External DAC with Connect

What is the truth, the philosopher would ask?
Every scientific fact is nothing but a hypothesis that is yet to be proved wrong, to be then substituted by a better one. The all swans are white hypothesis/fact is easy to be proved wrong by finding just the one black swan - but where is that in this case of amplifiers/DBTs?
Surely if you are going to make such postulations, they remain just speculations until you find the black swan? While the white swan hypothesis has been constructed by many people that have done these tests over the years and are satisfied to its quality. Anyone that doesn't agree has the burden of finding the black swan to destroy it.
Over at Harbeth UK, Alan Shaw for some time now has offered a free pair of his flagship speakers to anyone that can turn up one in controlled conditions, with any two amp of the claimants choice that are in working condition and meet a frequency response/distortion criteria that almost all modern ones do. No downsides, no entry fee, except for travel costs. And yet he has seen no takers, not even from people living in the UK where travel costs/time may not be the investment it would be for people elsewhere. Where are the black swans hiding?

You've conveniently elided my following comment to the effect that I don't really care who has the burden of proof. It's not important to the thrust of my argument, which is that both sides of the debate are actually discussing beliefs (about the real-world behaviour of engineering and technology) rather than scientific fact.

You've conveniently elided my following comment to the effect that I don't really care who has the burden of proof. It's not important to the thrust of my argument, which is that both sides of the debate are actually discussing beliefs (about the real-world behaviour of engineering and technology) rather than scientific fact.
I am sorry I misunderstood you to mean that the side arguing for conclusion validity based on controlled level matched DBTs has to prove the negative as well. I must have misread:D.
Leaving aside the question of whether this approach is based on a belief/fact/hypothesis, what exactly is the statement of the other side of the debate, never mind what it is called?

Well, Tell Alan Shaw to come to America with that offer, myself and millions of others would be glad to relieve him of some free Harbeths.

God knows, science never got anything wrong either. Logic would dictate that even if all amps where created equal, there would still be perceived differences because audio is certainly subjective and hearing itself is different from person to person.

Logic would dictate that even if all amps where created equal, there would still be perceived differences because audio is certainly subjective and hearing itself is different from person to person.
If you had said that in the above case there would still be perceived differences because of speaker load/behaviour differences, I'd say yes.
But to what you have said above, all I can do is shake my head in wonder:D.

Originally Posted by Kumar
Consider the two statements:
1. Two amplifiers that have the same measured flat frequency/distortion response of the kind that most modern amplifiers exhibit, that are working within designed limits - i.e, without clipping/distorting - cannot be distinguished from each other in a level matched DBT when operating with tone controls or the like not engaged.
2. All amplifiers sound the same.
People, of whom I am one, that support the first statement are often said to be saying the second, which they are not. The second statement is not a short version of the first, and isn't believed to be a true statement by them.
I am curious to know if there is anybody on this forum that disagrees with statement 1.

It is interesting - and to me revealing - that the only responses to the above display inadequate reading comprehension or deflect follow on explanations with a stab at humour.
OK, here are my comments.

Firstly, I think you have to talk about acoustic and electrical noise as well as distortion. Otherwise mains hum (for example) could be a trivially easy way to distinguish between amplifiers that satisfy the criteria you listed.

Secondly, in a later post you gave a list of amplifiers that you felt satisfied those criteria and had "modern levels of distortion" (to paraphrase). These included some solid state amplifiers and the AudioNote Ongaku. Looking at the specifications for that amplifier on AudioNote's web page, they rate it at:

27W + 27W @1kHz, 5% THD

That's not a typo, 5%. Are you saying that you believe 5% THD is not distinguishable from the typically sub-0.1% THD specification of a solid-state amplifier?

Also, the amplifier's distortion components will have a signature that could be used to distinguish between amplifiers. One reason people tend to like the sound of valve amps such as the Ongaku is that they have high levels of even-harmonic distortion.

Leaving that aside, as I understand it, amplifier specifications are usually quoted when running into ideal (resistive) loads with a steady-state sinusoidal input. When the same amplifier is connected to a loudspeaker in the real world to play music, it is likely to behave differently. The loudspeaker is a much more complex load whose characteristics change with frequency. The music programme will have a much higher crest factor than a sinusoid, which probably stresses the amplifier (power supply) much more. Also because it isn't steady state, thermal effects are likely to be more signficant.

So in order to be relevant to music reproduction rather than behaviour in a lab, I think your test criteria should state that the amplifier is playing a music programme into a given set of loudspeakers. My belief is that this would significantly narrow down the list of amplifiers that can fulfil the criteria for comparision. (Sad to say for ErikM I expect the Rotel vs Ongaku shoot-off would be ruled out pretty quickly.)

Originally Posted by Majik
Things have changed: these days building an amp that is, for practical purposes, "transparent" can be done by putting together a handful of off-the-shelf components.

This isn't recent and the amplifier problem was solved by engineers and manufacturers about 25 years ago. Recent real progress has been to drive down the cost of the solution to very low levels.

That's a very sweeping generalisation. What evidence do you have to support it? What happened 25 years ago that made the difference (I am genuinely interested)?

