The State of High End Audio

[treitz3]

Umm, no. It's not weird at all. Some don't believe the increased complexity and cost of a fully balanced system is worth it in a typical home system. Not sure why that's so hard to understand, as it happens to be the case that many, many people use single ended systems without noise issues.

Maybe I think it's weird you'd want to opt for a more complicated, expensive way of doing things that offers no benefits whatsoever to 99.9% of people with normal home systems?

Hello Dave and good evening to you. I can agree with this. If there is no benefit, then why bother? Like I said before, it is situation, gear and system dependant. Blanket statements that non balanced IC's are antiquated and "less" than a balanced cable I can not.

For those that have a need for a balanced system, no problem. That's what they were designed for but there are many a great to outstanding systems out there that do not have "the need".

Tom

 

[rbbert]

If the components on each side of the cable are both balanced, that would seem to be the obvious way to connect them. If not, you can probably go either way depending on how it sounds?

 

[Folsom]

Balanced for proaudio is best. When you just need to run the cables and have everything work the first time, they're gold. They also do much longer runs of cable. And they're mixed in with microphones and old instrument amplifiers, with no care for placement. On top of that they have many different circuit breakers potentially going to different boxes all over, so the pin1 ground reference in balanced is more important for them, than say, us. The environment is rough to say the least. If they didn't have to be worried about so many aspects all at once single ended could work.

In theory balanced is better because it rejects more noise, but the question remains, what sounds best? Was the noise even noticeable or did your equipment reject it? What about the noise from the added circuitry? (you double your circuitry for balanced, basically) For example if you have all transformers on the inputs and outputs (I think Rowland did this on a line of his) then they don't even transmit the higher RF bands so the point is fairly moot. You have to remember while it is a music signal it's in a frequency much higher than our equipment tends to operate in, except DAC's. However it's very problematic if it couples, through say parasitic means, to our stereos in certain locations because at that point it becomes line level frequency (you can literally hear it in your speaker, the radio station). To do this it's typically looking for a low impedance path. Our amplifier inputs are ideally not low impedance because they'd be too hard for the source to drive. This actually helps reject RF from the input, and drives it to drain on the ground/shield which is usually a better path. So you see single ended are not just magical sponges for noise.

It's important to note some equipment sounds better with balanced IC's because they were designed to operate with them. It's not because they couldn't sound just as good with singled ended, it was a choice, and probably based on what they expect their equipment and customer base to be using for their typical application. It's not bad, it offers some reassurances like no 60hz loops, etc, to avoid conflicts with other manufactures.

 

[Blizzard]

Bruno Putzeys take on balanced connections:

 

[Folsom]

Obviously Bruno has a different approach than Lamm. Bruno has some of the lowest distortion possible, and balanced IC's preferred. Lamm has 1% distortion and single ended.

Please don't try and tell us one is wrong or right.

Bruno's example is the tip of an iceberg.

 

[Blizzard]

Obviously Bruno has a different approach than Lamm. Bruno has some of the lowest distortion possible, and balanced IC's preferred. Lamm has 1% distortion and single ended.

Please don't try and tell us one is wrong or right.

Bruno's example is the tip of an iceberg.

There's no wrong or right in audio. If people are willing to pay for it, and reap enjoyment from it, then it's right. Some people's favorite food may be Kraft Dinner with chopped up chicken wieners inside, for others it may be prime New York Striploin steak and Alaskan King crab. Neither is right or wrong.

 

[dallasjustice]

There's no wrong or right in audio. If people are willing to pay for it, and reap enjoyment from it, then it's right. Some people's favorite food may be Kraft Dinner with chopped up chicken wieners inside, for others it may be prime New York Striploin steak and Alaskan King crab. Neither is right or wrong.

Me personally, king crab flown in fresh from Alaska to Salt Spring Island, BC overlooking the water is pretty much right.

 

[Blizzard]

Me personally, king crab flown in fresh from Alaska to Salt Spring Island, BC overlooking the water is pretty much as right.

Sounds about right to me as well

 

[Gregadd]

I think audiophiles
worry our the noise floor constantly . We are constantly ridiculed because of it.
My straw vote. I heard no noise when Steve^s system was at idle. His room is dead quiet.

