, then better quality caps, transistors, etc etc, is not going to affect the sound?
See above.

Quote
From the Clark testing, it seems that frequency response is basically the determining factor that results in people hearing differences.
Essentially yes these days. But with "broken" amplifiers, the resulting audible frequency response is a result of the interaction of their circuitry and the speaker.


What the stereophile article didn't cover openly is that many people prefer damaged sound to pure sound. Some types of distortion are subjectively preferred to a clean signal. Some types of frequency response error are sometimes preferred to an accurate signal. Where they drive themselves insane is in buying equipment that should have a specific simple job, and unconsciously hoping it will do these things too. Let the amplifier amplify. Use an EQ if you don't like the frequency response. Add some distortion if you like it (it's really easy in the analogue domain). What subjectively works is very recording dependent, so it's insane to have it built into the core equipment itself (IME). If people were more honest, there would be more of an open market for boxes that damaged sound (rather than the existing audiophile market for boxes that damage sound while claiming not to  ).

Cheers,
David.


Why do you keep using the word "damaged"? The equipment isn't damaged. It's designed. If the response is non-flat, it's not damaged, it's just different.

I also don't know how people can throw around terms like "broken". Clearly the equipment works just fine. But I think it is wrong to suggest that anything deviating from a flat frequency response is "broken" or "damaged". You may as well claim that many Class D and tube amps are damaged, but that's a ridiculous use of the word.

Richard Clark $10,000 Amplifier Challenge

Reply #32 –
I'm adopting "broken" from this (and other) threads. In the context of 2014, it's fair (if a bit provocative) to describe an amplifier with a non-flat frequency response as broken.

I 100% stand by "damaged". I'm talking about the sound, not the equipment. The aspirational goal of hi-fi is to reproduce something accurately - i.e. without change. Changing it is damaging it. Now, there are reasons you might want to change it, but the core circuitry of the amplifier (i.e. the part that's still in play even with tone controls etc defeated) is not the place to be doing it.

Cheers,
David.

Richard Clark $10,000 Amplifier Challenge

Reply #33 –
btw, I sense in all your posts here (apologies if I'm wrong) a kind of incredulity that we could believe this stuff. I'm guessing you come from a typical audiophile background (again, apologies if I'm wrong), and have even wondered if there's a thread somewhere on an audiophile forum where you re-post selected replies for everyone there to have a good laugh at.

Truth is, people here do believe (many of us know) that audio can be dramatically improved. Where we would differer from "typical" audiophiles is in what we believe can be dramatically improved. You see, to be commercially viable, the audiophile industry has to maintain the impression that every component is infinitely (and audibly) improvable. Meanwhile, in the real world (IMO!), some parts of the chain are already more than good enough and don't need that much money spending on them, while others are woefully inadequate. Interestingly, many of the parts that are inadequate are out of the control of typical audiophile companies, which partly explains their behaviour.

Cheers,
David.

Richard Clark $10,000 Amplifier Challenge

Reply #34 –
Quote from: Rich B on
Why do you keep using the word "damaged"? The equipment isn't damaged.


The equipment may be designed in some very broad sense, but what it does to the the music you play through it is not the result of any intelligent design. It is the random results of combining an arbitrarily source impedance with a speaker impedance curve that just happened.

Quote
If the response is non-flat, it's not damaged, it's just different.


At the very least the poorly-designed amp's FR is different with every different speaker that you attach to it.  Its bad enough that we don't have very good control over speaker FR as used, but adding random FR variations due to varying amp source impedance interacting with varying speaker impedance is way out of control.

Quote
I also don't know how people can throw around terms like "broken".


I don't know why people waste their time with such highly substandard junk.

Quote
Clearly the equipment works just fine.


No it does not. It claims to be an amplifier, but it surreptitiously adds a randomly-chosen equalizer. Frequency Response De Jour, what a concept! 

Please show me where the word amplifier includes a non-adjustable random equalizer function! ;-)

Quote
But I think it is wrong to suggest that anything deviating from a flat frequency response is "broken" or "damaged".


I think its wrong to secretly play craps with a system's frequency response, given how sensitive the ears are to such things.

Quote
You may as well claim that many Class D and tube amps are damaged, but that's a ridiculous use of the word.


OK, lets haggle over words. You object to damaged, so lets go with broken as designed or in short, substandard.

Class D amps that have high source impedances are substandard designs. There are ways to avoid that problem but as usual, not everybody got the memo.