There are still interesting discoveries being made about how audio amplifiers behave and should be designed. For example, going back to the topic of clipping that you mentioned, have you read Roger Sanders' white paper that discusses how much power headroom is required to avoid clipping typical music content? Here is a nice extract:

You will find that conventional, direct-radiator (not horn-loaded), magnetic speaker systems of around 90 dB sensitivity, require around 500 watts/channel to avoid clipping. More power is needed in larger rooms or if you like to play your music more loudly than most.

The key point I'm trying to make is that audiophiles usually are using underpowered amplifiers and are therefore listening to clipping amplifiers most of the time. When an amplifier is clipping, it is behaving (and sounding) grossly differently than its measured performance would suggest. This is because we always measure amplifiers when they are operating within their design parameters -- never when clipping. A clipping amp has horrible performance, so attempting to measure it is a waste of time.

In other words, we usually listen to an amplifier when it is clipping and we measure it when it is not. This is why amplifiers sound so different than their measurements would imply. It is not that measurements are wrong, it is simply that we are listening and measuring different conditions.

I don't own any Sanders products and don't have any affiliation to the company (other than admiration for the ingenuity of their design), but I highlight that extract simply to show how deceptive real-world behaviour of audio amplifiers can be. It was a surprise to me to discover that so much power was needed to avoid clipping under typical listening conditions.

The DAC problem was similarly solved within a few years of their appearance on the audio scene. In some ways, it is a simpler problem - none of the complex and dynamic speaker load interactions to be dealt with.

Again, that might be your opinion but I didn't see you put forward any supporting evidence. I suspect many electronic engineers working in the field might disagree with you.

The cable subject does not merit discussion. Along with green magic markers.

Your opinion again. jgatie appears to disagree since he's taking part in a discussion about cables in this very thread...

This is the internet though. I am sure there are still places on the net where the view that the Apollo missions landing on the moon was a hoax is actively debated.

Agreed. But completely irrelevant, isn't it? I know you're joking but it's attempting a slight at the same time.

I am sorry I misunderstood you to mean that the side arguing for conclusion validity based on controlled level matched DBTs has to prove the negative as well. I must have misread:D.
Leaving aside the question of whether this approach is based on a belief/fact/hypothesis, what exactly is the statement of the other side of the debate, never mind what it is called?

To keep it short: the statement is simply that cables may make an audible difference under some circumstances.

To me this seems more plausible than the converse statement, that there are no circumstances under which a cable can possibly make any audible difference.

To keep it short: the statement is simply that cables may make an audible difference under some circumstances.

To me this seems more plausible than the converse statement, that there are no circumstances under which a cable can possibly make any audible difference.

Again, I clarified that statement. I specifically said that digital cables, when operating properly, will show no discernable differences in a properly conducted DBT. By "operating properly" I mean both they and the system being used are exhibiting no errors and are properly grounded/shielded. Of freaking course a bad cable or bad system will sound different than a system/cable operating properly!

I also clarified that I am talking differences between generic, but to standard, cables and high-end boutique brands costing astronomical amounts.

Again, I clarified that statement. I specifically said that digital cables, when operating properly, will show no discernable differences in a properly conducted DBT. By "operating properly" I mean both they and the system being used are exhibiting no errors and are properly grounded/shielded. Of freaking course a bad cable or bad system will sound different than a system/cable operating properly!

I also clarified that I am talking differences between generic, but to standard, cables and high-end boutique brands costing astronomical amounts.

Yes, that's why I pointed out that I was keeping things short.

The difficulty is with your phrase "when operating properly". I think you are using it as a proxy for "when operating ideally", which turns your statement into a truism of the form "a component of type X, when operating as an ideal X, makes no difference". (I feel Kumar's hypothesis about amplifiers is heading in the same direction.)

If you're not using the phrase in that sense, you must define exactly what you mean by "exhibiting no errors" and "properly shielded/grounded", neither of which is trivial. Your statement then becomes much less general, and much more system-dependent, because you then have to demonstrate that there are in fact no errors (for example) rather than "infrequent errors".

To keep it short: the statement is simply that cables may make an audible difference under some circumstances.

To me this seems more plausible than the converse statement, that there are no circumstances under which a cable can possibly make any audible difference.
I owe you a reply to a longer earlier post, but this reply will help keep the debate on track.
First, let me address what you say is the converse. That isn't what is being said. Use too thin a cable core for a speaker cable run and it will make a difference. Just one example.
Second, to your statement of position, I would only ask two questions - if they appear to audibly do so, will this difference always survive a controlled DBT? What would be your response if it doesn't so survive?

Yes, that's why I pointed out that I was keeping things short.

The difficulty is with your phrase "when operating properly". I think you are using it as a proxy for "when operating ideally", which turns your statement into a truism of the form "a component of type X, when operating as an ideal X, makes no difference". (I feel Kumar's hypothesis about amplifiers is heading in the same direction.)

If you're not using the phrase in that sense, you must define exactly what you mean by "exhibiting no errors" and "properly shielded/grounded", neither of which is trivial. Your statement then becomes much less general, and much more system-dependent, because you then have to demonstrate that there are in fact no errors (for example) rather than "infrequent errors".

Why don't you clarify what you mean? Do you believe that digital cables can be "more revealing", cause "greater soundstage", "crisper highs", or cause one to describe the sound with other esoteric adjectives such as "wooden", "fuzzy", "orange peel", etc? Or do the transmission errors you describe affect the music in more deleterious ways? Do you think spending $1000 produces any audible gain over a properly constructed cable costing $10?