 

[Folsom]

I think audiophiles
worry our the noise floor constantly . We are constantly ridiculed because of it.
My straw vote. I heard no noise when Steve^s system was at idle. His room is dead quiet.

That's because it doesn't show up as artifacts unless something's wrong. There's noise, but whatever there is, is part of the music by the time it reaches you. But from the equipment that Steve uses, he has less noise in it than most people. Lamm is one of the few companies that does any type of main's AC filtration; particularly with linear power supplies.

 

[Whatmore]

Interesting article on balanced and unbalanced

http://sanderssoundsystems.com/tech...-balanced-vs-unbalanced-operation-white-paper

In short, the advantages of balanced systems will not be needed or used in most audiophile applications. But balanced operation still causes higher noise and distortion than what you find in the same electronics when run in unbalanced mode.

Now it must be said that the slight increase in noise and distortion caused by balanced operation cannot be heard by the human ear. It requires test equipment like a spectrum analyzer to reveal the difference. But if you don't want to compromise, and you want the most pure sound with the least noise and distortion, unbalanced operation is superior to balanced operation.

 

[amirm]

Interesting article on balanced and unbalanced

http://sanderssoundsystems.com/tech...-balanced-vs-unbalanced-operation-white-paper

As much as I have respect for Roger, this is a prime example of how this topic is misunderstood by even engineers. When we talk about noise, there are many versions of it. One for example is noise being picked up by the cable. In this discussion, we are *not* talking about that noise. Nor are we talking about noise inside of your electronics.

What we are talking about is that by mere fact that two power cords in two separate pieces of equipment creates two different ground potentials (voltages), by definition if you hook those two chasses together, current moves between them. If you then proceed to use unbalanced connection, this current becomes part of your signal. And because it runs at AC mains frequencies, it translates into hum.

This is *totally* induced noise by the user/system. It is on top of any and all other sources of noise. Its reason for existence cannot be eliminated. No matter what you do, the two chassis will be at different voltages. It is a problem created when we went from all-in-one systems to components. This problem was recognized and balanced connection were invented to eliminate it (plus other benefits I won't get into).

No goodness to a user comes from such a connection scheme. You are forcing the system into producing noise. Why on earth would you want to do that? You can't with a straight face.

Now if you are a designer/manufacturer, you have many reasons to want to avoid building a fully differential system. Such is life in mass market where component cost is everything. But in a $20,000+ amplifier or component? No way do I tolerate such compromises.

Given me both options and let me choose if it is not the cost that took balanced out of the system. Otherwise the arrow points directly at cost cutting. Or lack of understanding of the problem here which unfortunately is prevalent.

I say if there is one topic engineers like to forget about when they leave school is this one. It makes their heads hurt when they are taught the theory and walk out confusing all manner of grounding, transmission, emission, etc. This is behind so many bad designs and bad advice online. I myself had to relearn all of this.

 

[DaveC]

As much as I have respect for Roger, this is a prime example of how this topic is misunderstood by even engineers. When we talk about noise, there are many versions of it. One for example is noise being picked up by the cable. In this discussion, we are *not* talking about that noise. Nor are we talking about noise inside of your electronics.

What we are talking about is that by mere fact that two power cords in two separate pieces of equipment creates two different ground potentials (voltages), by definition if you hook those two chasses together, current moves between them. If you then proceed to use unbalanced connection, this current becomes part of your signal. And because it runs at AC mains frequencies, it translates into hum.

This is *totally* induced noise by the user/system. It is on top of any and all other sources of noise. Its reason for existence cannot be eliminated. No matter what you do, the two chassis will be at different voltages. It is a problem created when we went from all-in-one systems to components. This problem was recognized and balanced connection were invented to eliminate it (plus other benefits I won't get into).

No goodness to a user comes from such a connection scheme. You are forcing the system into producing noise. Why on earth would you want to do that? You can't with a straight face.

Now if you are a designer/manufacturer, you have many reasons to want to avoid building a fully differential system. Such is life in mass market where component cost is everything. But in a $20,000+ amplifier or component? No way do I tolerate such compromises.

