[b0dhi]

Quote:

Originally Posted by Koyaan I. Sqatsi /img/forum/go_quote.gif You can't quite say it was the same amplifier. It may have all been in the same box, but what was feeding the headphones in balanced mode wasn't exactly what was feeding it in single ended mode.

You do realise that this is the definition of the measurement? Ofcourse what was feeding the headphones in balanced mode was not exactly what was feeding it in SE mode - if it was, it would be the same mode.

Also, as far as I know, the amp uses half the amplifier circuitry when in SE mode and both halves when in balanced mode. So really, that is as close as you can get to measuring balanced vs SE while keeping all other factors the same as far as real-life performance goes.

 

[Steve Eddy]

Quote:

Originally Posted by b0dhi /img/forum/go_quote.gif You do realise that this is the definition of the measurement? Ofcourse what was feeding the headphones in balanced mode was not exactly what was feeding it in SE mode - if it was, it would be the same mode.

I think we need to back up here and ask the question, what exactly was intended to be evaluated?

Was it to evaluate the performance of a particular amplifier and its two modes of operation, or was it to evaluate the behavior of a headphone in response to being driven from a balanced source versus an unbalanced source?

I was under the impression it was the latter, and if that was indeed the case, then you can't make a proper evaluation of that if the headphone's being driven by two different source impedances or two different output levels.

Quote:

Also, as far as I know, the amp uses half the amplifier circuitry when in SE mode and both halves when in balanced mode. So really, that is as close as you can get to measuring balanced vs SE while keeping all other factors the same as far as real-life performance goes.

You can get closer if you assure that the headphone is being driven by the same source impedance in both cases. And it's not evident that that was the case here.

k

 

[b0dhi]

Quote:

Originally Posted by Koyaan I. Sqatsi /img/forum/go_quote.gif You can get closer if you assure that the headphone is being driven by the same source impedance in both cases. And it's not evident that that was the case here.

And you get closer still if you change the topology back to single-ended. Source impedance is a characteristic of the drive mode. What you're suggesting is like saying you should match the source impedance of a tube amp to compare it to a solid state amp. Doing so would alter the performance of the thing you're trying to test and defeat the purpose of the measurement.

Moreover, the FR changes seen in that graph don't look at all like it's the result of a change in source resistance.

 

[johnwmclean]

Balanced topology is the optimum configuration over single ended, whether it's audible/measurable is debatable. In theory it offers advantages over single ended, and certainly is not a step back - there are only advantages to be had.

One certainly cannot argue the merits of single ended, if you had the choice to choose one over the other.

 

[xnor]

Quote:

Originally Posted by johnwmclean /img/forum/go_quote.gif Balanced topology is the optimum configuration over single ended, whether it's audible/measurable is debatable. In theory it offers advantages over single ended, and certainly is not a step back - there are only advantages to be had. One certainly cannot argue the merits of single ended, if you had the choice to choose one over the other.

But that's exactly what the OP was asking for, "Can one hear differences between the two?".

In theory a lot of things have a lot of advantages over a lot of other things, but we're trying to figure out if you really can hear advantages.

 

[johnwmclean]

Quote:

Originally Posted by xnor /img/forum/go_quote.gif But that's exactly what the OP was asking for, "Can one hear differences between the two?". In theory a lot of things have a lot of advantages over a lot of other things, but we're trying to figure out if you really can hear advantages.

Which brings us back to subjective hearsay, even if measurable audible differences are unreliable. I have subjective opinions in regards to my listening with balanced but what good is that.

 

[xnor]

DBT is the keyword here.

 

[Steve Eddy]

Quote:

Originally Posted by b0dhi /img/forum/go_quote.gif And you get closer still if you change the topology back to single-ended. Source impedance is a characteristic of the drive mode.

But it's not an inherent characteristic of the drive mode. And there's nothing preventing one from maintaining the same output impedance in both balanced and single ended modes.

Quote:

What you're suggesting is like saying you should match the source impedance of a tube amp to compare it to a solid state amp. Doing so would alter the performance of the thing you're trying to test and defeat the purpose of the measurement.

And this goes back to what I'd asked previously.

What is the purpose of the measurement?

Is it to evaluate headphone amps, or is it to evaluate headphones vis a vis their being driven from a balanced or single ended source?

Again, I've been under the impression it is the latter. And if so, you can't make a meaningful evaluation unless the only thing that changes is drive mode, and not other things such as drive level and/or source impedance.

Quote:

Moreover, the FR changes seen in that graph don't look at all like it's the result of a change in source resistance.

Perhaps it's not. But without knowing what the source resistance was for both instances, it can't be ruled out.

