I am struggling to believe that people don't understand this TBH.
Imagine this:
You go to a test where you are asked to decide which of two tiles is black and which is white. There are 100 people in the audience. The tester holds up a tile. It is white. He asks for a vote. 100% say that tile A is white. He holds up the next tile. It is black. 100% of people say tile B is black. This is because there is a clear difference between A and B that everyone can see. Repeating the test gives the same result no matter how many trials are made.
Now, go to an amplifier test - same 100 people in the room. The tester plays two amplifiers for you. He doesn't know which is A or which is B (it is a 'blind' test).
He asks if anyone can hear a difference between A and B. 100% of people in the room say they can hear a difference. He plays another amplifier and asks the audience to say if it is A or B. (They just listened to both remember and said they can identify A from B). Some people say it is A and some say it is B. Someone (or a computer) makes a note of the number of people who say it is A and the number who say it is B.
He then plays another amplifier (he doesn't know which it is remember) and he asks the audience if it is A or B. Again, some say it is B and some say it is A. Note that it is not necessarily the same people who said A or B before.
He then plays another amplifier and asks if it is A or B. Nobody knows at this stage which is is (blind test right?) - it might be A or B. Again, some people say it is A and some say it is B.
Sometimes, because of the random selection of A or B, B is played 4 times in a row, say. People are asked which it is each time. Some say it is A and some say it is B. This goes on for a while and then test is finally over.
At the end of the test, the scorekeeper reveals the results. The first test the amp playing was A. Roughly 50 people said it was A and roughly 50 said it was B.
The second test the amp playing was A again. Roughly 50 people said it was A and roughly 50 said it was B. Note these are not the SAME people each time who are saying "it is A" or "it is B". It is just 50 people.
The third test the amp playing was B. Again 50 say it is A and 50 say it is B.
This pattern is repeated for all the tests. What conclusion can we draw? The answer is that we have to conclude that nobody could reliably tell A from B. If they could have, just as with the black and white tile test, a very high number would consistently get it right.
If A was definitely different to B (say they believed it had a 'warmer' sound) then they would identify that warmer sound most of the time - so the scores would be more like 80% against 20% each time. The fact that the scores were more like 50% against 50% each time shows that people cannot reliably tell A from B. IOW, if people hadn't even tried to tell one from the other and had just picked A or B each time, at random, you would still get roughly 50% picking A and 50% picking B. That is the probability when there are only two choices, just like if you toss a coin 100 times, it will come down heads roughly 50 times and tails roughly 50 times (try it and see).
I hope this has helped explain it.
With headphones, the room has been removed from the equation. That is where the real difference is, rather than the difference between one amp and another.
I know nothing about headphone amps BTW, but assume that that they have reached a stage, like other amps, where there is little to choose, SQ-wise, between them. It would be interesting, as you say, to see a blind ABX test of headphone amps. Removing the influence of the room should make it easier to hear the differences, if there are differences.
It's the last time I promise you :)
It's not the love of audio... rather the endless desire to lecture and all of what that entails... in their mind. The simplistic theory simply gives them a platform to stand on and disregard the real world.
You are very welcome.
Yes, I agree that the outcome could be different. It's just that, in all the dozens of reports of AB tests I have read, it never is. They have all been inconclusive and come in at or around the same result that random chance gives. I think you would be surprised if you could ABX your monoblocks against the AVR amps, so long as the latter are working to spec and not clipping and are able to drive your speakers correctly (ie they can handle a 4 ohm load if you have 4 ohm speakers, or a difficult load as presented by some 'audiophile' brands).
I am not sure if I wish this was different and that all amps sound different to each other, or if I am happy in the knowledge that they don't. If the former, I could know that I would get an improvement in SQ by spending more on a 'super amp'. If the latter, I am happy to know that I don't need anything other than my Emotiva amps because nothing else will give me a sonic improvement.
At the end of the day, it doesn't really matter much because I know that the real way to get BIG improvements in SQ is to get really, really good speakers (which I believe I have done) and really, really good subs (which I have definitely done) and concentrate on measuring and treating the room. Especially the latter as it is so cheap, comparatively, to do. That is where the real changes will be heard - not in modern electronics.
There are dozens of similar tests all over the net - I posted a link to about 50 recently in this thread. They all come to broadly the same conclusion - nobody can reliably detect a difference between the amps under test. Some tests may have flaws, but not all of them.
Incidentally, sample size in a survey isn't all that important so long as the sample is representative of the whole. That's why polling companies can predict elections etc very accurately by surveying 1000 people, even though the universe is over 100 million (in the US).
Incidentally again :) - that the audience had preconceived biases helps the test's validity not harms it. Even with their bias or predeliction to prove that amp A is different to amp B, they still couldn't reliably hear any differences.
Originally Posted by Charles R
It's not the love of audio... rather the endless desire to lecture and all of what that entails... in their mind. The simplistic theory simply gives them a platform to stand on and disregard the real world.
It's neither in my case. It's a genuine desire to help people via the medium of these forums, in the same way that I have been helped so much by them in the past. It's called "giving something back". Of course, everyone is entitled to their views and beliefs and I have never said differently. The real issue is that you will find people on these threads who will pass of their opinions as facts. So they say things like "All amps sound different to each other and I know that for a fact because I can hear the differences with my own ears".
Now this is not only wrong, but it doesn’t even enter the arena of sensible discussion because someone's opinion is just that - an opinion. It's like me saying that "I know all French people smell of garlic because I have smelled it with my own nose". People who say these things will never offer any objective evidence to support their position (usually because there isn't any) and will just repeat, ad infinitum, that it must be so because their ears say so. Worse, when presented with objective evidence that they are wrong, they will refuse to accept it and cling to their original position regardless. Even to the extent of saying "You will never convince me otherwise". IOW, their mind is made up that they must be correct n matter what.
Why does this matter? Because for every person who contributes to a thread, there are many more who silently read it. And many of them are looking for information to help them with their purchasing decisions. And many of them are on tight budgets. And many of them have very little knowledge of audio or acoustic science. So when they read statements that tell them that one of the most important things they can do to improve their SQ is to buy a new amp, many of them will go out and do that. They will do this even though it is almost always a waste of their money and they would benefit far more from rearranging their rooms or adding some simple room treatments, at very low cost if they are made on a DIY basis (which is very easy). So I, personally, think it is important to provide the right information, backed by objective science, in order to provide a counter-perspective. Hopefully, some people will read this and think "hmm, maybe that $2,000 amp would be a waste of money after all" and they might investigate other ways which have been objectively proven to yield what they are after: better SQ.
So I, personally, do object to being told I am on a "crusade" or have "an endless desire to lecture others" because neither position is remotely true and you can verify my aims and intents on AVS by searching my posts, especially in the Audyssey thread where I usually hang out. To offer objective, evidence-backed views is not disregarding the "real world" in any way IMO. It is very much trying to help one of the most 'real world' issues there is: spending one's hard-earned cash wisely.