Given me both options and let me choose if it is not the cost that took balanced out of the system. Otherwise the arrow points directly at cost cutting. Or lack of understanding of the problem here which unfortunately is prevalent.

I say if there is one topic engineers like to forget about when they leave school is this one. It makes their heads hurt when they are taught the theory and walk out confusing all manner of grounding, transmission, emission, etc. This is behind so many bad designs and bad advice online. I myself had to relearn all of this.

That's absolutely not true, you do not have a correct picture of what is happening in a single ended system. Your whole post is misinformation... It IS true that the ground on single ended cables carries a lot more than just the return signal, but under normal circumstances it doesn't couple to the signal, otherwise single ended systems would be completely unuseable.

The truth is, many people happily own and listen to single ended systems with minimal noise. One of my systems is ridiculously quiet, you can't hear noise from a 102 dB speaker with your ear INSIDE THE HORN. And that's with separate components, a tube pre and amp, that I built myself using simple star grounding.

A far larger issue wrt noise is the typically excessive gain of modern systems, but with single ended vs balanced, both can be done right, it's the implementation that's important much like most things in audio. Whenever someone makes a blanket statement "XXX is always better" you know it's information you can safely disregard 99% of the time.

Another data point is Odyssey, who makes amps that won TAS best product of the year award two years running, he recommends single ended connections. Amir, there are just too many people who manage to do exactly what you are saying is impossible, and in fact PREFER single ended systems. Your attitude that they just don't know better (and you do) is a bit ridiculous. As far as I can see, you are the one who is mistaken on this topic, and you're contradicting a lot of folks who actually design and build quality gear. Your ideas simply don't match up with reality.

Not only that, balanced systems were not invented to deal with SCIN issues, they were invented to deal with cable interference issues, hence the shielded, twisted pair standard. Balanced components are not as immune to SCIN as you seem to think either.

 

[amirm]

That's absolutely not true, you do not have a correct picture of what is happening in a single ended system. Your whole post is misinformation... It IS true that the ground on single ended cables carries a lot more than just the return signal, but under normal circumstances it doesn't couple to the signal, otherwise single ended systems would be completely unuseable.

Don't know why you say this. I showed that with 25 feet of cable and nice Belden cable, you can achieve better than 80 db signal to noise ratio. I don't call that unusable by any stretch. It exceeds the signal to noise ratio of my Reel to Reel tape deck.

And it is not a matter of coupling. This is not induced external noise. The current simply needs a return path and it will use the shield in the interconnect cable. It will do that whether one likes it or not.

The truth is, many people happily own and listen to single ended systems with minimal noise.

That's true. Try to keep your cable lengths short, and minimize differential between outlets.

One of my systems is ridiculously quiet, you can't hear noise from a 102 dB speaker with your ear INSIDE THE HORN. And that's with separate components, a tube pre and amp, that I built myself using simple star grounding.

Take that exact system and put in someone else's home and you may easily drop many dbs due to use of unbalanced interconnect. Use balanced and it will not be location and use dependent.

If you woke up tomorrow and had a clean sheet of paper, you would design an audio interconnect where it would rely on vagaries of how equipment is wired to mains and how much current leaks to their chassis???

 

[PeterA]

This detour into cable typology is interesting and the more I read the passion with which different opinions are expressed, the more I think that this somehow reflects the state of the High End as it represents how strongly entrenched certain camps are in their own positions. It is a large tent, but there are many different tables inside.

 

[ack]

As much as I have respect for Roger, this is a prime example of how this topic is misunderstood by even engineers. When we talk about noise, there are many versions of it. One for example is noise being picked up by the cable. In this discussion, we are *not* talking about that noise. Nor are we talking about noise inside of your electronics.

What we are talking about is that by mere fact that two power cords in two separate pieces of equipment creates two different ground potentials (voltages), by definition if you hook those two chasses together, current moves between them. If you then proceed to use unbalanced connection, this current becomes part of your signal. And because it runs at AC mains frequencies, it translates into hum.

