Which companies have the deepest roots in science of audio?

[Gregadd]

Actually I see the premium parts upgrade offered on the the Atma-Sphere website for the MA-1 not the MA-2. Teflon CAPS are standard on the MA-2. That might be considered a premium part.

 

[Phelonious Ponk]

I''m still learning about the HK 990. I never said anything about "audiophile parts" in either amp. I did say the the Atma-Sphere offered more premium parts as an option. Maybe that's where you went astray. Interestingly the Harmon Kardon website offers much more information than it did when I first discovered the product.
http://www.harmankardon.com/EN-US/Products/Pages/ProductDetails.aspx?PID=HK 990

I said an "impressive set of stats and parts." If you look at the insides you will see Siemens caps and very impressive dual toroidal transformers. They claim low negative feedback. My point is you can't get those kind of stats from ok parts and a casual approach.

Got it. I think it is a matter of semantics. I would consider normal parts those that deliver the target specifications and meet the design goals. "Premium parts," as in the Atma-Sphere options, would be, in my view, things like expensive audiophile caps and resistors, silver wire, etc. There is no doubt that the HK uses good components in the 990. I'd be very surprised to learn that they use what most audiophiles would call premium components. But I could be wrong. It has happened before.

Tim

 

[microstrip]

... "Premium parts," as in the Atma-Sphere options, would be, in my view, things like expensive audiophile caps and resistors, silver wire, etc.
Tim

Tim,
Are you considering that premium parts - "non normal parts", since we consider the others normal parts - are the parts which affect the sound in a way that it can not be shown by a classical measurable way how they are doing it?

 

[Gregadd]

I would just add meets or exceeds the required or target specifications.

The problem is arriving at a consensus as to what that means. Hence the ambiguity of the definition. Obviously if your philosophy is that only certain parameters matter and that they are easily achieved you are going to have problems with my definition.


Both companies have gone to great lengths to achieve their results. They have taken different paths. Both are rooted in science.

 

[microstrip]

I would just add meets or exceeds the required or target specifications.
The problem is arriving at a consensus as to what that means. Hence the ambiguity of the definition. Obviously if your philosophy is that only certain parameters matter and that they are easily achieved you are going to have problems with my definition.

Greg,
I do not know of any of us not having problems with your definition ... As far as I know, no one in this forum has been able to define the "required or target spefications in high-end audio".

BTW, most components used as premium parts were not developed for audio, but for other critical applications - e.g. Teflon capacitors were used in precision peak holder circuits due to their low leakage, but only now are manufactured with large values for audio. Not to speak about all the high reliability and stability military components.

 

[Ron Party]

Both companies have gone to great lengths to achieve their results. They have taken different paths. Both are rooted in science.

Greg, I think you nailed it, not just for those 2 companies but for many. It should almost go without stating that many (most?) use science to varying degrees in the design, manufacture, testing, and tweaking - did I leave something out? - of their products.

In the marketing department, we also may see different levels of emphasis on the role science played.

 

[Gregadd]

...voicing might be a word you are looking for. Even though listening is a qualitative analysis, the resuting adjustment is quantitative. For example determing how much feedback to use.

 

[Ethan Winer]

Any part in any circuit needs to meet several engineering parameters. So once we find parts that meet these parameters, we don't expect that more expensive parts will add anything (except cost) to the product.

Exactly. For most circuits, standard 5% carbon film resistors (2 cents each) are fine, unless it's a balanced input stage where higher tolerance is needed. Or a MM photo input where low noise metal films might be better. Same for normal capacitors (6 cents to 14 cents each) of whatever type is appropriate.

--Ethan

 

[Ethan Winer]

Are you open to the idea that some parts may measure identically yet sound different?

Absolutely not. Assuming the right things are measured. If two parts measure identically in all regards, how could they possibly sound different? This is a serious question.

Or 2 digital filters designed (at the same cost and same engineering rigor) to produce different sounds?