Richard Clark $10,000 Amplifier Challenge

Reply #35 –
Quote from: Rich B on
No, the other thread was about flat frequency response for speakers, not for amplifiers.



But this isn't the only forum you're on.  And whether 'amps  sound different' certianly has been discussed to death on the other forums that I know you are on. 

So why do you continue to indulge in these exercises?


Quote
Guys here are saying "no, no, amps must not have non-flat response. I don't see the harm in it.


Well the 'high end' loves guys like you.  What a brilliant market you make.  The high-end can claim to give every component a 'sound'.  (Rather than just sell you one EQ unit that could replicate all those different sounds).  And sell them to you.

Quote
But the thing about the Richard Clark testing is that it's not something that is reproducible in people's homes. People aren't going to sit there with a resistor fiddling around trying to match two different frequency responses of two different amplifiers.



And neither is Clark, in most cases (e.g., solid state compared to solid state).  What he will do in those cases, is level match the outputs. Because there is no need to match the EQ.

Quote
It's a highly contrived set of test conditions.



Your objections are even more highly contrived.

How many people actually compare  amps at home? 
How many do it sighted?
How many level match?

Don't those constitute 'contrived' conditions too, compared to the run of the mill?

Richard Clark $10,000 Amplifier Challenge

Reply #36 –
Quote from: Rich B on
Guys here are saying "no, no, amps must not have non-flat response. I don't see the harm in it.


Electric guitar players, anyone? Then distortion is a point in itself. It is about creating a sound, not about reproducing it.

Instrument amp specialist Mesa/Boogie did this thing: http://www.mesaboogie.com/Product_Info/Out...n/the_baron.htm
Of course that is about tweaking the sound to what you like.

Richard Clark $10,000 Amplifier Challenge

Reply #37 –
Quote from: Rich B on
So as far as amplifier "sound" goes, expensive parts don't have any effect on quality below clipping?


What is an "Expensive Part"?  Are thy all the same? ;-)

Like many things, parts quality has been evolving over the life of the art of electronics.

When I was a boy, the mainstream resistor in commodity audio gear was a crude 20% tolerance carbon composition part with modest reliability and poor noise performance not to mention the broad tolerance.  The upscale part was a 1% or 5% carbon or metal film resistor that cost many times as much.  Today the mainstream resistor in commodity audio gear is 5% carbon film resistor that in practice has usually tighter tolerances than specified. 5% and 1% parts and parts using more sophisticated technology such as metal film elements command only nominal price premiums.

Yet other than the broad tolerances which could audibly impact the performance of the tiny minority of equalization networks and the like,  the passive parts of old were not the main stumbling block to the performance of finished equipment. The main stumbling blocks to sound quality were the basic  technologies such as tubes, vinyl, analog tape, and AM/FM radio.  I remember doing a wall-to-wall parts upgrade on a integrated amp and measuring no change in overall performance.

One of the things you almost never see is before and after measurements and reliable listening tests related to the alleged parts upgrades.  Remember that the commodity parts of today are better components than the premium parts of not that long ago.

Real engineers know about sensitivity analysis - determining the end results of changes to internal parts. Often big changes inside make no difference outside. I've done what seemed like major op-amp upgrades on pro audio gear with absolutely no measurable differences at the output terminals. 

The history of electronics includes Earl "Madman" Muntz. http://electronicdesign.com/boards/whats-a...ff-anyhow\

In brief Earl Muntz was a professional used car salesman who looked at the then-mainstream RCA TV set and decided that it was vastly overbuilt for use in metropolitan areas. Muntz did not merely upgrade or downgrade parts quality. He totally removed parts  by the dozen - whole functional units of the TV set went missing. Entire amplification stages went away. The results were a far less complex, far less expensive TV receiver that still delivered a good picture for most customers.  I believe that the tube count was nearly cut in half, so the number of potential points of failure was cut in half. Due to the lack of redundancy the consequences of probable failures were more severe, but the failures themselves were rarer.

Quote
So if its all frequency response linearity, then better quality caps, transistors, etc etc, is not going to affect the sound?


In many cases the only thing that is better about those parts is their profits for the retailer, and the hype.  Remember that even making a measurable difference is usually elusive., and just because there is a measurable difference proves nothing about actual audible quality.

Quote
From the Clark testing, it seems that frequency response is basically the determining factor that results in people hearing differences.


This has been known for decades. I've already explained the reason why in this thread - frequency response differences usually affect sound quality at all SPLs, while nonlinear distortion is inherently level-sensitive.