This is *totally* induced noise by the user/system. It is on top of any and all other sources of noise. Its reason for existence cannot be eliminated. No matter what you do, the two chassis will be at different voltages. It is a problem created when we went from all-in-one systems to components. This problem was recognized and balanced connection were invented to eliminate it (plus other benefits I won't get into).

No goodness to a user comes from such a connection scheme. You are forcing the system into producing noise. Why on earth would you want to do that? You can't with a straight face.

Now if you are a designer/manufacturer, you have many reasons to want to avoid building a fully differential system. Such is life in mass market where component cost is everything. But in a $20,000+ amplifier or component? No way do I tolerate such compromises.

Given me both options and let me choose if it is not the cost that took balanced out of the system. Otherwise the arrow points directly at cost cutting. Or lack of understanding of the problem here which unfortunately is prevalent.

I say if there is one topic engineers like to forget about when they leave school is this one. It makes their heads hurt when they are taught the theory and walk out confusing all manner of grounding, transmission, emission, etc. This is behind so many bad designs and bad advice online. I myself had to relearn all of this.

As DaceC also said, the highlighted is just simply not true: you get rid of hum induced by ground loops by going balanced, and the proof is in the pudding. One of the points of balanced connections is to decouple signal ground from chassis ground and ground loops there, isn't it?

Also, regarding your first sentence about Roger, "this is a prime example of how this topic is misunderstood by even engineers", I find nothing exhibiting a misunderstanding in Roger's article. Yes, he's talking about noise picked up by interconnects cables and you are talking about ground-induced noise - that doesn't make him or you wrong; you are both covering different types of noise. He's also talking about the added complexity of going balanced and why the two phases' circuits must be matched, and be a mirror of each other, to get the total desired effect of noise cancellations. He refers to such mismatches between the two circuits for either phase as "distortions", in order to put it in layman's terms. None of what he writes is him "misunderstanding" anything. I think he's exactly right.

 

[amirm]

As DaceC also said, the highlighted is just simply not true: you get rid of hum induced by ground loops by going balanced, and the proof is in the pudding. One of the points of balanced connections is to decouple signal ground from chassis ground and ground loops there, isn't it?

I am sorry but I am being dense this morning as you are agreeing with me yet you say Dave was right???

Also, regarding your first sentence about Roger, "this is a prime example of how this topic is misunderstood by even engineers", I find nothing exhibiting a misunderstanding in Roger's article. Yes, he's talking about noise picked up by interconnects cables and you are talking about ground-induced noise - that doesn't make him or you wrong; you are both covering different types of noise. He's also talking about the added complexity of going balanced and why the two phases' circuits must be matched, and be a mirror of each other, to get the total desired effect of noise cancellations. He refers to such mismatches between the two circuits for either phase as "distortions", in order to put it in layman's terms. None of what he writes is him "misunderstanding" anything. I think he's exactly right.

My objection to his paper was that in the same sentence he mixes two different aspects of the topic at the same time. Notice how I never did this in my explanation. I made no mention of RFI, equipment noise, etc. These are independent topics. The key here and the revelation for me was being able to decouple of them and with it, simplify the topic into simple application of ohm's law. This is the brilliance of Bill Whitlock and the gift of understanding he has given us. And why it is impossible to challenge him on this. It is not like he is applying field theory and someone can look for mistakes in his math. It is simple act of measuring currents and resistance, multiplying and getting your induced noise voltage.

As to Roger talking about difficulty of designing balanced gear, that is his problem, not mine as a customer . There are plenty of high-end equipment with exemplary performance using balanced topology. So unless he can prove impossibility, as a selfish consumer, I don't care .

Read any proper paper on this topic and you see enragement and rant about lack of understanding of this topic. When I managed development of computer systems, we would always hire expensive consultants to advise on such things as grounding, EMI/RFI, etc. I would not trust my army of senior engineers to know any of this properly and they knew it just the same.

Anyway, the point here is not huge. If you are using short interconnects, unbalanced is fine. I just like to see high-end equipment where I have paid 10 to 100 times more for than mass market, give me the option of balanced. Let's not murder each other over it .