There are many types of digital filters, so I don't understand the question.

--Ethan

 

[Ethan Winer]

One additional thought on "parts" quality that came up earlier today in another forum. Someone asked about rock wool versus rigid fiberglass for making bass traps. Both of these are about the same acoustically, but rigid fiberglass costs 3 to 4 times more. However, rigid fiberglass is a much "better" material because it has a smoother front surface and doesn't look lumpy behind fabric, and it doesn't crumble over time, or sag when panels are hung from a ceiling. It's also a lot easier to work with and cut accurately etc.

So even when two parts are audibly the same, there are still reasons to use better parts. Same for parts where life expectancy is a factor, such as potentiometers and switches. Higher quality pots and switches also feel better, which should not be dismissed. I recall hearing of research Toyota did years ago about how important the "feel" of door levers etc is in customer satisfaction. I like the feel of viscous-damped pots and solid feeling switches too. I wouldn't pay $100 per switch! And I wouldn't expect even a tiny difference in sound quality. But "nice stuff" is always nice to have!

--Ethan

 

[caesar]

Absolutely not. Assuming the right things are measured. If two parts measure identically in all regards, how could they possibly sound different? This is a serious question.



There are many types of digital filters, so I don't understand the question.

--Ethan

I guess you not allowing the possibility that something measuring virtually identically to 0.00001 can sound different (as the President of Bryston and NOT a shyster told me some of their stuff does) pretty much ends the conversation.... But what if they really did? Are you not potentially closing your mind to a factor that you did not consider?

As to the digital filters, I am talking about flipping a switch on a dcs or an esoteric cd player and getting different sonic signature from each button press. Some will sound overly detailed, while others may sound snappy with a lot of rhythm and pace.

 

[Gregadd]

Exactly. For most circuits, standard 5% carbon film resistors (2 cents each) are fine, unless it's a balanced input stage where higher tolerance is needed. Or a MM photo input where low noise metal films might be better. Same for normal capacitors (6 cents to 14 cents each) of whatever type is appropriate.

--Ethan

I don't mean to be disrespectful. It's just that statement is so qualified that lawyers would call it illusory. I don't know what fine means or normal. Are those scientific terms?

 

[Ethan Winer]

I guess you not allowing the possibility that something measuring virtually identically to 0.00001 can sound different

Correct. Again I ask, if two components measure exactly the same, how could they possibly sound different? What exactly would account for the difference in sound?

Are you not potentially closing your mind to a factor that you did not consider?

My mind is wide open. All I need is proof. In this case, side by side measurements of the same audio device with one component used inside versus another component. Or a blind test showing that anyone can reliably identify a difference.

As to the digital filters, I am talking about flipping a switch on a dcs or an esoteric cd player and getting different sonic signature from each button press. Some will sound overly detailed, while others may sound snappy with a lot of rhythm and pace.

What switch? What is the switch labeled? What does the switch purport to do?

--Ethan

 

[Gregadd]

Let me try to explain a little better. Let's suppose we have two integrated solid state integrated amps using standard parts with a given tolerance and one built to the nth degree. The one built to the nth degree sounds decidedly better. For the sake of argument it measures markedly better. The manufactures of the better one tells me that it sounds better because it is built to mil-specs. The maker of the lesser product says there is no way the other product could sound better because even thought his does not use the same parts his product measures as well as is reasonably practicable to affect the sound.

What is the answer?
As I have pointed out before don't fight the hypothetical,
Assume the parts are measurably better.
That unt A does sound demonstrably sound better than unt B.
How are we going to determine the cause of the superior sound?
We know its not magic.
Is it possible that the designer of the superior unit is just a better designer and could have achieved the superior sound without the better parts?
Or is it that the better parts did in fact result in better sound?

 

[Ethan Winer]

I don't know what fine means or normal. Are those scientific terms?

They could be scientific terms.