And by the same token, there's nothing about balanced versus single ended drive that would account for the change provided source resistance and levels were kept constant.

k

 

[Steve Eddy]

Quote:

Originally Posted by johnwmclean /img/forum/go_quote.gif Balanced topology is the optimum configuration over single ended, whether it's audible/measurable is debatable. In theory it offers advantages over single ended, and certainly is not a step back - there are only advantages to be had.

But from what perspective?

From the perspective of the amplifier, or the perspective of the headphone?

I got the impression that it was to be from the perspective of the headphone. And from that perspective, I see no particular advantage at all, save perhaps for common-mode noise rejection, which arguably isn't an issue at all.

And if it's to be from the perspective of the amplifier, then I think we should really drop the word "balanced" and instead use the much more appropriate term, "bridged," as that's how most headphone amps achieve what they call "balanced" and because bridging doesn't offer the one benefit that has traditionally been the raison d'etre of balanced interfaces long before people started making "balanced" headphone amps.

k

 

[upstateguy]

One would think that high end manufacturers like Luxman would be aware of any advantages "balanced" might have to offer and include it in their new $2-3K headphone amplifiers.

One might also think that the pro world would have adopted "balanced" for their headphone monitoring gear, but that does not seem to be the case either.

It seems as though the Japanese market and the pro market know "6dB snake oil" when they see it and although "balanced" has its place connecting gear to gear, they have avoided it for headphone use.

Unfortunately, as more and more people get taken in by the volume increase, manufacturers will see the opportunity to make expensive amplifiers with twice the electronics to take advantage of it.

In making decisions about "balanced", a good rule of thumb is to follow the dollar, especially when an entire cottage industry, devoted to making money from "balanced" and "balanced accoutrements", is heavily promoting a particular placebo.

USG

 

[b0dhi]

Koyaan, going from SE to balanced mode halves the output impedance. Low output impedance is a sought-after feature of an amplifier. Artificially increasing this impedance for the test is modifying the amplifier and does not reflect real-world performance. That's as straightforward as I can make it for you.

 

[Steve Eddy]

Quote:

Originally Posted by b0dhi /img/forum/go_quote.gif Koyaan, going from SE to balanced mode halves the output impedance.

No, it doesn't. It doubles the output impedance.

And even that's not anything inherent but rather a matter of implementation.

For example, You can take what would otherwise be a single-ended amplifier and hitch it to an output transformer, giving you a balanced output. Short the bottoms of the primaries and secondaries, effectively turning the transformer into an autoformer, and now you've got a single ended output. But the output impedance will be identical in both cases.

Quote:

Low output impedance is a sought-after feature of an amplifier. Artificially increasing this impedance for the test is modifying the amplifier and does not reflect real-world performance.

That depends on what exactly it is you're "testing."

If what you're testing is a headphone's behavior when driven from a balanced versus single-ended source, which is what I understand the issue to be here, then you need to assure that all else remains equal. Including output impedance.

k

 

[b0dhi]

Quote:

Originally Posted by Koyaan I. Sqatsi /img/forum/go_quote.gif No, it doesn't. It doubles the output impedance. And even that's not anything inherent but rather a matter of implementation.

I stand corrected about the half/double.

Let's take a step back here. Tell me if I understand you correctly: -

You seem to be coming from the position that everything other than the variable being tested should remain constant, and that there should only be one variable changing. I completely agree with that.

Then you seem to be saying that the "variable" being changed is the balanced/SE operation. I do agree with that, but this is where things get murky.

You yourself said that the output impedance can double (depending on implementation). So the question is; do we include this particular change as part of the "variable"? You seem to be saying no, we shouldn't include this as part of the "variable" because it isn't necessary that the output impedance be a certain value for the operation to be balanced, therefore we should keep it constant. I.e., it isn't inherent to balanced operation so it should be kept constant.

If that's correct, then: -

If the output impedance changes due to going from SE to balanced (whether it halves or doubles is irrelevant), that is a propogative effect of the variable under test. If you then change the output impedance, you are no longer changing only one variable in the test, you are changing two - the balanced/SE operation, and the output impedance. The fact that you're matching the output impedances does not change the fact that you're altering the system.

Note that I'm not saying it's impossible that output impedance is causing audible differences between bal./SE. What I'm saying is that the experiment should start out by first capturing the effect, then refining the test to determine the cause.

 

[DarKu]

In theory tube based amps and dacs have a lot worse THD, and bigger distorsion on the paper, however a lot of people do enjoy them more than solid state dacs or amps

Not every time measurements is the only thing to believe, i believe my own ears, others? I don't know
The biggest advantage i see going balanced is double the Voltage swing, but....some SE amps have the same V swing as some balanced amps...
I need to hear a balanced amp first to believe it...or not

 

[Currawong]

Quote:

Originally Posted by DarKu /img/forum/go_quote.gif In theory tube based amps and dacs have a lot worse THD, and bigger distorsion on the paper,

Whose theory and what paper?