 

[amirm]

Way ahead of you Amir. Jim Smith measured this a couple of years ago when he was here. And I just remeasured now. I first tried my digital Radio Shack meter, but it does not go below 60dB. It just reads "low". So I tried the "SoundMeter" app on my iPhone and just got a reading of 27 dB (A weighted) and 29 dB (C weighted) between wind gusts at my listening position. There is currently a winter storm, so it is very windy. It peaks at about 36 dB during loud gusts. At night, when listening, I have measured 25 dB. The house was built in 1795 and my listening room is on the first floor near a street corner, though there is not much traffic. Being in the historic district, I can not install insulated double pane windows, so I have only historically correct single pane true divided 6/6 sashes. I did replace the glass with laminated glass to cut some of the noise when restoring the windows. I'm sure other rooms are much quieter.

Additional information is that at normal listening levels I do not hear any noise out of my speakers from the listening seat. At 1" from tweeter or mid range driver, the noise level is 29 dB (A) and 31 (C) at normal listening volume, so about a 2 dB gain when the system is powered on relative to system off. If I turn the volume to max, I can hear 37 dB (A) and 40 (C) noise 1" in front of the tweeter and midrange drivers, but I can not listen at this volume as it is way too loud.

I've never heard of a room approaching 0dB SPL. What is the measurement of your room? And do you know what the noise floor measurement is for the typical listening room?

Your enthusiasm for using measurements to back the performance your system is palpable. It is a powerful tool to cut through arguments, isn't it?

Alas, this is one situation where you don't want to get ahead of me. SPL numbers in your room are of zero value in determining how noisy your room is *to your ear*. Your ear is highly non-linear with respect to frequency range. Its sensitivity dramatically changes depending on the same frequency. Not only that, the sensitive curve itself is level sensitive as we see in the classical Fletcher-Munson listening tests:

Take the bottom curve that is the threshold of audibility. 70 db sine wave at the 20 Hz is silent just as nearly -10 db is at 3000 Hz!!!

Your SPL meter is "dumb." It gives you a sum total, single value number, for some unknown spectrum further modified by the weighting you used. If there is a freeway rumble five miles away, you would get it to show 45 dB at 20 Hz even though to your ears that would be totally silent.

I explain all of this in my article that I linked to earlier: http://www.madronadigital.com/Library/RoomDynamicRange.html

You have to get the full spectrum, then convert the noise into equivalent tone power and then compare it to Fletcher-Munson graph.

So erase this commonly talked about metric from your vocabulary. Use SPL meter to set up sub levels and such but not how noisy your room is.

 

[ack]

I am sorry but I am being dense this morning as you are agreeing with me yet you say Dave was right???

Apologies, I misread your post. I thought you said using "balanced" connections one would still get hum. But you said "unbalanced", and yes, in that case, one is open to hum. Sorry!

 

[PeterA]

Your enthusiasm for using measurements to back the performance your system is palpable. It is a powerful tool to cut through arguments, isn't it?

Alas, this is one situation where you don't want to get ahead of me. SPL numbers in your room are of zero value in determining how noisy your room is *to your ear*. Your ear is highly non-linear with respect to frequency range. Its sensitivity dramatically changes depending on the same frequency. Not only that, the sensitive curve itself is level sensitive as we see in the classical Fletcher-Munson listening tests:


So erase this commonly talked about metric from your vocabulary. Use SPL meter to set up sub levels and such but not how noisy your room is.

Good grief, Amir. You asked me to take an SPL noise level of my room with an iPhone app. I did so. I never claimed it would be close to 0dB. I wrote that I was way ahead of you simply because I had done this two years ago with Jim Smith, meaning precisely that, that I had already done this before you asked me to do it. I did it again yesterday with the same results. It is not a question of wanting to get ahead of you. Your tone is like a veiled threat or something, kind of weird.

For some reason, I think you are the one grinning. These exchanges are very tiresome. And you never responded to my question of what the noise level in your room is with a similar iPhone app or what the reading would be for a typical listening room. Please answer those questions for me. Data is a powerful tool to cut through arguments. What is the noise level in your room using a similar iPhone SPL app?