A "fine" resistor or capacitor means audio passing through it is not degraded audibly. Whether by an audibly large change in frequency response, an audible increase in the noise floor, or audibly more distortion. Note my repeated use of "audible" and "audibly." We can measure stuff accurately to well below what is audible. So if the frequency response changes 0.01 dB at 30 KHz, that's not audible. Same for a 0.001% increase in distortion. It's easy to measure that with the right test gear, but nobody will ever hear it.

--Ethan

 

[Ethan Winer]

What is the answer?

Proof in the form of a blind test.

How are we going to determine the cause of the superior sound?

Once superior sound is actually proven with a blind test, we can find out why by measuring!

--Ethan

 

[Gregadd]

I think we have been here before.

 

[caesar]

Correct. Again I ask, if two components measure exactly the same, how could they possibly sound different? What exactly would account for the difference in sound?



My mind is wide open. All I need is proof. In this case, side by side measurements of the same audio device with one component used inside versus another component. Or a blind test showing that anyone can reliably identify a difference.



What switch? What is the switch labeled? What does the switch purport to do?

--Ethan

In the case of dcs Puccini, a choice of four reconstruction filters is provided for CD playback, these areselected by the Menu button. Some Esoteric models have something similar.

Going back to measurements, and I am not trying to be disresctful with this analogy, consider an individual standing and looking out into the horizon and measuring a piece of land. The earth looks flat. But as science progressed we found out it's not. What if there are factors or measurements that have not yet been discovered? Again, I am not being disrectful with this either, but your speakers cost $500, as you mention in another thread. Say you take hold of a most revealing speaker like a Martin Logan CLX electrostat and run it with a good measuring amp in your home. Are you open to the fact that you may be able to hear more musical information and are missing something in your analysis?

 

[Ethan Winer]

In the case of dcs Puccini, a choice of four reconstruction filters is provided for CD playback, these areselected by the Menu button. Some Esoteric models have something similar.

Well, if four different filter types are offered, then maybe they do sound different. I have no idea. I've never heard of a DAC offering user-selectable filter types. All the high-end pro gear I'm aware of employs whatever single filter type the designers felt works best.

consider an individual standing and looking out into the horizon and measuring a piece of land. The earth looks flat. But as science progressed we found out it's not. What if there are factors or measurements that have not yet been discovered?

That's not a valid comparison for many reasons. This comes up a lot, and I always explain it the same way. I'm surprised it keeps coming up! Here's the proof:

If there were something else that "science" didn't know to look for when measuring, it would have been revealed long ago as a residue in a null test. Nulling goes back 50 years or more, and is the basis for the first distortion meters sold by Hewlett-Packard. You null out the test frequency, and whatever residual remains is the distortion. These days nulling is common, and anyone with audio editor software can do it. So I am 100 percent totally confident there's no "curved earth beyond the horizon" that hasn't yet been discovered.

your speakers cost $500 ... Say you take hold of a most revealing speaker like a Martin Logan CLX electrostat

Why do you think my Mackie HR624 speakers are not every bit as "revealing" (whatever that means) as an ML CLX? How do you know the Mackies are not even more revealing?

Are you open to the fact that you may be able to hear more musical information and are missing something in your analysis?

Not really. It's easy to prove that test gear can measure to far better accuracy and resolution than even the finest ears, and test gear is much more repeatable. So if I wanted to compare the improvement of using one resistor or other component compared to another in a circuit design, I'd simply measure all the relevant parameters with each proposed component. Yes, it really is that easy. Professional audio circuit designers do this every day.

--Ethan

 

[microstrip]

Not really. It's easy to prove that test gear can measure to far better accuracy and resolution than even the finest ears, and test gear is much more repeatable. So if I wanted to compare the improvement of using one resistor or other component compared to another in a circuit design, I'd simply measure all the relevant parameters with each proposed component. Yes, it really is that easy. Professional audio circuit designers do this every day.
--Ethan

According tho your opinion what are the "relevant parameters" for capacitors and resistors?